Search

The remaining useful life (RUL) of transformer insulation paper is largely determined by the winding hot-spot temperature (HST). Frequently the HST is not directly monitored and it is inferred from other measurements. However, measurement errors affect prediction models and if uncertain variables are not taken into account this can lead to incorrect maintenance decisions. Additionally, existing analytic models for HST calculation are not always accurate because they cannot generalize the properties of transformers operating in different contexts. In this context, this paper presents a novel transformer condition assessment approach integrating uncertainty modeling, data-driven forecasting models, and model-based experimental models to increase the prediction accuracy and handle uncertainty. The proposed approach quantifies the effect of measurement errors on transformer RUL predictions and confirms that temperature and load measurement errors affect the RUL estimation. Forecasting results show that the extreme gradient boosting (XGB) algorithm best captures the nonlinearities of the thermal model and improves the prediction accuracy among a number of forecasting approaches. Accordingly, the XGB model is integrated with experimental models in a particle filtering framework to improve thermal modeling and RUL prediction tasks. Models are tested and validated using a real dataset from a power transformer operating in a nuclear power plant. [ABSTRACT FROM AUTHOR]

Short-term household electricity load forecasting is important for utility companies to ensure reliable power supplies. Traditional methods for load forecasting relied on historical records from one single data source and have limitations with insufficient or missing data. Recently, an emerging family of machine learning algorithms, multitask learning (MTL), has been developed and has the potential for load forecasting. By MTL, the electricity consumption data from multiple communities can be fused to improve forecasting accuracy. However, appropriate modeling of the relatedness to enable the between-community knowledge transfer remains a challenge. This paper proposes an improved MTL algorithm for a Bayesian spatiotemporal Gaussian process model (BSGP) to characterize the relatedness among the different residential communities. It hypothesizes on the similar impacts of environmental and traffic conditions on electricity consumption in improving short-term load forecasting. Furthermore, this paper proposes a low-ranked dirty model along with an iterative algorithm to improve the learning of model parameters under an MTL framework. This paper uses a real-world case study from two residential communities in Tallahassee, Fl, USA, to demonstrate the method effectiveness. The proposed method significantly outperforms state-of-the-art forecasting methods and effectively captures the relatedness to provide between-community knowledge transfer compared with other MTL methods. [ABSTRACT FROM AUTHOR]

As one of the most promising types of fuel cells, the proton exchange membrane fuel cells (PEMFCs) can be utilized in many applications. However, it still faces two main challenges before large-scale industrial applications, namely short lifetime and high costs. The aim of this paper is to establish an accurate online short-term prognostics method to help users extend the lifetime and reduce the cost of PEMFCs. First, we compare the short-term prognostics accuracy and computational efficiency of several different methods including the Elman neural network, the group method of data handling, the adaptive neuro-fuzzy inference system (ANFIS) with different fuzzy inference system creation strategies, and the wavelet decomposition approach. Test results show that the ANFIS with fuzzy c-means (ANFIS-FCM) strategy has the best short-term prognostics performance. Then, we propose an automatic parameter adjustment method for ANFIS-FCM by using the particle swarm optimization (PSO) algorithm. Test results show that the PSO algorithm can effectively adjust the parameters and achieve improved prognostics results. Finally, the proposed prognostics methods are verified on a PEMFC experimental platform. Experimental results show that the proposed methods have great potential for practical applications. [ABSTRACT FROM AUTHOR]

The employment of solar microconverter allows a more detailed monitoring of the photovoltaic (PV) output power at the single module level; thus, machine learning techniques are capable to track the peculiarities of modules in the PV plants, such as regular shadings. In this way, it is possible to compare in real time the day-ahead forecast power with the actual one in order to better evaluate faults or anomalous trends that might have occurred in the PV plant. This paper presents a method for an effective fault diagnosis; this method is based on the day-ahead forecast of the output power from an existing PV module, linked to a microconverter, and on the outcome of the neighbor PV modules. Finally, this paper also proposes the analysis of the most common error definitions with new mathematical formulations, by comparing their effectiveness and immediate comprehension, in view of increasing power forecasting accuracy and performing both real-time and offline analysis of PV modules performance and possible faults. [ABSTRACT FROM AUTHOR]

This paper proposes a novel causality analysis approach called the causal Markov Elman network (CMEN) to characterize the interdependence among heterogeneous time series in multinetwork systems. The CMEN performance, which comprises inputs filtered by Markov property, successfully characterizes various multivariate dependencies in an urban environment. This paper also proposes a novel hypothesis of characterizing joint information between interconnected systems such as electricity and transportation networks. The proposed methodology and the hypotheses are then validated by information theory distance-based metrics. For cross validation, the CMEN is applied to the electricity load forecasting problem using actual data from Tallahassee, Florida. [ABSTRACT FROM AUTHOR]

A grid-tied microgrid (MG) with a battery energy storage system (BESS) is studied in this paper. The energy storage efficiencies of the BESS are considered to optimize the operational cost of the MG. Two quadratic functions are verified and utilized to formulate the efficiencies of BESS in both charge and discharge process. Afterward, constraints of MG power scheme are investigated based on aforementioned equations. Furthermore, the 24-h ahead forecasting data of photovoltaic (PV) generation and loads demand are also utilized during MG modeling. To minimize the operational electricity cost of the MG in the next 24 h, a nonlinear programming with discontinuous derivatives solver is applied based on the proposed constraints. Additionally, to balance the power flow of MG and reduce the effects of the forecasting error of PV generation, a two steps MG management strategy is therefore developed based on the scheduled power. Experiments are conducted to verify the relationship between battery energy storage efficiency and charging/discharging current of the lithium-ion battery. Moreover, the proposed energy management strategy is validated by the hardware-in-the-loop experiments for real-time MG operation. [ABSTRACT FROM AUTHOR]

Measurement and control of electric power quality (PQ) parameters in off-grid systems has played an important role in recent years. The purpose is to detect or forecast the presence of PQ parameter disturbances to be able to suppress or to avoid their negative effects on the power grid and appliances. This paper focuses on several PQ parameters in off-grid systems and it defines three evaluation criteria that are supposed to estimate the performance of a new forecasting model combining all the involved PQ parameters. These criteria are based on common statistical evaluations of computational models from the machine learning field of study. The studied PQ parameters are voltage, power frequency, total harmonic distortion, and flicker severity. The approach presented in this paper also applies a machine learning based model of random decision forest for PQ forecasting. The database applied in this task contains real off-grid data from long-term one-minute measurements. The hyperparameters of the model are optimized by multiobjective optimization toward the defined evaluation criteria. [ABSTRACT FROM AUTHOR]

To effectively find an optimal solution in the development process of electrical machines, it is essential to predict machine iron loss and thermal behavior in the predesign stage. Most existing publications only consider thermal behavior by roughly limiting the maximum current density. However, especially for high-speed application, iron loss can also be an essential part in thermal-behavior estimation. This paper proposes a simplified model to efficiently calculate iron loss and thermal behavior in the design procedure of switched reluctance machines. For this purpose, a maximum coenergy loop control method is implemented to facilitate the estimation of flux waveform and machine torque. Three different iron loss calculation methods are compared in terms of accuracy and time. Besides, a simplified lumped parameter thermal network is applied for machine sizing. Furthermore, the impact of number of turns and peak magnetomotive force on the losses and machine sizing is discussed. The efficiency and power density of various machine geometries are comparatively analyzed. Finally, the accuracy of the simplified model is verified by comparison with finite element simulations and experiment results. [ABSTRACT FROM AUTHOR]

An enormous challenge for state estimation (SE) in power distribution networks with high penetration of photovoltaic generators lies in how to deal with measurement uncertainty. Driven by this motivation, an interval arithmetic (IA) based SE is proposed in this paper, which considers interval modeling for measurement uncertainty. The proposed SE is formulated as two interval optimization models based on the unknown-but-bounded theory, and solution bounds of state variables are obtained using a two-stage linear programming approach. In particular, the proposed SE is extended to realistic power distribution networks with mixed microphasor measurement units (μPMU), voltage magnitude, smart meters and feeder terminal units. Case studies and results are illustrated and discussed to demonstrate the performance of the proposed IA-based SE. It could provide the operators intuitive and effective bounds of system states for advanced power industrial applications under consideration of uncertainty. [ABSTRACT FROM AUTHOR]

Energy imbalance market (EIM) provides an opportunity that allows larger shares of variable renewable energy sources in the grid. Under highly volatile weather conditions, an accurate forecasting of photovoltaic (PV) power is necessary for grid stability and market operation. Most of existing forecasting methods strongly rely on the accuracy of measurements, and the adaptability of these methods to complex weather conditions is rarely discussed. In this paper, a weather classification multivariate adaptive regression spline (MARS) forecasting model is introduced for complex weather conditions in all seasons. It can be updated incrementally and its high computational efficiency satisfies EIM operations. A data set that consists of the historical power and meteorological parameters produced by a small-scale PV platform is classified and used to train MARS models with forecast horizons ranging from 15 min to 24 h in different seasons. The tests and analyses results indicate higher accuracy, adaptability, and efficiency of the novel model. [ABSTRACT FROM AUTHOR]

The performance of the logistic system is one of the most important aspects in regional economy, and the freight volume is the biggest part of the logistic system. In this paper, an information granulation method is introduced to represent the freight volume in a fuzzy manner. After the characteristic features have been extracted from the raw time-series data and represented as information granules, the granules are modeled with the support vector machine (SVM). In consideration of both algorithm efficiency and prediction accuracy, an efficient version of SVM called least square (LS) SVM is employed and integrated with a parameter optimization algorithm, the particle swarm optimization. Simulation results on a real dataset illustrate the performance of the proposed method, and comparison studies are carried out with LS and partial LS-based methods. [ABSTRACT FROM AUTHOR]

In this paper, a generalized formulation for intelligent energy management of a microgrid is proposed using artificial intelligence techniques jointly with linear-programming-based multiobjective optimization. The proposed multiobjective intelligent energy management aims to minimize the operation cost and the environmental impact of a microgrid, taking into account its preoperational variables as future availability of renewable energies and load demand (LD). An artificial neural network ensemble is developed to predict 24-h-ahead photovoltaic generation and 1-h-ahead wind power generation and LD. The proposed machine learning is characterized by enhanced learning model and generalization capability. The efficiency of the microgrid operation strongly depends on the battery scheduling process, which cannot be achieved through conventional optimization formulation. In this paper, a fuzzy logic expert system is used for battery scheduling. The proposed approach can handle uncertainties regarding to the fuzzy environment of the overall microgrid operation and the uncertainty related to the forecasted parameters. The results show considerable minimization on operation cost and emission level compared to literature microgrid energy management approaches based on opportunity charging and Heuristic Flowchart (HF) battery management. [ABSTRACT FROM PUBLISHER]

Short-term forecasting of solar irradiance at a targeted site with consideration of its time series and measurements at neighboring sites is studied in this paper. A data-driven framework for forecasting solar irradiance based on fusing spatial and temporal information is proposed. In the framework, data-driven approaches including boosted regression trees (BRT), artificial neural network, support vector machine, and least absolute shrinkage and selection operator are applied to model spatial dependence among solar irradiance time series thereby to generate forecasts of solar irradiance at the targeted site. A comprehensive comparison among data-driven forecasting models is performed using 30-min averaged data of recent two years. Moreover, benchmarking models including scaled persistence model, autoregressive model, and autoregressive exogenous model are employed to further validate the effectiveness of data-driven forecasting models. Computational results of multiple-steps ahead forecasting demonstrate that the BRT model gives the best performance with the lowest normalized root mean squared error of 18.4%, 24.3%, 27.9%, and 30.6% for forecasting horizons of 30, 60, 90, and 120-min, respectively. [ABSTRACT FROM AUTHOR]

The remaining useful life (RUL) prediction of rolling element bearings has attracted substantial attention recently due to its importance for the bearing health management. The exponential model is one of the most widely used methods for RUL prediction of rolling element bearings. However, two shortcomings exist in the exponential model: 1) the first predicting time (FPT) is selected subjectively; and 2) random errors of the stochastic process decrease the prediction accuracy. To deal with these two shortcomings, an improved exponential model is proposed in this paper. In the improved model, an adaptive FPT selection approach is established based on the $3\sigma$ interval, and particle filtering is utilized to reduce random errors of the stochastic process. In order to demonstrate the effectiveness of the improved model, a simulation and four tests of bearing degradation processes are utilized for the RUL prediction. The results show that the improved model is able to select an appropriate FPT and reduce random errors of the stochastic process. Consequently, it performs better in the RUL prediction of rolling element bearings than the original exponential model. [ABSTRACT FROM PUBLISHER]

Finite control set model predictive control (FCS-MPC) strategies for power conversion devices benefit from extending the prediction horizon length. Solving this problem relies on the definition of the underlying integer least-squares problem. Sphere decoding algorithm (SDA) has been extensively used in previous works as an approach to solve this problem. In this article, a parallel and fully scalable K-best SDA hardware design is proposed as an alternative. The K-best SDA establishes a different breadth-first search strategy, which addresses some of the main drawbacks of the SDA. Through experimental tests based on an uninterruptible power supply, the K-best SDA performance for long prediction horizon FCS-MPC is assessed and verified. Results demonstrate key beneficial aspects through which the K-best SDA is capable of rendering an improved control performance when compared to the conventional SDA. [ABSTRACT FROM AUTHOR]

Passing cloud results in rapid changes of irradiance. The intermittency of photovoltaic (PV) power output has drawn serious concern especially for utility-scale PV system. Consequently, power ramp-rate control (PRRC) has been introduced to avoid significant PV power fluctuations. PRRC is usually implemented either by curtailing active power output or implementing energy storage system (ESS). However, current active power curtailment cannot deal with the irradiance ramp-down fluctuations, and the high cost of the ESS is still hindering its extensive application. In this paper, two innovative PRRC strategies are presented, which utilize the short-term forecasting. The first solution does not require any ESS, during the power ramp-down event, the PV generation will be curtailed before the actual shading occurs. The second solution requires only one-quarter of the energy capacity of the conventional ESS control strategy. To provide the PV generation forecasts, a dynamic model based on spatio-temporal theory is formulated. The effectiveness of the proposed forecasting model and control strategies have been verified through experiment and case studies. [ABSTRACT FROM AUTHOR]

The importance of railway transportation has been increasing in the world. Considering the current and future estimates of high cargo and passenger transportation volume in railways, prevention or reduction of delays due to any failure is becoming ever more crucial. Railway turnout systems are one of the most critical pieces of equipment in railway infrastructure. When incipient failures occur, they mostly progress slowly from the fault-free to the failure state. Although studies focusing on the identification of possible failures in railway turnout systems exist in literature, neither the detection nor forecasting of failure progression has been reported. This paper presents a simple state-based prognostic (SSBP) method that aims to detect and forecast failure progression in electromechanical systems. The method is compared with Hidden-Markov-Model-based methods on real data collected from a railway turnout system. Obtaining statistically sufficient failure progression samples is difficult, considering that the natural progression of failures in electromechanical systems may take years. In addition, validating the classification model is difficult when the degradation is not observable. Data collection and model validation strategies for failure progression are also presented. [ABSTRACT FROM AUTHOR]

On-road sensor systems installed on freeways are used to capture traffic flow data for short-term traffic flow predictors for traffic management, to reduce traffic congestion and improve vehicular mobility. This paper intends to tackle the impractical time-invariant assumptions which underlie the methods currently used to develop short-term traffic flow predictors: 1) the characteristics of current data captured by on-road sensors are assumed to be time invariant with respect to those of the historical data, which is used to developed short-term traffic flow predictors; and 2) the configuration of the on-road sensor systems is assumed to be time-invariant. In fact, both assumptions are impractical in the real world, as the current traffic flow characteristics can be very different from the historical ones, and also the on-road sensor systems are time varying in nature due to damaged sensors or component wear. Therefore, misleading forecasting results are likely to be produced when short-term traffic flow predictors are designed using these two time-invariant assumptions. To tackle these time-invariant assumptions, an intelligent particle swarm optimization (IPSO) algorithm is proposed to develop short-term traffic flow predictors by integrating the mechanisms of PSO, neural network and fuzzy inference system, to adapt to the time-varying traffic flow characteristics and the time-varying configurations of the on-road sensor systems. The proposed IPSO was applied to forecast traffic flow conditions on a section of freeway in Western Australia, whose traffic flow information can be captured on-line by the on-road sensor system. These results clearly demonstrate the effectiveness of using the proposed IPSO for real-time traffic flow forecasting based on traffic flow data captured by on-road sensor systems. [ABSTRACT FROM PUBLISHER]

Relevance vector regression (RVR) is a useful tool for degradation modeling and remaining useful life (RUL) prediction. However, most RVR models are for 1-D degradation processes and can only handle univariate observations. This article proposes a degradation path-based RUL prediction framework using a dynamic multivariate relevance vector regression model. Specifically, a multistep regression model is established for describing the degradation dynamics and extending the classical RVR into a multivariate one with consideration of the multivariate environment. The article introduces a matrix Gaussian distribution-based RVR approach and then estimates the hyperparameters with Nesterov’s accelerated gradient method to avoid the exhausting re-estimation phenomenon in seeking analytical solutions. It further forecasts the degradation path for monitoring the degradation status. Based on the forecasted path, the RUL is predicted by the first hitting time method. Finally, the proposed methods are illustrated by two case studies, one is presented in this article and the other in the supplement, which investigate the capacitors’ and bearings’ performance degradations. [ABSTRACT FROM AUTHOR]

A lithium-ion battery has advantages such as high energy density and long calendar life, but it suffers from the risk of thermal runaway. Overcharge-induced thermal runaway accidents hold a considerable percentage. This article discovers that the slope of the dynamic impedance in the frequency band of 30–90 Hz turns positive from negative when the cell just starts to overcharge and proposes the theoretical explanation. Taking 70 Hz impedance as an example, the thermal runaway accident can be successfully avoided by cutting off the charging when the slope turns positive from negative during charging. The warning time is 580 s ahead of the thermal runaway. This feature is easy to identify and requires no complex mathematical models and parameters. Besides, the prediction method based on this feature can be conducted by using an online dynamic impedance measurement device designed by us, which is suitable for large-scale applications. Thus, the overcharge-induced thermal runaway accidents can be avoided. [ABSTRACT FROM AUTHOR]

Machine remaining useful life (RUL) prediction is vital in improving the reliability of industrial systems and reducing maintenance cost. Recently, long short-term memory (LSTM) based algorithms have achieved state-of-the-art performance for RUL prediction due to their strong capability of modeling sequential sensory data. In many cases, the RUL prediction algorithms are required to be deployed on edge devices to support real-time decision making, reduce the data communication cost, and preserve the data privacy. However, the powerful LSTM-based methods which have high complexity cannot be deployed to edge devices with limited computational power and memory. To solve this problem, we propose a knowledge distillation framework, entitled KDnet-RUL, to compress a complex LSTM-based method for RUL prediction. Specifically, it includes a generative adversarial network based knowledge distillation (GAN-KD) for disparate architecture knowledge transfer, a learning-during-teaching based knowledge distillation (LDT-KD) for identical architecture knowledge transfer, and a sequential distillation upon LDT-KD for complicated datasets. We leverage simple and complicated datasets to verify the effectiveness of the proposed KDnet-RUL. The results demonstrate that the proposed method significantly outperforms state-of-the-art KD methods. The compressed model with 12.8 times less weights and 46.2 times less total float point operations even achieves a comparable performance with the complex LSTM model for RUL prediction. [ABSTRACT FROM AUTHOR]

Semi-observable systems are referred to as a kind of widely used industrial equipment whose physical degradation state is only observable via shutdown inspection. To monitor the degradation process of semi-observable systems online, different types of sensors are generally employed to collect monitoring signals. Lots of studies have been conducted to fuse multi-sensor signals to predict remaining useful life (RUL). Majority of them, however, ignored the partially available state observations which can be viewed as ground truth measurements of physical degradation. To deal with this problem, this article proposes a multi-sensor data-driven RUL prediction method for semi-observable systems, which leverages degradation information from online multi-sensor signals as well as offline state observations. This method is developed based on a generalizable state-space model combined with particle filtering framework. In the framework, a state transition function is used to describe the degradation process of system states. A multidimensional measurement function is constructed to describe the mapping between states and multi-sensor signals. To enhance the performance of prediction, an algorithm named prioritized sensor group selection is also proposed to select the optimal sensor group for RUL prediction. The effectiveness of the proposed method is demonstrated using an experiment of cutting tool wear. [ABSTRACT FROM AUTHOR]

In order to simultaneously provide excellent tracking performance as well as good generalization for precision motion control of magnetically levitated (maglev) planar motors, this article proposes a feedforward compensation strategy with precise reference modification prediction using gated recurrent units (GRU). First, a GRU network is constructed, which is capable of precisely predicting the reference modification after training with data collected from a maglev planar motor. Essentially, the GRU network can be viewed as a data-based model which captures the dynamic characteristics of appropriate reference modification and gives a prediction even before implementing the real trajectory. Consequently, an excellent feedforward compensation part can be formed, which is important in the whole control structure. Comparative experimental investigation validates the effectiveness of the GRU-based modification prediction ability and demonstrates the excellent tracking performances of the proposed control strategy. The proposed strategy not only performs as well as iterative learning control, but also does not need any time-consuming iterations, which is valuable for industrial applications. [ABSTRACT FROM AUTHOR]

Uncertainty modeling of renewable energy sources, load demand, electricity price, etc. create a high volume of data in smart grids. Accordingly, in this article, a precise forecasting method based on a deep learning concept with microclustering (MC) task is presented. The MC method is structured based on hybrid unsupervised and supervised clustering tasks by K-means and Gaussian support vector machine, respectively. In the proposed method, the input data sequence is clustered by the MC task, and then the forecasting process is employed. By applying the MC, input data in each hour are categorized into different groups, and a distinctive forecasting unit is allocated to each one. In this way, more clusters and forecasting networks are earmarked for the hours with higher fluctuation rates. The bi-directional long short-term memory (B-LSTM), which is one of the newest recurrent artificial neural networks, is proposed as the forecasting unit. The B-LSTM has bidirectional memory—feedforward and feedback loops—that helps us to investigate both previous and future hidden layers data. The optimal number of clusters in each hour is determined based on the Davies–Bouldin index. To evaluate the performance of the proposed method, in this study, three forecasting tasks including the wind speed, load demand, and electricity price are studied in different periods using the Ontario province, Canada, data set. The results are compared with other benchmarking methods to verify the robustness and effectiveness of the proposed method. In fact, the proposed method, which is equipped with the MC technique and B-LSTM networks, significantly promotes the forecasting results, especially in spike points. [ABSTRACT FROM AUTHOR]

Surface Electromyography (sEMG) based human–robot interaction has been widely studied, where prediction of human voluntary torques is one of the key issues that have not been well addressed. In this article, a torque prediction method based on collaborative neuromusculoskeletal modeling and adaptive learning, is proposed to overcome the limitation of existing methods. First, an sEMG-torque model is designed in comprehensive consideration of the previous research results, the requirement for subject-specific adjustment and the coupling between the muscle or muscle-tendon length and the adjacent joint angles, where the latter two factors have rarely been considered in the literature. Then, by combining the advantages of the stochastic particle swarm optimization and conjugate gradient algorithms, a collaborative optimization method is designed to calibrate simultaneously the undetermined parameters. Moreover, an adaptive learning method based on Gaussian process regression is proposed to learn and predict the estimation errors in real time, by which it is supposed that the torque prediction accuracy can be improved efficiently. Finally, experiments were carried out to validate the performance of the proposed method. [ABSTRACT FROM AUTHOR]

Finite-set model predictive current control (FS-MPCC) has been employed for machine drives owing to the good dynamic performance. Sensitivity to parameter variation is one of the main barriers to its widespread application. To overcome this barrier, this article proposes an improved FS-MPCC for surfaced permanent magnet synchronous machines (SPMSMs). The contribution of this article is developing a novel current update mechanism, where the variation of resistance, rotor flux linkage, and inductance is considered, in which a modified coordinate system that contains a stationary and rotary axis frame is introduced. In this case, the predicted parameter can be obtained more accurately based on this mechanism, which can suppress the disturbances caused by model parameter mismatch. In addition, SPMSM parameter mismatch effect on FS-MPCC performance is analyzed in this article, which testifies that resistance mismatch influence on current prediction error can be neglected in practical system. The proposed FS-MPCC is validated and compared with different state-of-the-art methods. Improvement of the proposed control strategy is validated by means of experimental results on a 2-kW test rig. [ABSTRACT FROM AUTHOR]

For prognostics and health management of mechanical systems, a core task is to predict the machine remaining useful life (RUL). Currently, deep structures with automatic feature learning, such as long short-term memory (LSTM), have achieved great performances for the RUL prediction. However, the conventional LSTM network only uses the learned features at last time step for regression or classification, which is not efficient. Besides, some handcrafted features with domain knowledge may convey additional information for the prediction of RUL. It is thus highly motivated to integrate both those handcrafted features and automatically learned features for the RUL prediction. In this article, we propose an attention-based deep learning framework for machine's RUL prediction. The LSTM network is employed to learn sequential features from raw sensory data. Meanwhile, the proposed attention mechanism is able to learn the importance of features and time steps, and assign larger weights to more important ones. Moreover, a feature fusion framework is developed to combine the handcrafted features with automatically learned features to boost the performance of the RUL prediction. Extensive experiments have been conducted on two real datasets and experimental results demonstrate that our proposed approach outperforms the state-of-the-arts. [ABSTRACT FROM AUTHOR]

In this article, a variable switching frequency full zero-voltage switching (ZVS) range bidirectional three-phase voltage-source inverter is proposed. Discontinuous pulsewidth modulation is used so that at any time only two phase legs need to achieve ZVS. The boundary switching frequency can be calculated in a digital controller based on current ripple prediction instead of current zero-crossing detection. No additional high-frequency sensor or auxiliary circuit is needed. The frequency variation range is narrow in a line cycle at any load condition. An LCL filter is used to attenuate the inverter-side high ripple current and improve the grid current quality. Owing to the high switching frequency and the small filter parameters, the power density can be greatly increased. Both the device and the filter losses are reduced, and the efficiency is high. A 3.5 kW simulation and experimental prototype with silicon carbide (SiC) device interfacing 400 V dc with three-phase 110 V ac grid is developed to verify the effectiveness of the proposed control strategy. [ABSTRACT FROM AUTHOR]

This article proposes a novel model predictive control for a three-phase permanent-magnet synchronous motor (PMSM) with enhanced robustness against parameter variation and higher current control precision. First, the extended set of voltage vectors are adopted to increase the current control precision. Then, the parameter sensitivity problem is avoided in the first stage using an indirect reference vector strategy for the sake of computation time reduction. Subsequently, the inductance disturbance observer is adopted in the current prediction process to enhance the robustness against machine parameter mismatch. The intrinsic connection between the discrete control sets precision and disturbance observer is analyzed. The proposed method improves the robustness against parameter mismatch and current control precision simultaneously. Experimentations are conducted to verify the effectiveness of the proposed method. [ABSTRACT FROM AUTHOR]

To simultaneously achieve accurate tracking error prediction, rigorous motion accuracy, and certain robustness to parameter variations and unknown disturbances, this article proposes a data-based learning adaptive robust control (LARC) strategy based on gated recurrent unit (GRU) neural network. Firstly, parameter adaptive control and robust control are utilized to guarantee the robustness against parametric uncertainties and unknown disturbances. A GRU neural network is then constructed and capable of precisely predicting the tracking error after training with data collected from a linear-motor-driven stage. Essentially, the GRU network can be viewed as a data-based model, which captures the tracking error dynamic characteristics and provides a prediction even before implementing the real trajectory. Consequently, a reference modification and a feedforward compensation part can be formed, which is the significant part of the whole LARC control structure. Comparative experimental investigation not only validates the effectiveness of the tracking error prediction ability, but also demonstrates the practically satisfactory transient/steady-state tracking performance of the proposed control strategy. [ABSTRACT FROM AUTHOR]

Object detection is significant for event analysis in various intelligent multimedia processing systems. Although there have been many studies conducting research in this area, effective and efficient object detection methods for video sequences are still much desired. In this article, we investigate salient object detection in real-time multimedia processing systems. Considering the intrinsic relationship between top-down and bottom-up saliency features, we present a new effective method for video salient object detection based on deep semantic and spatiotemporal cues. After extracting top-down semantic features for object perception by a 2-D convolutional network, we concatenate them with bottom-up spatiotemporal cues for motion perception extracted by a 3-D convolutional network. In order to combine these features effectively, we feed them into a 3-D deconvolutional network for feature-sharing learning between semantic features and spatiotemporal cues for the final saliency prediction. Additionally, we propose a novel Gaussian-like loss function with an L2-norm regularization term for parameter learning. Experimental results show that the proposed salient object detection approach performs better in terms of both effectiveness and efficiency for video sequences compared with the state-of-the-art models. [ABSTRACT FROM AUTHOR]

In this article, a kind of predictive control (PC) for permanent magnet synchronous motor (PMSM) rotor position control is introduced, and more stable and rapid performances are achieved comparing with other PCs. Nonlinear predictive control (NLPC) is shown to outperform the linear in terms of performance for nonlinear controlled plants because of containing nonlinear part and longer prediction horizons. A novel direct speed NLPC scheme for PMSM is presented based on this strategy. The response and reference signals are predicted with Taylor series expansion to guarantee the accuracy of the system. Prediction horizons are selected according to Shannon with its maximum value to eliminate steady-state error and delay between actual and reference signals. Furthermore, the proposed method is designed and analyzed in detail, and tested by simulations and experiments. Results with respect to the rotor position trace tracking, control performance, and weighting factors sensitivity demonstrate the efficiency of the proposed method. [ABSTRACT FROM AUTHOR]

This article presents a novel degradation modeling and prognostic method for a class of closed-loop feedback systems with degrading actuators. Toward this end, we first present a degradation modeling framework by integrating the stochastic degradation process model of the actuator and the state transition model of the system. This takes into consideration the mutual effects between the component-level degradation and system-level state. Then, the particle filter algorithm is utilized to jointly estimate the hidden degradation state of the actuator and the system state through indirect observations. Further, a time-varying nonlinear diffusion process equipped with two-stage parameters updating procedure is used to learn the evolving progression of the hidden degradation state. As such, a residual-threshold-based remaining useful life (RUL) prediction method is presented by simulating future system states and degradation trajectories based on the learned degradation process. As the application of the predicted RUL, a fault tolerant control method is presented by adjusting the controller parameter so as to extend the life of the system. Finally, a simulation study is conducted using a closed-loop control system in an inertial platform to verify the proposed method. The results indicate that the proposed method can reduce prognosis error, improve robustness, and extend the lifetime of the system. [ABSTRACT FROM AUTHOR]

In this article, a fully passive, wireless solution for out-of-sight pipeline monitoring is presented. By predicting corrosion before its occurrence, a proactive response may be taken to mitigate the chance of an environmental disaster. The chipless radio-frequency identification system is a combination of a tag ID, consisting of an array of six rectangular spiral resonators, and a tag antenna consisting of two cross-polarized novel patch antennas etched on a skin-thin microwave laminate. The tags are grounded on a carbon steel pipeline coated with 3 mm Teflon (polyethylene-like electrical characteristics). The proposed system exhibits robust capability to create frequency signatures to detect and monitor defects underneath the pipeline coating owing to water ingress that could eventuate to corrode the pipeline. The reader antenna above the pipeline comprises two identically cross-polarized log periodic dipole antennas with the intent to remotely capture resonances of the tag ID in real-time for early detection and prediction of potential pipeline exposure to moisture. The proposed structure provides low-cost real-time solution. [ABSTRACT FROM AUTHOR]

For the open-winding permanent-magnet synchronous generator system with a common dc bus, in order to improve the steady-state control performance of conventional prediction control and simultaneously reduce the calculation burden, a zero-sequence current suppression method based on the prediction of three-dimensional reference voltage is proposed in this article. First, the voltage vector used by prediction control is extended from the plane to the three-dimensional (3-D) space. The voltage vectors of two converters are analyzed and predicted under the 3-D space. Moreover, to improve steady-state performance, two voltage vectors are applied in one control period, and the voltage selection principle that determines candidate vectors is presented. Finally, experimental results test the effectiveness of this method. [ABSTRACT FROM AUTHOR]

Recently, machine learning (ML) has been widely adopted for fingerprint-based indoor localization because of its potency in delineating relationships between received signal strength (RSS) information and labels accurately. Existing ML-based indoor localization systems are less robust because they only adopt the output with the highest probability. This affects the final location estimate, hence compromising accuracy due to the severity of RSS fluctuations. Since different ML algorithms (MLAs) yield different performances, it is therefore intuitive to fuse predictions from multiple MLAs to improve the positioning performance in the presence of signal fluctuation. In this article, we propose SmartLoc, a smart wireless indoor localization framework to enhance indoor localization. In the offline phase, multiple MLAs are trained by utilizing an offline database. We further apply probability alignment to guarantee the predicted probabilities of each MLA at the same confidence level. In the online phase, given a testing RSS sample of a user at an unknown location, we extract the labels with probabilities greater than a certain threshold from each MLA to construct the space of candidate labels (SCL). The size of SCL can be adaptively determined by using our proposed dynamic size determination algorithm. Based on the SCL, we propose a probabilistic model to intelligently estimate the user's location by evaluating the label credibility simultaneously. A high label credibility indicates that the frequently occurred label is more likely to be true. Experimental results in a real changing environment verify the superiority of SmartLoc, outperforming the best among comparative methods by 10.8% in 75th percentile accuracy. [ABSTRACT FROM AUTHOR]

In hybrid dynamical systems, degradation evolution of components depends on the operating modes of the system, which are activated by discrete events. In a hybrid system's dynamics, some parameter degradations may not be observable from the measured variables in every state of the discrete components. Thus, the degradation model identifications (DMI) for hybrid components’ parameters are difficult and challenging due to unobservable conditions in some particular working modes of the system. This article proposes a distributed parameter estimation-based adaptive prognosis for multicomponent hybrid system by using bond graph modeling as a unified framework. Here, degradation states of different parameters are progressively estimated by including the information of operating modes of the system. A bank of mode-dependent degradation models are identified and extrapolated for remaining useful life (RUL) prediction of the degrading components. In this article, special emphasis is given to DMI and RUL prediction for hybrid components. The results are supported by an in-depth experimental study. [ABSTRACT FROM AUTHOR]

This article proposes a double-line-frequency current ripple (DLFCR) reduction strategy for the photovoltaic microinverter. A power predictive scheme is established according to the output power mathematical model of the prestage full-bridge dc–dc converter (PFDDC), which is the basis of the proposed DLFCR reduction strategy. The control effect is analyzed with considering parameter estimation error, and the analysis results show that the proposed strategy has good dynamic performance and robust performance. The small-signal model of the PFDDC has been established to design the voltage-loop parameters. The bandwidth of the system controlled by the power predictive scheme is large, which can effectively reduce the design complexity because no parameter is needed to design for power prediction. Experimental verifications are presented to verify the analysis. [ABSTRACT FROM AUTHOR]

Photovoltaic (PV) power is volatile in nature and raises the level of uncertainty in power systems. PV power forecasting is an important measure to solve this problem. It helps to improve the reliability and reduces the generation cost. Advances in computer technology and sensors make the numeric modeling methods a hotspot in the field of PV power forecasting. However, data modeling methods strongly rely on the accuracy of measurement data. Unavoidable outliers in the measured meteorological data have an adverse effect on the model due to their heteroscedasticity. Although many studies can be found focusing on outlier detection, only a few have incorporated outlier detection with regression models. In this study, an innovative method employing the weighted Gaussian process regression approach is proposed, such that data samples with higher outlier potential have a low weight. A density-based local outlier detection approach is introduced to compensate the deterioration of Euclidean distance for high-dimensional data. A novel concept of the degree of nonlinear correlation is incorporated to compute the contribution of every individual data attribute. Effectiveness of the proposed method is demonstrated by performing an experimental analysis and making comparisons with other typical data-based approaches, and the results exhibit higher estimation accuracy. [ABSTRACT FROM PUBLISHER]