Search

Reliability of blades has a profound impact on both serviceability and safety of offshore wind turbines. Reliability estimation of wind turbine blade is a complex and time-consuming problem with multi-physics coupling and multi-failure modes correlation. Developing physics-of-failure modeling and reliability analysis methods with high efficiency and accuracy is a long-term challenge. In this work, a high availability and cost-effectiveness reliability estimation framework for offshore wind turbine blade by combining transfer learning (TL) and direct probability integral method (DPIM) is proposed. Extensive performance simulation of offshore wind turbine blade is a complex and necessary task, this paper develops a new adaptive sampling strategy to improve the validity of sample selection in the design space; On this basis, a physics-of-failure surrogate modeling approach is proposed by introducing TL method to fuse two kinds of multi-physics coupling analysis data, then the performance of all critical loads can be predicted efficiently in advance to provide a reliable design; Further, this paper provides an efficient reliability estimation method for offshore wind turbine blades by combining DPIM and surrogate model. Finally, the validity of the proposed approach is illustrated by numerical example and offshore wind turbine blade reliability estimation. The proposed framework provides a cost-effective alternative to higher loads simulation efforts and safety factors selection. • A new adaptive sampling strategy for multi-failure modeling is proposed. • A failure modeling method is developed for multi-physics coupling system. • A reliability estimation method is proposed based on Dirac delta function. [ABSTRACT FROM AUTHOR]

The existing decentralized wind farm control incorporating the wind turbine wake interaction has almost all concentrated on the condition of fixed wind speed/direction without considering the variability of wind in nature. Meanwhile, most of the current wake models cannot meet the requirement for wind turbine control optimization research with the high demand of wake calculation speed and accuracy. In this paper, a novel decentralized control strategy, named the wind-interval-based (WIB) control which adopts the uniform operation among the same interval of variable wind speeds/directions, is proposed for the control optimization study of in-line two scaling wind turbines for demonstration. To prove the effectiveness of the new control mechanism, the ideal instantaneous control is introduced for the comparative study. At the same time, a wake model incorporating the two directly controllable parameters (i.e., the tip speed ratio λ and pitch angle γ) is established based on artificial neural network (ANN). The comparative results show that the total power output applying the instantaneous control and the new WIB control mechanisms are increased by 0.1%–4.1% and 0.45%–3.9% depending on the tested wind scenarios, respectively. By comparing the optimized power output with the two controls, it is found that the error between them is generally lower than 3%, while the proposed WIB control reduces the operational difficulty to a large extent facilitating its application to the real wind turbine operation in reality. In summary, this paper shows that the new proposed WIB control not only reduces the difficulties of the wind turbine control mechanism, but maintains the control effectiveness by achieving a comparable total power output with respect to the traditional instantaneous control, which is of great significance to the wind farm developer. [ABSTRACT FROM AUTHOR]

Wind turbine power curve cleaning, by way of removing curtailment, stoppage, and other anomalies, is an essential step in making raw data useable for further analysis, such as determining turbine performance, site characteristics, or improving forecasting models. Typically, data comes as SCADA (Supervisory Control and Data Acquisition) data, so contains not only environmental and turbine performance data but also the control action imposed on the turbine by the operator. Many different anomaly detection (AD) methods have been proposed to clean power curves; however, few papers have explored filtering explicit and obvious anomalies from the SCADA prior to running AD. This paper actively explores this filtering impact by comparing the performances of 4 different AD methods with/without filtering. These are: iForest, Local Outlier Factor, Gaussian Mixture Models, and k-Nearest Neighbours. Each approach is evaluated in terms of prediction error, data removal rates, and ability to maintain the underlying wind statistical characteristics. The results show the effectiveness of filtering with every technique showing improvement compared to its unfiltered counterpart. Furthermore, Gaussian Mixture Models are shown to provide favourable accuracy whilst maintaining wind variability, however, with the wide range of performances of methods, a user's choice may be different depending on their needs. [ABSTRACT FROM AUTHOR]

Offshore wind turbines (OWTs) must withstand harsh environmental loads over their 20- to 30-year design life. Fragility analysis investigates the probability of damage over a range of hazard intensities and is integral to a performance-based engineering approach. The focus of this paper is on monopiles, which are widely used to support OWTs in water depths up to around 40m. The paper presents a fragility analysis of monopiles in extreme storms in terms of geotechnical performance, measured by permanent rotation of the foundation. Geotechnical fragility has so far not been comprehensively addressed due to the challenge of predicting soil behaviour under cyclic loading and estimating the probability of extreme responses. On the latter, the paper develops efficient Karhunen-Loeve representations of wind and wave loading that can be combined with inexpensive probabilistic methods to compute fragility. The framework was demonstrated using a representative scenario of a 5 MW OWT installed in clay. Non-linear foundation response was captured by a dynamic 3D finite element model. Fragility curves were generated using subset simulation for storms with return periods (RPs) from 1 to 100 years. Fragility during extreme storms (with 50- and 100-year RPs) was significantly higher than storms with RPs of 10 years or less. • Fragility analysis of monopile foundations for offshore wind turbines. • Geotechnical performance in extreme storms assessed by permanent rotation. • Efficient Karhunen-Loeve representations of wind and wave loading. • Fragility curves generated by subset simulation. • Failure probability increases significantly for storms with 50+ year return periods. [ABSTRACT FROM AUTHOR]

A fully-submerged bottom-standing 1.5 MW wave energy converter (WEC) equipped with a membrane-pump system is under development for installation at a test site off the coast of Pembrokeshire, Wales, UK. The system comprises several flexible-membrane air cells that are compressed and sucked by the action of the waves, rectifying valves, low-pressure and high-pressure ducts, and a unidirectional air turbine that drives an electrical generator. This two-part paper reports the design and testing of an effective control law to be implemented into the programmable logic controller of the membrane-pump turbine-generator set, allowing efficient and safe operation for the wave climate at the Pembrokeshire test site. Part 1 of this paper concerns developing the design strategy of the control laws, the numerical model, and the performance evaluation of the WEC power take-off system. A genetic algorithm is used to design control laws to maximise each sea state's WEC time-average efficiency. An adaptive control algorithm is devised by cross-correlating the control laws parameters and the rotational speed. The final result is a control law that uses input the rotational speed signal and airflow density. It adapts to the sea state conditions while maximising the time-average total efficiency of the membrane-pump WEC for deployment off the Pembrokeshire coast. • A genetic algorithm used to design control laws. • Torque versus rotational speed control-laws studied using Bézier curves. • Adaptive control devised by correlating control parameters and rotational speed. • The control law uses as input the rotational speed and airflow density. • A one-parameter adaptive control law provides good performance. [ABSTRACT FROM AUTHOR]

This paper presents an experimental study on the effects of joint wind and operating loads (JWOLs) on the long-term performance of monopole-supported wind turbines (MWTs) standing on soils. In this paper, a novel loading method that can apply ambient excitation without constraints and generate random fluctuating loads around a non-zero mean is proposed to simulate JWOLs. The method was realized using a wind tunnel. A series of wind tunnel tests with different JWOLs were designed and performed on a 1:100 MWT model of the NREL 5 MW wind turbine. The effects of the average load in the fore-to-aft (F–A) direction and the fluctuating loads in the F–A and side-to-side (S–S) directions on the long-term performance of MWTs (including the variations in the natural frequency, damping ratio, and the accumulated deformation) were studied. The test results show that the variation in the natural frequency is marginal (within −2.1% to 3.6%), whereas that in the damping ratio is within −63.9% to 38.4%. The accumulated deformation curve under different conditions can be represented using a power function. The long-term performance problems (the frequency shift and the accumulated deformation) of MWTs observed under simple harmonic loads may be high to an extent compared with those observed under certain JWOLs. [ABSTRACT FROM AUTHOR]

Systems involving more than one energy source, namely hybrid systems, can be arranged in different configurations and layouts depending on the available or used energy source, application and scale of the system. Despite an increasing awareness and penetration of renewable energy sources at different scales in many countries over the world, the adoption of hybrid and novel renewable systems is relatively scarce, especially at a small scale. This scenario is also presented from the point of view of the research of such systems. Therefore, the scope of the paper is to improve the knowledge regarding small-scale hybrid renewable energy systems performance and operation by investigating a novel trigeneration system based on a biomass fired steam cycle, wind turbine, photovoltaic panels and adsorption chiller. In the proposed system, the thermal energy recovered by the steam cycle is used to match the user thermal demand by means of a storage tank supplying directly heat to the user during winter, and driving the adsorption unit during summer. The electrical energy produced by the steam turbine, photovoltaic field and wind turbine is managed by a control system that integrates a bidirectional connection with the electric grid. The grid allows one to virtually store the electrical energy produced in excess and to recover it in part when needed by the user. In order to assess the system performance, a farm and residential buildings are considered as case study. The proposed system is modelled and dynamically simulated by means of Transient System Simulation software. In the paper, the dynamic operation of the system and the energy and economic performance of the system is assessed. The operation of the system is presented in a daily and monthly basis, while the global performance parameters of the system are reported considering a one-year period. The primary energy saving of the system is above 70% when a reference system based on natural gas, electric chiller and grid is considered. For this scenario, the Simple Pay Back of about 10 years is achieved. The results of the analysis show that the system is feasible from the technical and energy point of view, whereas its economic profitability is significantly affected by its cost and the reference system that is adopted by the user. • Modeling and dynamic simulation was performed with TRNSYS software. • Energy and economic performance of the system was assessed. • Different reference systems are considered for heating. • Primary Energy Saving ratio was between 57.4 and 73.3%. • The minimum Simple Pay Back is about 10 years. [ABSTRACT FROM AUTHOR]

Porous discs are the experimental counterparts of the numerical actuator disc concept and are frequently used in wind tunnel studies to reproduce the far-wake characteristics of wind turbines. The potential of using porous discs instead of wind turbines comes with the benefits such as operational easiness and design simplicity. This paper presents the results of an experimental study focusing on the wake development characteristics of a single porous disc and a single wind turbine under different freestream turbulence intensities but at similar Reynolds number conditions. The measurements are performed using two-dimensional two-component particle image velocimetry (2D2C PIV) technique up to seven diameters downstream distance in the wake of the models. The wake flows are analyzed in terms of the development of mean wake velocity deficit, wake decay, and wake turbulence characteristics. The results show that the wake recovery occurs faster for both the porous disc and the turbine with increasing freestream turbulence intensity. The wake width increases with the freestream turbulence intensity, indicating faster diffusion of the wake for both models. On the other hand, the wake width of the porous disc and the turbine is similar at the same turbulence intensity conditions, although they operate with slightly different thrust coefficients. Turbulent kinetic energy profiles in the far wake of the porous disc and the turbine, which are significantly different for the low turbulence intensity conditions, become similar for the high turbulence intensity case. The comparison of the turbulent kinetic energy, decay rate and production characteristics points to different turbulence mechanisms in the wake of the porous disc and the turbine. In the near wake, the wake decay rate and turbulent kinetic energy levels are higher in the porous disc case due to enhanced mixing levels associated with the grid-generated turbulence. In the far wake, the decay rate of the turbine wake is higher, most probably due to the breakdown of the tip vortices. When the freestream turbulence is increased, the differences in the turbulent wake flow characteristics diminish in the far wake, particularly after a 4.5-diameter downstream distance. [ABSTRACT FROM AUTHOR]

The paper analyses the spring-like air compressibility effect in coaxial-duct oscillating-water-column (CD-OWC) wave energy converters (WECs). This is accomplished through a novel non-linear time-domain model for OWC WECs implemented in the object-oriented language Modelica. Good agreement was observed between numerical and physical model testing results. The air chamber volume significantly affected the spring-like air compressibility effect and the converter performance. Another significant factor is the damping level of the power take-off system. Both fixed and floating CD-OWC versions were numerically investigated, focusing on the compressibility effect. Differences were found mainly due to the additional degrees of freedom in floating configurations. In general, numerical and experimental results showed that air compressibility might positively or negatively affect device power performance in regular waves, depending on whether the frequency is within or out of an interval defined by critical frequencies. Some particular points were observed and categorised as Equicompressum Nullum and Equicompressum critical points, representing different characteristics. Knowledge of these critical values might be important in OWC control. Finally, some applications of the CD-OWC concept are discussed, including considerations on the power output level provided by biradial and Wells air turbines. • A new time-domain model for OWCs implemented in Modelica language is presented. • Air compressibility effects in fixed and floating CD-OWCs were investigated. • Regions of positive and negative compressibility effects were identified. • Two types of Equicompressum critical points were found. • Air turbine selection and Blue Economy applications for CD-OWCs are discussed. [ABSTRACT FROM AUTHOR]

This paper presents a novel approach for detection and diagnosis of the rotor imbalance types pitch misalignment, yaw misalignment and mass imbalance by monitoring the drivetrain vibration response. Traditionally, only SCADA signals including nacelle accelerations, rotor speed and electrical power are utilized for this purpose, while drivetrain condition monitoring signals are mainly used for fault detection in gears and bearings. A diagnostic method is proposed using statistical change detection methods for fault detection, phase angle estimation for localizing the faulty blade, and physics-based decision criteria for fault classification. The proposed method is tested in a numerical case study with aeroelastic and drivetrain multi-body models of the 10 MW DTU reference wind turbine. The results suggest that drivetrain condition monitoring signals are particularly beneficial for detecting and diagnosing pitch misalignment, since this fault type uniquely induces periodic out-of-plane bending moments that excite drivetrain bending modes. Drivetrain signals improved the detection rate of a 1° pitch error from 19% to near 100% and reduced the standard error in locating the faulty blade from 71. 5 ° to 11. 2 °. In addition, by using drivetrain vibration amplitudes as a decision criterion, all considered pitch error cases are correctly distinguished from other fault types. • Drivetrain CMS outperform SCADA signals for the diagnosis of pitch misalignment. • The detection rate of a 1° pitch error is increased from 19% to near 100%. • The standard error in locating the faulty blade is reduced from 71.5° to 11.2°. • All pitch errors are correctly classified regardless of environmental conditions. [ABSTRACT FROM AUTHOR]

Offshore wind farms are built in coastal areas prone to have earthquakes. When offshore wind turbines (OWTs) erected on cohesionless sediments are subjected to earthquakes, liquefaction might occur in the shallow deposits that might lead to severe overturning and settlement of the foundation structure. A novel hybrid foundation structure is proposed in this paper consisting of three major components: monopile, friction wheel, and suction bucket. This paper focuses on the performance of the hybrid foundation under earthquake loads. Hybrid foundation models with different dimensions were installed on both dry and saturated loose sand deposits and subjected to dynamic geotechnical centrifuge tests. Pore-water transducers were embedded in the soil to monitor its susceptibility to liquefaction. The dynamic responses of the superstructure were also monitored by linear variable differential transducers (LVDTs) and accelerometers. The experimental results indicated that the new hybrid foundation improved the ground liquefaction resistance. The extent of improvements was more significant in saturated sand than in dry sand. In saturated sand, it was observed that the hybrid foundation with larger bucket depth densified the surrounding soil and helped to mitigate the liquefaction. Consequently, the hybrid system showed responses of a stiffer foundation where the lateral displacement and settlement were reduced significantly. • A new concept of hybrid foundation for offshore wind turbine is proposed. • Dynamic response of the surrounding soil and the superstructure under seismic load is studied through centrifuge tests. • The mechanism of the hybrid curbing the liquefaction of upper layer soil is investigated. • The effects of the dimensional parameters of hybrid foundation on the seismic-resisting performance is analyzed. [ABSTRACT FROM AUTHOR]

This paper presents an academia-industry joint case study, which was conducted to quantify and compare multi-year changes in power production performance of multiple turbines scattered over a mid-size wind farm. This analysis is referred to as a space-time performance comparison. One key aspect in power performance analysis is to have the wind and environmental inputs controlled for. This research employs, in a sequential fashion, two principal modeling components to exercise tight control of multiple input conditions—a covariate matching method, followed by a Gaussian process model-based functional comparison. The analysis method is applied to a wind farm that houses 66 turbines on a moderately complex terrain. The power production and environmental data span nearly four years, during which period the turbines have gone through multiple technical upgrades. The space-time analysis presents a quantitative and global picture showing how turbines differ relative to each other as well as how each of them changes over time. • This paper presents a case study for quantifying the turbine performance change and comparing turbines in space and time. • The method entails two major modeling components to control for stochastic environmental changes. • The outcome presents a useful decision-support for wind farm owners and operators. • The study used more than 1 GB actual wind farm operation data from 66 turbines spanning nearly four years. [ABSTRACT FROM AUTHOR]

A fully-submerged bottom-standing 1.5 MW wave energy converter (WEC) equipped with a membrane-pump system is being developed for installation at a test site off the coast of Pembrokeshire, Wales, UK. The system comprises several flexible-membrane air-cells that are compressed and sucked by the action of the waves, a system of rectifying valves, low-pressure and high-pressure ducts and a unidirectional air turbine that drives an electrical generator. This two-part paper reports the design and testing of an effective control law to be implemented into the programmable logic controller of the membrane-pump turbine-generator set, allowing efficient and safe operation for the wave climate at the Pembrokeshire test-site. Part 2 of this paper focuses on the experimental investigation at the IST variable flow test rig of the real-time control algorithm described in Part 1 of this paper, using a 1:2.4 turbine scale model. Tests were performed for five sea-states and two turbine model average rotational speeds. Experimental results reveal that the control law proposed for the Pembrokeshire mWave turbine-generator set, based only on the turbine rotational speed and airflow density measurements, is effective and of a simple implementation. Fine fitting between the model testing results validates the performed scaling of the variables related to the time-dependent turbine-generator operation, as well as the numerical model of the system described in Part 1. • The turbine-generator set real-time control algorithm was experimentally studied. • Tests were performed for five sea-states and two turbine model average rotational speeds. • The control law based on the turbine rotational speed and airflow density is effective. • Model testing results validate the scaling of the turbine-generator operation time-dependent variables. [ABSTRACT FROM AUTHOR]

The paper discusses the response and fatigue damage of the supporting tower of a small vertical axis wind turbine subject to stationary and non-stationary excitation due to wind, turbine rotation, emergency stop and start. Using in-field measurements and the Hot Spot Stress Approach, the study analyses a number of aspects that are usually disregarded in design verifications. The results highlight the importance of the detail modeling, the fundamental role played by the non-stationary conditions and the errors committed when using conventional models of the load. The emphasis of this paper is put on two main aspects: the use of full scale data and the detection of isolated events of non-stationary vibrations, the effects of which can be disregarded when using stress time-histories of 10 min - 1 h. • Importance of correct characterization of FAT category of the structural details. • Non-stationary loads give rise to critical conditions for fatigue crack propagation. • Crucial role of low-frequency stress cycles in fatigue damage. • Uncertain estimate of the fatigue damage when using few 10-min stress time-histories. • Limits of current standard methods for small wind turbine design. [ABSTRACT FROM AUTHOR]

Condition monitoring systems (CMS) are essential to reduce costs in the wind energy sector. This paper proposes a method based on Support Vector Machine (SVM) to detect rotor mass imbalance for a multi-class imbalance problem, using the estimated speed as an input variable, obtained through a combination of electrical quantities (currents and voltages). Moreover, it is sought to obtain the magnitude of the rotor mass imbalance. With the aid of statistical tools, intermediate classes can be estimated, other than the ones proposed for the SVM. Besides, if the azimuth position is provided, the angular position of the mass imbalance can be also obtained. A 1.5 MW three-bladed wind turbine model with a permanent magnet synchronous generator, was considered, and a database was built numerically using the software Turbsim, FAST, and Simulink. From the database, the Power Spectral Density (PSD) technique was used to transform the input data from the time to the frequency domain. Then, the SVM algorithm and statistical analysis were used to classify the magnitude and the angular position of the imbalance. Different scenarios of mass imbalance were tested under different wind speeds and turbulence intensities. The results demonstrate the satisfactory performance of the proposed method. [Display omitted] • This paper proposes a method to detect rotor mass imbalance in wind turbines. • Support Vector Machine (SVM) is used to classify imbalance from electrical signals. • The method locates the equivalent angular position of the rotor mass imbalance. • The method estimates intermediate imbalance classes from the SVM classes. • The impact of the wind turbulence intensity on the SVM performance was investigated. [ABSTRACT FROM AUTHOR]

With over 200 wind farm projects underway in 33 states in America, now more than ever, innovation and precision are necessary in wind farm design. The Wind Farm Layout Optimization Problem (WFLOP) calls for optimally positioning turbines within a wind farm so that a particular objective function is optimized in the presence of wake effect. To make the WFLOP tractable, many solution methods begin by modeling the wind farm as a n x n square grid where the centers of each of the n 2 cells serve as potential locations for wind turbines. After making this modeling assumption, there are 2 n 2 potential layouts to consider, and so heuristic algorithms are often employed in order to search the solution space and generate a near-optimal layout. In this paper, we propose a local, continuous refinement technique that seeks to improve the layouts generated by these heuristic algorithms. In particular, we consider the objective function Levelized Cost of Energy (LCOE) and capture wake using the simple, yet effective, Jensen Model. Given the anemometer data for two potential wind farm sites, we begin by generating initial layouts using a specific heuristic algorithm (the Distributed Genetic Algorithm) and then refine these layouts using our proposed continuous, deterministic refinement scheme. We compare the performance of this deterministic refinement technique to another deterministic refinement technique in the space: a Heuristic Hill-climbing approach. Results suggest that our refinement technique outperforms the compared refinement technique by generating layouts that increase overall energy production, consequently reducing the LCOE. To our knowledge, this is the first paper to compare refinement techniques in wind farm layout optimization , and in doing so, we employ a simulation framework that examines the energy production of the generated wind farm layouts for 10 min intervals. • Refinement algorithm. • Derivative-free optimization. • Wind farm layout. • Simulation framework. [ABSTRACT FROM AUTHOR]

This paper proposes the use of Artificial Neural Networks (ANNs), specifically Physics-Informed Neural Networks (PINNs), for dynamic surrogate modelling of wind turbines. PINNs offer the flexibility to model complex relationships while incorporating physics-based constraints, enabling accurate representation of wind turbine dynamics. In this paper, a PINN-based surrogate model is developed for the Blade Element Momentum (BEM) aerodynamic model used in state-of-the-art numerical wind turbine simulations. The PINN model replaces the time-consuming root-finding process in BEM with high-dimensional regression, significantly improving computational efficiency. The PINN model is trained using data generated from a numerical model of the IEA-15MW reference wind turbine, and its performance is compared against conventional data-driven Neural Network (DDNN) models. The proposed surrogate model provides more efficient and accurate evaluations of wind turbine responses compared with traditional surrogate modelling approaches. A significant computational advantage is obtained by using the developed surrogate models with a forty-fold speedup demonstrated compared to the BEM model. Replacing the BEM model with the PINN-based surrogate model for load computation in the numerical model used for dynamic analysis results in an overall reduction of 35% in computational time for a complete dynamic simulation. This is a substantial improvement in efficiency without sacrificing accuracy — the maximum Mean Absolute Error (MAE) values for the surrogate models are of the order of 1 0 − 2 , which shows that the surrogate models can predict the angle of attack at any blade node with a discrepancy of less than 0. 5 °. The surrogate models significantly reduce computational time while maintaining high accuracy, making them a promising approach for simulating wind turbine dynamics, especially in fields such as reliability analysis or fatigue estimation where many simulations are necessary. [ABSTRACT FROM AUTHOR]

The expansion of floating offshore wind brings the industry closer to achieving commercial viability. However, the challenging environment characterised by strong winds, waves, and currents, along with the growing size of wind turbines and the dynamic behaviour of floaters, introduces concerns about power production efficiency and system durability due to increased fatigue loads, which subsequently impacts overall costs. In an attempt to mitigate the financial implications coming from alterations made to control strategies and structural elements during the initial design phase, this paper propounds an all-encompassing simulation framework for offshore wind turbines. The current study thoroughly explores the various capabilities of the tool, with a focus on its simulation models. Importantly, the paper highlights the complex interactions between tool models and different controllers. Carefully designed, this tool offers users a variety of functions to enhance system design, fine-tune control strategies, and thoroughly assess performance metrics. The paper elaborates on these aspects, providing an explanation of the tool's capabilities and enhancing the dynamic comparison between the models. [ABSTRACT FROM AUTHOR]

The increasing demand for renewable energy sources has led to the development of HRES that combine multiple sources of renewable energy to provide a reliable and efficient energy supply. However, renewable energy sources typically come with higher costs than conventional energy production methods. Hence, the optimal sizing of HRES is essential to prevent the oversizing of the system while ensuring a reliable supply of load demand, accommodating variations in meteorological conditions and changes in power demand. Numerous research papers addressing the optimal sizing problem are available in the literature. However, none of the mentioned studies take into account the power degradation of all renewable sources. The degradation of renewable sources, such as PV and WT, can have a significant impact on the performance of HRES. The degradation of PV panels can result in reduced efficiency and output, leading to a decrease in the overall energy production of the HRES. Similarly, WT may experience wear and tear over time, leading to reduced performance and output. The reduction in the output of renewables in the standalone system causes an increase in the loss of power supply probability (LPSP). This paper analyzes the optimal sizing of the HRES, considering the degradation of the sources. Further, analysis has also been carried out based on the number of BESS to be added to HRES to meet the derived constraints and their impact on the cost of energy as well. [Display omitted] [ABSTRACT FROM AUTHOR]

The expansion of floating offshore wind brings the industry closer to achieving commercial viability. However, the challenging environment characterised by strong winds, waves, and currents, along with the growing size of wind turbines and the dynamic behaviour of floaters, introduces concerns about power production efficiency and system durability due to increased fatigue loads, which subsequently impacts overall costs. In an attempt to mitigate the financial implications coming from alterations made to control strategies and structural elements during the initial design phase, this paper propounds an all-encompassing simulation framework for offshore wind turbines. The current study thoroughly explores the various capabilities of the tool, with a focus on its simulation models. Importantly, the paper highlights the complex interactions between tool models and different controllers. Carefully designed, this tool offers users a variety of functions to enhance system design, fine-tune control strategies, and thoroughly assess performance metrics. The paper elaborates on these aspects, providing an explanation of the tool's capabilities and enhancing the dynamic comparison between the models. [ABSTRACT FROM AUTHOR]

The increasing demand for renewable energy sources has led to the development of HRES that combine multiple sources of renewable energy to provide a reliable and efficient energy supply. However, renewable energy sources typically come with higher costs than conventional energy production methods. Hence, the optimal sizing of HRES is essential to prevent the oversizing of the system while ensuring a reliable supply of load demand, accommodating variations in meteorological conditions and changes in power demand. Numerous research papers addressing the optimal sizing problem are available in the literature. However, none of the mentioned studies take into account the power degradation of all renewable sources. The degradation of renewable sources, such as PV and WT, can have a significant impact on the performance of HRES. The degradation of PV panels can result in reduced efficiency and output, leading to a decrease in the overall energy production of the HRES. Similarly, WT may experience wear and tear over time, leading to reduced performance and output. The reduction in the output of renewables in the standalone system causes an increase in the loss of power supply probability (LPSP). This paper analyzes the optimal sizing of the HRES, considering the degradation of the sources. Further, analysis has also been carried out based on the number of BESS to be added to HRES to meet the derived constraints and their impact on the cost of energy as well. [Display omitted] [ABSTRACT FROM AUTHOR]

Low temperature and high humidity in winter can easily lead to blade icing, which severely impacts the actual power output and safe operation of wind turbines. However, existing blade icing diagnostic methods suffer from inadequacies in feature extraction and the accuracy. Therefore, we propose an efficient and accurate blade icing diagnosis method. Initially, we develop hybrid features that include both short-time and long-time features. RFE is combined with icing mechanism to extract short-time features, and the sliding window algorithm is used to extract long-time features. Secondly, the B-SMOTE method is used to synthesize minority icing samples. Finally, the B-SMOTE-Bi-GRU model is applied to validate the superiority of the proposed method. Compared to the features extracted based on the icing mechanism and the sliding window algorithm, the average improvements in recall, TNR, and accuracy for each wind turbine is between 0.26% and 4.87 %, 0.23%–4.60 %, and 0.05%–2.73 %, respectively. In comparison to the other models, the B-SMOTE-Bi-GRU method proposed in this paper enhances the recall and TNR of each wind turbine by between 0.06%-13.10 % and 0.07%–12.01 %, respectively. The results indicate that the hybrid features and B-SMOTE-Bi-GRU model proposed in this paper offer a more effective and accurate approach to blade icing diagnosis. • A new feature extraction method is proposed based on RFE combined with icing mechanism. • Icing diagnostic model based on the B-SMOTE-Bi-GRU model is proposed. • The proposed new method of feature extraction enhances the accuracy of blade icing diagnosis. • The B-SMOTE-Bi-GRU model is superior to existing models. • The proposed method achieves higher accuracy for blade icing diagnosis. [ABSTRACT FROM AUTHOR]

Wind speed forecasting is the basis of wind farm operation, which provides a reference for the future operation status evaluation of wind farms. For the wind speed forecast of wind turbines in the whole wind farm, a strategy combining unified forecast and single site error correction is proposed in this paper. The unified forecast framework is composed of a unified forecast model and multiple single site error correction models, which combines the forecasted grids of numerical weather prediction (NWP) with the monitoring data of wind farms. The proposed unified forecast model is called spatiotemporal conversion deep predictive network (STC-DPN), which is composed of temporal convolution network (TCN) and 2D convolution long short-term memory network (ConvLSTM). Firstly, the NWP forecasted grids are interpolated to the fan location, and the sequence matrix is composed of the NWP data and the monitored data of each wind turbine according to the time series, which is entered into the TCN network for time sequence feature extraction. Then, the output of the TCN network is converted into a regular spatio-temporal data matrix, which is entered into the ConvLSTM network for joint learning of spatio-temporal features to obtain the wind speed sequence forecasted in the whole wind farm. Finally, an independent TCN-LSTM error correction model is added for each site. Variational modal decomposition (VMD) is used to process data series, and different processing methods are adopted in unified forecast and single site error correction. In the 96 steps forecast test of a wind farm from Jining City, China, the proposed method is superior to several baseline methods and has important practical application value. [ABSTRACT FROM AUTHOR]

Condition monitoring (CM) of wind turbines (WTs) is commonly accepted as an effective way to increase the availability and reduce the operation and maintenance (O&M) costs of wind farms. CM methods based on data from Supervisory Control and Data Acquisition (SCADA) system have garnered a great deal of interest due to their accessibility and low cost. Existing studies based on SCADA data provide whether the WT is normal or abnormal, and further the fault locations if an anomaly occurs. Nevertheless, few studies have attempted to identify the underlying causes of the anomaly. In this paper, a novel WTCM method utilizing a self-attention-based mechanism embedded with a multivariable query pattern is proposed for the anomaly detection and underlying causes identification. Firstly, an anomaly detection model composed of multiple cascaded encoders and decoders with a multi-head self-attention mechanism is proposed to extract cross-variable correlations. Furthermore, the multivariable query pattern is designed to evaluate the influences of different features on the target, which is essential for fault identification. Finally, the relative anomaly index (RAI) is proposed for each feature to quantify the impact of each feature on the anomaly. RAIs are adopted to analyze the underlying causes of anomalies. Experimental results confirmed the effectiveness of the proposed method for the WT fault early detection and identification of anomaly causes. [ABSTRACT FROM AUTHOR]

Floating offshore wind turbines (FOWTs) are the largest rotating structures on the earth. Dynamically sensitive structures such as these must be protected in these environments to ensure that they can continue to operate reliably and safely. In this paper structural dynamic models and probabilistic assessment tools are combined to demonstrate improvements in structural reliability when FOWT towers are equipped with a new type of damper — the tuned mass damper inerter (TMDI). A multi-body dynamic approach is used to model the wind turbine and the TMDI installed in the tower. The model is subjected to stochastically generated wind and wave loads of varying magnitudes to develop wind-induced probabilistic demand models for towers of FOWTs under model and load uncertainties. A focus is placed on the impact of the wind-wave misalignment on the lightly damped side-to-side mode. Numerical simulations are carried out to construct fragility curves which illustrate reductions in the vulnerability of FOWTs to wind and wave loading owing to the inclusion of the new damper. Results show that the TMDI delivers significant increases in structural reliability of FOWT towers. [ABSTRACT FROM AUTHOR]

Wind tunnel experiments provide worthwhile insights for designing efficient micro wind energy harvesters and large-scale wind turbines. As wind tunnel tests with large-scale wind turbines are expensive and not always feasible, most experiments are conducted with geometrically scaled rotors. Furthermore, micro-scale runners used for wind energy harvesting face the issue of lower efficiency than large turbines. A better understanding of Reynolds number effects induced by the downsizing of a turbine would help to design more efficient wind energy harvesters and more faithfully scaled experiments. This paper reports on Reynolds number effects on the performance and wake of micro-scale wind turbines. Wind turbines' power and torque coefficients are measured in a wind tunnel for a wide range of Reynolds numbers. The wake axial velocity fields and the vortex core locations are collected for three Reynolds numbers using phase-averaged and phase-locked particle image velocimetry technique. The results emphasize that an increase in the Reynolds number leads to larger power coefficients, torque coefficients, and optimum tip-speed ratios. Higher Reynolds numbers induce wider wake expansion and a larger axial velocity defect. This quantitative analysis will contribute to a clearer understanding of the scaling effects and help to design more efficient wind energy harvesters. [ABSTRACT FROM AUTHOR]

The data imbalance problem extensively exists in wind turbine fault diagnosis, resulting in the compromise between learning attention to majority and minority classes. In this paper, a deep neural network method is proposed to resolve the mentioned problem. Specifically, convolutional and recurrent neural networks are designed to extract spatial and temporal features within supervisory control and data acquisition (SCADA) measurements. To improve the reliability of fault diagnosis results by collective decision, a coarse learner and multiple fine learners are established. With the consideration of data imbalance and learning diversity, fault-related information can be revealed. Moreover, a learner selection scheme is designed to ensure high computational efficiency. The effectiveness of the proposed method is demonstrated by experiments based on simulated data and real-world SCADA measurements from a wind farm. Experimental results show that the accuracy in identifying health conditions can be improved by the proposed method regardless of the data imbalance. On the two datasets, the proposed method outperforms four benchmark approaches as the learning attention to all classes can be enhanced. Therefore, the proposed method is a promising solution to wind turbine fault diagnosis. [ABSTRACT FROM AUTHOR]

Offshore wind is becoming the way forward for green energy harnessing worldwide. However, frequent typhoons are a major constraint for the development of offshore wind power for some regions in Asia and North America. Typhoons may pose a huge challenge to offshore wind farm development in southern China. In this paper, a fully-coupled analysis was carried out for a 10 MW large-scale monopile offshore wind turbine (OWT) using SIMO-Riflex-Aerodyn (SRA) code. The response characteristics of the OWTs in different typhoon regions are investigated based on the measured typhoon conditions in China. The effect of aerodynamic damping on the response and the load effect is analyzed in detail. Two different distribution methods are used to statistically extrapolate the response value and get the short-term extreme response. Cumulative linear fatigue damage is evaluated by the rain-flow counting method to explore the possible failure modes of large wind turbines during typhoons. The results show that aerodynamic loads play an important role in large monopile OWTs during high wind speeds in parked conditions. The extreme response and fatigue analyses from this study indicate that fatigue is a dominant failure mode for the large OWT tower during typhoons, while buckling is unlikely. • The wind field at the wind turbine was analyzed when the typhoon passed by. • The response of large monopile wind turbine in each typhoon region of is analyzed. • Aerodynamic damping play an important role for the structural response of parked monopile OWT. • The extreme moments at the tower base follows the Gumbel distribution and lognormal distribution. • The cumulative fatigue damage during different typhoon stages is evaluated. [ABSTRACT FROM AUTHOR]

The aerodynamic simulations of wind turbines are typically carried out using a steady inflow condition. However, the aerodynamics and aeroelasticity of wind turbine blades can be significantly affected by inflow wakes due to the environmental conditions or the presence of neighbouring wind turbines. In this paper, the effects of flow unsteadiness on the aerodynamics and aeroelasticity of the wind turbine rotor are investigated. It is found that the unsteadiness of the wake can have an impact on the aerodynamic flow field around the wind turbine rotor and it could also influence the aeroelasticity of the wind turbine. One of the distinctive features of this paper is the application of the highly efficient nonlinear frequency domain solution method for modelling harmonic disturbances for the aerodynamic and aeromechanical analysis of wind turbines. A test case wind turbine is selected for the aerodynamic and aeromechanical analysis as well as for the validation of the method used. The effects of different material properties along with a large vibration amplitude on the aeroelasticity parameter known as aerodynamic damping of the wind turbine blade are also investigated in the present work. Compared to the conventional time domain solution methods, which require prohibitively large computational cost for modelling and solving aerodynamics and aeroelasticity of wind turbines, the proposed frequency domain solution method can reduce the computational cost by one to two orders of magnitude. • A frequency domain method is proposed for aeromechanical analysis of wind turbines. • Proposed method reduces computation time by one to two orders of magnitude. • The effect of material properties on aerodynamic damping is investigated. • Composite material provides greater aerodynamic damping. • Inflow wakes have significant impact on wind turbine aerodynamics. [ABSTRACT FROM AUTHOR]

A framework of using transfer learning algorithms, Inception V3 and TrAdaBoost, for fault diagnosis of two wind turbine faults is presented and verified. Two failure modes, blade icing accretion and gear cog belt fracture, are analyzed using SCADA data. A new index named 'Comprehensive Index' is defined to evaluate performance of different algorithms. Traditional machine learning algorithms do not perform well for data sets that are unbalanced and follow different distributions. The former causes bias in classification and the latter leads to poor adaptability of algorithms. A novel transfer learning algorithm studied in this paper, TrAdaBoost, has been proved to have superior performance on dealing with data imbalance and different distributions. A new approach to calibrate data labels using transfer learning algorithms is also proposed, which provides important insights into unsupervised learning for wind turbine fault diagnosis. • A framework of using transfer learning algorithms on SCADA data for WT fault diagnosis is proposed. • A new index named 'Comprehensive Index' is defined to fairly evaluate performance of different algorithms. • TrAdaBoost shows its superior performance on dealing with data imbalance and different distributions. • A new framework to calibrate data labels is proposed, which can solve data improper labelling problem. • This paper provides important insights for unsupervised learning. [ABSTRACT FROM AUTHOR]

This paper investigates a novel control strategy that reduces the fatigue load of wind turbines during active power control. Based on the analysis of the reference power tracking of wind turbines, the proposed new control strategy considers simultaneous control of rotor speed and pitch angle. For this goal, a two-degree-of-freedom wind turbine active power control model based on the small-signal method is established, and the linearized fatigue load calculation is also completed in the model. For the effective control under this two-degree-of-freedom control framework, a model predictive control method suitable for multi-objective optimization is applied. Simultaneously, with the help of simulation tools, adaptive weight selection for multi-objective optimization is discussed and implemented. Simulation results show that the two-degree-of-freedom control system proposed in this paper can accurately implement reference power tracking and significantly reduce the fatigue load of wind turbines. The proposed method is valid not only on a single wind turbine but also in the wind farm's active power dispatching process. • Two-degree-of-freedom wind turbine control model is studied. • Fatigue load of wind turbines is modeled linearly. • Optimized multi-input multi-output MPC optimization control is adopted. • Fatigue load of wind turbine operation is significantly reduced. [ABSTRACT FROM AUTHOR]

Blades icing would reduce the aerodynamic performance, cause power generation loss of WTGs, and even affect the safety of production and operation. By analyzing the relationship between blades icing and Supervisory Control And Data Acquisition (SCADA) data characteristic parameters at different stages in the process of wind turbines generating energy transfer, the paper calculated the accompanying changes in blades icing timing of Wind Turbine Generator System (WTGs) output power performance, mechanical performance and aerodynamic performance characteristic parameters. This paper then applies the residual to describe the deviation degree of each characteristic parameters value, and establishes the blades icing identification model by progressive parameterization judgment form. In addition, combined with the statistical properties of historical meteorological parameters of blades icing, the timely and accurate determination of blades icing was achieved. The identification results of the model are reliable and accurate through the verification of blades icing examples in different regions and of different wind turbines. The blades icing identification model based on SCADA data variation characteristics does not require adding new hardware and software investment, but it has even higher sensitivity. It can make accurate judgment in the early icing process, which is conducive to implement the control strategy and develop de-icing plan in advance. • The determining blade icing was established based on the cascade correlation of performance parameters. • A model of blade icing identification based on multi-parameter deviation was proposed. • Blade icing can be confirmed 30 min earlier than current methods. • There is no additional cost to leveraging existing SCADA data. [ABSTRACT FROM AUTHOR]

Due to constant change of rotational speed and low frequency property of vibration signal, it is difficult to identify the weak time-varying fault features of direct-driven wind turbines during early period. In view of the problems of variable speed and strong time-varying characteristics, the traditional Time-Frequency (TF) analysis methods always generate a low TF resolution. Rearrangement Method (RM), as a mathematically excellent and efficient solution, is used to improve the TF energy localization, whereas it cannot reconstruct signals. To make TF plane more readable, RM can be utilized to make up the deficiency of the frequency-reassigned Synchrosqueezing Transform (SST). On such basis, this paper puts forward a novel TF analysis method, namely RFSST2, which has drawn on the advantages of RM and Second-Order Synchrosqueezing Transform based on STFT (FSST2). This method is used to sharpen the blurry TF energy for strong time-varying modulated signals. Besides, it can reconstruct the perfect signals. After the extraction the rotation speed curve from high-resolution TF plane generated by RFSST2, the order analysis without tachometer is performed to address the problem of variably rotating speed. To prove the proposed method is superior in processing complex multi-component signals, we respectively estimate the variable speed of wind turbines and deal with the fault signals from the gear fault simulation test-bed and from actual 1.5 MW direct-drive wind turbines. The analysis result obtained by comparing different TF methods shows that the method put forward by this paper can effectively extract time-varying fault characteristics under non-stationary condition. Image 1 • RM can be utilized to make up the deficiency of the frequency-reassigned SST. • RFSST2 is proposed to produce a more concentrated TF while remaining reversible. • Order analysis of 1.5 MW direct-drive wind turbine is performed without tachometer. [ABSTRACT FROM AUTHOR]

In this paper, an improved Sliding Mode Control(SMC) is proposed based on the Wind Energy Conversion System(WECS), which is with offshore wind turbine Permanent Magnet Synchronous Generator(PMSG) connected to the grid. Moreover, an aerodynamic torque observer is presented to improve the performance of Maximum Power Point Tracking(MPPT). Since the wind energy conversion system is a complicated nonlinear system, an enhanced reach law of sliding mode controller is designed to reduce the chattering issue. Compared to traditional control methods, the simulation results show that the control method proposed in this paper significantly improves the stability and robustness of PMSG-Based WECS in the area of MPPT. • A sliding mode control method with extant state observer is proposed for MPPT. • A sliding mode control method based on a new approach rate is designed. • The anti-interference ability is signi_cantly improved by the proposed method. • The feed-forward compensation is added on grid voltage orientation. [ABSTRACT FROM AUTHOR]

Maintenance scheduling (MS) for wind turbines (WTs) is an emerging investigation area in recent years. The MS of WTs is more complex than that of traditional thermal generators, because maintenance activities of WTs are affected by stochastic weather conditions, e.g., wind speed and precipitation. This paper proposes a long-term MS method to obtain the joint preventive maintenance plan during the whole warranty period of WTs on multiple wind farms. Both the labour cost and production loss are used as objective functions of the MS. Historical weather data are analysed, and a statistical model is developed to describe the weather conditions. Then, the MS problem is formulated compactly as a mixed integer linear programming model. Finally, a detailed practical case study is demonstrated to validate the effectiveness of the proposed MS method. The result confirms that cost-effective joint preventive maintenance (PM) plans of three wind farms can be derived through the proposed MS method. Compared with the periodic PM plan, the expected labour cost and production loss are reduced by approximately 30% and 20%, respectively. • A practical maintenance scheduling problem of wind farms is solved in this paper. • Practical meteorological data are statistically analysed and modelled. • The long-term maintenance scheduling problem considers multiple wind farms. • The scheduling problem is compactly formulated using a MILP model. [ABSTRACT FROM AUTHOR]

This study explored the concept of a floating offshore wind farm (FOWF) that self-reconfigures wind turbine (WT) positions. The self-reconfiguration (SR) mechanism moves degraded turbines to different farm positions to delay failures occurring and reduce power loss. A 40-turbine agent-based simulation was created utilising wind and turbine performance data. For the first time, the SR mechanism was designed and optimised with principles of self-engineering complexity. The paper demonstrates the effectiveness of the SR mechanism through an agent-based simulation approach. The optimised SR was used in FOWF simulations of 50 years of operation; SR balanced fatigue across the FOWF and led to an increased income of £5–40M at years 27–33 of operation; however, outside of these years, there is no net positive income, and by year 50 SR has cost the FOWF £4–7M more in movement costs. Lastly, simulations with repair and maintenance restricted to 10 months or less were conducted. SR delayed turbine failures, so they occurred in months when repair could be conducted. The SR reduced power loss and increased net income by up to £20M, indicating that SR could be useful when repair and maintenance times are limited. In the absence of significant operational data, a qualitative validation with experts confirmed the approach and the validity of the simulation model for a range of FOWF scenarios. [ABSTRACT FROM AUTHOR]

Abstract —The fault diagnosis of wind turbines under nonstationary conditions is still challenging. This paper proposes a novel tacho-less generalized demodulation (NTLGD) method for the wind turbine fault diagnosis. First, one instantaneous frequency is extracted from the time-frequency representation of the vibration signal. Second, a novel phase function design method is developed based on the extracted frequency, by which, different from traditional methods, the fault-related frequencies are mapped into the fixed predefined values. Then, the bandpass filters are designed according to the designed phase functions to separate the mapped frequencies. Finally, the diagnosis template is constructed, and the fault is localized by matching the peaks in the demodulated spectrum with the spectral lines in the template. The method is evaluated by the vibration signals of support bearings and the planetary gearbox in a test rig of wind turbine drive train. The results demonstrate that the proposed method can well pinpoint the fault-related frequency components without a tachometer and the demodulated values are independent of speed profiles. • A novel tacho-less generalized demodulation method is proposed. • A phase function design method is developed based on one extracted frequency. • The time-varying frequencies of signals are mapped into fixed values. • The method is evaluated by the signals of bearings and a planetary gearbox. [ABSTRACT FROM AUTHOR]

The unsteady aerodynamic characteristics and interference effects of floating offshore wind turbine (FOWT) are mainly affected by the pitch motion of the ocean platform. Based on computational fluid dynamics (CFD) and advanced overset grid technology, the aerodynamic performance and wake characteristics of a fully configured wind turbine with rotating blades, nacelle , and tower are studied in this paper. The effects of the amplitude and frequency of pitch motion on the wind turbine aerodynamic loads and flow field are investigated herein in detail. The power and thrust between numerical simulation and experiment are compared. The results show that the grid and simulation parameters used in this study can accurately capture the aerodynamic characteristics and the flow field around wind turbines. The influence of the pitch amplitude and frequency on the performance of wind turbines is discussed. The complex flow interaction between the tip vortex, tower shedding vortex, and the turbulent wake was observed. The present results indicate that the pitch motion amplitude and frequency have a great influence on the power, thrust, and wake characteristics. [ABSTRACT FROM AUTHOR]

In the growing wind energy sector, as in other high investment sectors, the need to make assets profitable has put the spotlight on maintenance. Efficient solutions which leverage from condition or performance based maintenance policies have been proposed during the last decades, but the proposed methods generally focus on individual components or stand for specific application areas. This paper aims to contribute to the development of performance based maintenance strategies within the wind energy sector by providing a condition monitoring based generic methodology for wind turbine performance assessment at system level. The proposed methodology is based on the detection of critical periods in which low performance is detected repeatedly. Multiple machine learning methods and models are applied to assess the wind turbine performance. This methodology has been applied in a case study with SCADA data of eight wind turbines. An analyst could benefit from the implementation of the methodology and the easy-to-interpret results shown in the proposed control chart, especially in cases in which there is less know-how about which variables have higher impact on systems performance. • A methodology for performance-based maintenance of wind turbines is proposed. • The methodology has been applied in a case study with eight wind turbines. • Machine learning methods are applied to assess the wind turbine performance. [ABSTRACT FROM AUTHOR]

Oscillating water columns (OWCs) are widely regarded as the simplest and most reliable type of wave energy converter. An economically viable OWC wave power plant requires maximizing the performance of the power take-off system, composed of an air turbine coupled to an electrical generator. Besides, the OWC power plant's inherent nonlinearities and modeling uncertainties must be dealt with. One solution is the use of fuzzy controllers since they capture both nonlinearities and uncertainties of the system, which can largely contribute to improving the overall efficiency. This paper shows how an inverse fuzzy model optimized by a genetic algorithm (GA) can improve the control of an OWC and, consequently, its performance. As a case study, the Mutriku power plant with a biradial turbine was addressed, considering 11 different sea states of the characteristic wave climate. Two optimization objectives were considered: maximizing turbine and generator output average powers. Compared with a well-known baseline control strategy, the GA-fuzzy control did not significantly contribute to increasing average turbine power. However, when maximizing generator yearly-accumulated power, there were improvements of over 9%. Ultimately, these research findings may play an essential role in control strategies for wave power plants. • An inverse model controller was proposed for wave energy control. • The inverse model was obtained with fuzzy modeling and optimized by genetic algorithms. • Focus on a real case study: An oscillating water column at the Mutriku power plant. • Comparison of the proposed algorithm with a well-known baseline strategy. • The proposed strategy showed yearly improvements of over 9% on average generator power. [ABSTRACT FROM AUTHOR]

Improving the efficiency of wind turbine state prediction is an important goal of wind energy utilization. But much of abnormal data existing in supervisory control and data acquisition (SCADA) seriously affects the health state prediction of wind turbine. In this paper, a new composed method is proposed to clean SACAD data according to abnormal data type of wind turbine. In proposed composed method, a preprocessing method is first presented to get rid of outliers of power curve based on operational mechanism, and a new data cleaning method called TTLOF (Thompson tau-local outlier factor) is proposed to quantify particularly data points and eliminate outliers by setting correlation parameter thresholds. In TTLOF cleaning data, Empirical copula-based mutual information (ECMI) is used to select correlation parameters for anomaly characteristic assessments, and each parameter interval is divided for performing segmentation fine cleaning which can reduce the model complexity of identifying anomaly characteristics. Finally, a deep learning network which is long short-term memory (LSTM) is used to verify the effectiveness of the proposed data cleaning method. By analyzing the state monitoring results, it is shown the proposed composed method is more effective for cleaning anomy data than other methods. [ABSTRACT FROM AUTHOR]

Condition monitoring, fault diagnosis, and scheduled maintenance of wind turbines (WTs) are becoming a necessity to maximize their economic benefits and reduce their downtime. One of the critical faults in WTs is inter-turn short-circuit faults (ITSCF) in their rotating machines. Being dependent on higher wind speeds, offshore wind farms are more efficient than onshore ones, yet at the cost of higher ITSCF, more maintenance requirements, and difficulty in accessibility. Thus, they require an efficient non-contact fault diagnosis technique to enable operators to inspect all WT components with minimal contact. Among existing fault detection technologies, Infrared thermography (IRT) is considered as a non-invasive and non-destructive anomaly detection technique based on capturing thermal images by IR cameras, making it suitable for offshore harsh environments. Meanwhile, deep learning (DL) fault diagnosis approaches have proven to be promising intelligent tools with minimum human labor. This paper proposes an efficient and robust ensemble DL-based diagnosis method for ITSCF detection in induction rotating machines, besides identifying fault locations and short-circuit severity. The proposed technique is quite convenient for early diagnosis of offshore WT generator faults since it depends on IRT technology. Compared to previous studies, the proposed approach outperforms its counterparts since it integrates eight DL models, rather than using a single model, thus merging all their structural benefits altogether. Moreover, a cascaded feature fusion and selection procedure is introduced. First, the deep features extracted from the DL models are fused, then the most influential features are selected using a hybrid feature selection scheme. Thus, the classification accuracy is enhanced to reach 100% and meanwhile, the input features' size to the classifier is reduced, decreasing classification complexity and training time. Finally, no clustering or segmentation phases are required in the proposed technique, resulting in a further decrease in the diagnosis time and computation burden. • A CAD is proposed for faults diagnosis in induction generator-based offshore windfarms. • It fuses 8 DL models instead of using an individual one, thus boosting the classification ability. • The CAD employs a few features to perform diagnosis through a cascaded procedure. • First, features are fused, then significant features are selected via hybrid feature selection. • No segmentation or clustering stages are required which adds to the system's simplicity. [ABSTRACT FROM AUTHOR]

Accurate image stitching is crucial to wind turbine blade visualization and defect analysis. It is inevitable that drone-captured images for blade inspection are high resolution and heavily overlapped. This also necessitates the stitching-based deduplication process on detected defects. However, the stitching task suffers from texture-poor blade surfaces, unstable drone pose (especially off-shore), and the lack of public blade datasets that cater to real-world challenges. In this paper, we present a simple yet efficient algorithm for robust and accurate blade image stitching. To promote further research, we also introduce a new dataset, Blade30, which contains 1,302 real drone-captured images covering 30 full blades captured under various conditions (both on- and off-shore), accompanied by a rich set of annotations such as defects and contaminations, etc. The proposed stitching algorithm generates the initial blade panorama based on blade edges and drone-blade distances at the coarse-grained level, followed by fine-grained adjustments optimized by regression-based texture and shape losses. Our method also fully utilizes the properties of blade images and prior information of the drone. Experiments report promising accuracy in blade stitching and defect deduplication tasks in the vision-based wind turbine blade inspection scenario, surpassing the performance of existing methods. • An efficient stitching method is introduced for accurate defect and blade analysis. • Our proposed stitching method yields 2–3 times better accuracy than existing methods. • Our new and comprehensive dataset, Blade30, leads in both quality and quantity. • Blade30 covers 30 full blades, contains various annotated defects and contaminations. [ABSTRACT FROM AUTHOR]

Accurate estimation of extreme wind and wave conditions is critical for ocean engineering activities and applications. Various renewable energy offshore structures, particularly floating wind turbines are designed to sustain extreme wind and wave induced loads. Statistics of wind speeds and wave heights is the key input for structural safety and reliability study. Consequently, development of novel robust methods, able to predict extreme wind-wave conditions is essential. This paper discusses criteria for selecting design point by applying recently developed method for estimating extreme wave statistics, based on the hourly wave height and wind speed maxima at the location of interest. Wave and wind data, analyzed in this paper, was obtained from the hindcast model applied to the SEM-REV offshore sea location, near the coast of France, during years 2001–2010. The ECMWF (European Centre for Medium range Weather Forecasting) framework along with the atmospheric model SKIRON were employed to generate accurate hindcast wind-wave hourly data at the location of interest. Note that the SEM-REV site was built within the framework of the CPER (Contrat de Projet Etat-Région) 2007–2013 for the Pays de la Loire region, therefore it is important to note that 2001–2010 data studied in this paper was obtained by hindcast into the time period before SEM-REV began operations. Structural design values are often based on univariate statistical analysis, while actually multivariate statistics is more appropriate for modelling the whole structure. The bivariate analysis of extremes is often poorly understood and generally not adequately considered in most practical measurements/situations, therefore it is important to utilize recently developed bivariate average conditional exceedance rate (ACER2D) method. This paper studies extreme wind speeds and wave heights, that are simultaneously obtained at the same location. Due to less than full correlation between wind speed and wave height, application of the multivariate, or bivariate in the simplest case, extreme value theory is of practical importance. This paper focuses on application of the bivariate ACER2D method for prediction of bivariate extreme value statistics. Finally, this paper suggests how the design point should be chosed based on bivariate analysis. The latter is of particular engineering importance as it presents first application of bivariate wind-wave statistics to a raw SEM-REV site data. • First application of extreme wind-wave statistics to a raw SEM-REV site data. • Predictions of extreme in situ wave heights and wind speeds are obtained. • Engineering advantages of the proposed method are discussed. • The novel technique of improving correlated extreme wind speed and wave height predictions has been presented. [ABSTRACT FROM AUTHOR]

The typhoon-induced vibration characteristics of large wind turbines are significantly different in different travelling stages of typhoons due to the structural complexity of typhoons. Influences of multi-stage typhoon-induced effects on structural safety of wind turbines have not been studied yet. The objective of this paper is to investigate the vibration characteristics of wind turbines in different stages of the typhoon as well as the influencing rules of the structural design standards. For this purpose, a framework was established for predicting multi-stage typhoon-induced effects of large wind turbines, which includes a new typhoon-induced multi-stage wind field simulation method and an advanced multi-body model for large wind turbines. On this basis, aerodynamic loads and dynamic response of large wind turbines during different travelling stages of typhoon were analyzed systematically based on the blade element momentum, multi-body dynamic methods, spectral analysis and data statistics. The working mechanisms of multi-stage effects on vibration characteristics of the large wind turbine were revealed. Finally, an evaluation method of vibration amplification effects for large wind turbines with considerations to multi-stage effects was established. Research results demonstrate that the proposed method can predict vibration characteristics of large wind turbines considering the multi-stage effects efficiently. The multi-stage typhoon-induced effects can influence the value of peak factor and the extremum of wind-induced force and vibration responses of large wind turbines significantly. Conversely, the wind vibration coefficient of structural design was affected slightly. Instead of using a uniform structural design standard for large wind turbines, the influence rule of multi-stage effects on anti-typhoon safety performance was summarized in this paper. • A method for predicting multi-stage typhoon-induced effects of large wind turbine. • The working mechanisms of typhoon-induced vibration of wind turbine are revealed. • The nonlinearity, non-Gaussian and resonance effects of responses are presented. • The structural design factors are studied considering multi-stage typhoon effects. [ABSTRACT FROM AUTHOR]

In recent years, the intelligent diagnosis technology of wind turbines has made great progress. However, in practical engineering applications, the operating states of wind turbine are various, accompanied by a large number of noise interference, which leads to the decline of the discrimination accuracy of intelligent diagnosis. In order to solve this problem, inspired by the Google team Inception model, this paper proposes a concurrent convolution neural network (C–CNN), the raw data is fed into the network without any prior knowledge, and the characteristics are learned directly and adaptively from the input. Even if the data is accompanied by noise, the model still has high accuracy and strong generalization ability. The model is composed of a CNN with multiple branches. Meanwhile, the convolutional layer of different branches selects the kernels with different scales in same level, thus improving the learning ability of entire network. In this paper, the feasibility of this method for fault diagnosis of bearings in wind turbines is demonstrated by three bearing datasets. The results show that the proposed method can extract discriminative features and classify bearing data accurately under the disturbance of different rotating speed, different load and random noise. • The C–CNN realize each layer has multi-scale kernels, so as to data features in a multi-angle and deep level. • The C–CNN network not only has good classification accuracy, but also has good generalization ability. • The model can balance the depth and width of the network, and thus control the growth of parameters and calculations. • The feasibility of this method for fault diagnosis of bearings in wind turbines is demonstrated by three data sets. [ABSTRACT FROM AUTHOR]

In recent years, the intelligent diagnosis technology of wind turbines has made great progress. However, in practical engineering applications, the operating states of wind turbine are various, accompanied by a large number of noise interference, which leads to the decline of the discrimination accuracy of intelligent diagnosis. In order to solve this problem, inspired by the Google team Inception model, this paper proposes a concurrent convolution neural network (C–CNN), the raw data is fed into the network without any prior knowledge, and the characteristics are learned directly and adaptively from the input. Even if the data is accompanied by noise, the model still has high accuracy and strong generalization ability. The model is composed of a CNN with multiple branches. Meanwhile, the convolutional layer of different branches selects the kernels with different scales in same level, thus improving the learning ability of entire network. In this paper, the feasibility of this method for fault diagnosis of bearings in wind turbines is demonstrated by three bearing datasets. The results show that the proposed method can extract discriminative features and classify bearing data accurately under the disturbance of different rotating speed, different load and random noise. • The C–CNN realize each layer has multi-scale kernels, so as to data features in a multi-angle and deep level. • The C–CNN network not only has good classification accuracy, but also has good generalization ability. • The model can balance the depth and width of the network, and thus control the growth of parameters and calculations. • The feasibility of this method for fault diagnosis of bearings in wind turbines is demonstrated by three data sets. [ABSTRACT FROM AUTHOR]

The typhoon-induced vibration characteristics of large wind turbines are significantly different in different travelling stages of typhoons due to the structural complexity of typhoons. Influences of multi-stage typhoon-induced effects on structural safety of wind turbines have not been studied yet. The objective of this paper is to investigate the vibration characteristics of wind turbines in different stages of the typhoon as well as the influencing rules of the structural design standards. For this purpose, a framework was established for predicting multi-stage typhoon-induced effects of large wind turbines, which includes a new typhoon-induced multi-stage wind field simulation method and an advanced multi-body model for large wind turbines. On this basis, aerodynamic loads and dynamic response of large wind turbines during different travelling stages of typhoon were analyzed systematically based on the blade element momentum, multi-body dynamic methods, spectral analysis and data statistics. The working mechanisms of multi-stage effects on vibration characteristics of the large wind turbine were revealed. Finally, an evaluation method of vibration amplification effects for large wind turbines with considerations to multi-stage effects was established. Research results demonstrate that the proposed method can predict vibration characteristics of large wind turbines considering the multi-stage effects efficiently. The multi-stage typhoon-induced effects can influence the value of peak factor and the extremum of wind-induced force and vibration responses of large wind turbines significantly. Conversely, the wind vibration coefficient of structural design was affected slightly. Instead of using a uniform structural design standard for large wind turbines, the influence rule of multi-stage effects on anti-typhoon safety performance was summarized in this paper. • A method for predicting multi-stage typhoon-induced effects of large wind turbine. • The working mechanisms of typhoon-induced vibration of wind turbine are revealed. • The nonlinearity, non-Gaussian and resonance effects of responses are presented. • The structural design factors are studied considering multi-stage typhoon effects. [ABSTRACT FROM AUTHOR]

Economic nonlinear model predictive control (ENMPC) is a strong candidate for controlling wind turbines (WTs). In the model predictive control (MPC) group, the model is the crucial component for the true controller performance. It is common to use simplified models to reduce the problem complexity. These models neglect some of the underlying dynamic responses of real wind turbines. This paper simulates the case in which high accuracy nonlinear models describe both the plant and the controller. The results will be compared to reduced-order models in order to extract conclusions and decide the most appropriate model for WT control. On the other hand, one of the main features of MPC and ENMPC is the concept of receding prediction horizon, which considers the future evolution of the plant to compute the control action. The error of prediction will drastically reduce MPC performance. Also, rapid variation in wind speed can cause problems since wind turbines cannot easily follow these sudden variations due to their high inertia and aerodynamic characteristics. This paper provides an advanced control approach to improve the energy extraction from turbulent wind and enhance wind turbine durability. By implementing this method, the wind speed forecasting is done with a combination of artificial neural networks (ANN) and dynamic equations applied in ENMPC. The results show a significant enhancement of the control performance. • Enhancing the net energy of variable speed wind turbines (WTs). • Considering high accuracy nonlinear models for wind turbine simulation. • Comparing reduced-order models to choose the most appropriate model for WT control. • Suggesting an advanced control approach combining wind speed forecasting with ENMPC. • Improving ENMPC performance, and thereby increasing the power output of WT. [ABSTRACT FROM AUTHOR]

This paper discussed the procedure of an optimization design method involving a small lab-scale wind turbine rotor. To simulate the aerodynamic performance of a prototype wind turbine at utility-scale by scaled model experiments (especially, the visualization measurements of wake flow in small wind tunnels), an appropriate similar design of the small lab-scale rotor is needed. However, the similar design of blade performance from a prototype wind turbine for small scaled model is rare. In this paper, using a 2.5 MW utility-scale wind turbine as the prototype, we set up a new optimization process of similar design for a model wind turbine with a 320 mm rotor based on the Lifting-Line Theory (LLT) with wake-induced corrections. Results show a dramatic deviation in performance of the geometric scaled model compared to the objective values, which demonstrates the significance of the Reynolds number effects. With the optimized blade, the distribution of normal thrust force is very similar to the objective values. Geometric characteristics variations along the blade span are similar to the prototype. With the tip speed ratio runs at a roughly matched rated condition, the optimized rotor gives a better performance than other models used in previous wind tunnel studies. • A new small lab-scale wind turbine similar to the prototype has been designed. • The design process provides a hybrid model blade similar to the prototype. • Performance of geometric scaled model is significantly less than the objective value. • A better performance of the optimized rotor has been comparably reached. [ABSTRACT FROM AUTHOR]