Search

Evaluation and analysis of wind energy in large-scale wind farms is the pivotal link with regard to the wind farm's siting and economic benefit evaluation. For numerous years, several statistical distributions have been utilized to assess and analyze wind energy. However, the determination of the optimal distribution for the effective evaluation and analysis of wind energy is still a difficult and challenging task. In this study, the selection strategy is developed for the establishment of the optimal statistical model for wind energy assessment and analysis on the basis of the root-mean-square error, and the best outcome is obtained through simulation calculations based on the Weibull distribution. To further improve the fitting accuracy, three artificial intelligence algorithms—namely, the grey wolf optimizer, particle swarm optimization, and cuckoo search algorithms—and four numerical methods, are utilized to ascertain the optimal parameters for the Weibull model. The experimental results indicate that the grey wolf optimizer algorithm presents the most efficient and accurate methodology for the estimation of the Weibull distribution parameters. Therefore, the grey wolf optimizer algorithm is particularly suitable for the assessment and analysis of wind energy in large-scale wind farms.

Wind power forecasting is critical to the safe running of the power grid. However, due to the strong intermittence and instability of wind, reliable forecast of wind power remains a significant difficulty. In this study, a novel multivariable machine learning hybrid prediction system that incorporates data preprocessing, prediction, and multi-objective system optimization is designed to quantify the certainty and uncertainty of wind power. To increase the quality of data input, the data preparation module performs outlier tests based on the correlation between wind power and wind speed, as well as feature extraction, on the original data. In the prediction process, this paper offers an incremental kernel extreme learning machine (IK-elm), the parameters of which are set synchronously by an enhanced multi-objective optimization technique (MOCEHHO) developed in this paper. It overcomes the restrictions of duplicated hidden layer nodes and low learning efficiency caused by classic ELM and successfully maximizes the model's prediction capabilities. The simulation results on four datasets from Turkish wind farms show that the hybrid forecasting system outperforms the benchmark and may be utilized as a useful tool for wind power forecasting.

Electrical power system (EPS) forecasting plays a significant role in economic and social development but it remains an extremely challenging task. Because of its significance, relevant studies on EPS are especially needed. More specifically, only a few of the previous studies in this area conducted in-depth investigations of the entire EPS forecasting and merely focused on modeling individual signals centered on wind speed or electrical load. Moreover, most of these past studies concentrated on accuracy improvements and usually ignore the significance of forecasting stability. Therefore, to simultaneously achieve high accuracy and dependable stability, a hybrid forecasting framework based on the multi-objective dragonfly algorithm (MODA) was successfully developed in this study. The framework consists of four modules—data preprocessing, optimization, forecasting, and evaluation modules. In this framework, MODA is employed to optimize the Elman neural network (ENN) model as a part of the optimization module to overcome the drawbacks of single-objective optimization algorithms. In addition, data preprocessing and evaluation modules are incorporated to improve forecasting performance and conduct a comprehensive evaluation for this framework, respectively. Empirical results reveal that the developed forecasting framework can be an effective tool for the planning and management of power grids.

Forecasting petroleum consumption is a complicated and challenging task because many parameters affect the oil consumption. Whereas a highly accurate prediction model can help one utilize data resources reasonably, an inaccurate model will lead to a waste of resources. Thus, choosing an optimization model with the best forecasting accuracy is not only a challenging task but also a remarkable problem for oil consumption forecasting. However, a single model cannot always satisfy time series forecasting and the variations in oil consumption. In this paper, a total of 26 combination models using traditional combination method were developed to increase the prediction accuracy and avoid the problem of individual risk prediction methods "over-fitting", which would reduce the accuracy. Our conclusion is that the proposed combination models provide desirable forecasting results compared to the traditional combination model, and the combination method of TCM-NNCT is the most feasible and effective one. This paper also discussed the factors related to the statistical models and the results can be used by policy makers to plan strategies. Numerical results demonstrated that the proposed combined model is not only robust but able to approximate the actual consumption satisfactorily, which is an effective tool in analysis for the energy market.

Accurate wind speed forecasting is increasingly essential for improving the operating efficiency of electric power systems. Numerous models have been proposed to obtain the accurate and stable wind speed forecasting results. However, previous proposed models are limited by single predictive model or cannot deal with complex nonlinear data characteristic, which resulted in poor and unstable prediction results. In this paper, a novel forecasting model that combines noise processing, statistical approaches, deep learning frameworks and multi-objective optimization algorithm is proposed. Multi-objective optimization algorithms can take advantage of the merits of benchmark prediction models to address nonlinear characteristics of wind speed series. The 10-min real wind speed data from three Sites in China are adopted for verifying the effectiveness of this proposed model. The experimental results of multi-step prediction show that the model achieves MAPE1-step = 2.2109%, MAPE2-step = 3.0309%, and MAPE3-step = 4.2536% at Site 1; MAPE1-step = 2.4586%, MAPE2-step = 3.2034%, and MAPE3-step = 4.6843% at Site 2; MAPE1-step = 2.3180%, MAPE2-step = 3.0846%, and MAPE3-step = 4.4193% at Site 3. Therefore, the forecasting performance of this model is excellent, and it is beneficial to the dispatching and planning of power grid.

Reliable wind speed forecasting is crucial for the operation of wind power systems, and many efforts have been made to develop methods for wind speed prediction. However, most of them ignored feature extraction from original data, leading to poor performance. In this study, a novel combined forecasting system is proposed based on a two-stage data preprocessing technique, three component forecasting models and a novel combination method of a multi-objective optimization algorithm to compensate for their shortcomings. Through the two-stage data preprocessing, the raw data is decomposed and reshaped to reduce noisy and chaotic disturbance, which improves the quality of data input. The forecasting module uses component forecasting and a combination strategy that takes advantages of each model to achieve both accurate and stable results. Four 10-min wind speed datasets are employed for experiments, and the results of deterministic and probabilistic forecasting indicate that the proposed system achieves optimal accuracy and robustness comparing with contrastive models. For point and interval forecasting, the system achieves 3.1112%, 4.7375%, 2.7459%, and 2.1110% mean absolute percent errors and 96.6667%, 100%, 97.3333%, and 98% interval coverage probabilities for spring, summer, autumn and winter dataset, respectively, connoting a considerable potential for application in wind power production.

Wind speed forecasting plays a vital role in power system management, planning and integration. In previous studies, most forecasting models have focused on improving the accuracy or stability of wind speed forecasting. However, for an effective forecasting model, considering only one criterion (accuracy or stability) is insufficient. In this paper, a novel combined forecasting model was proposed and successfully employed to solve the problem of simultaneously obtaining both high accuracy and strong stability in wind speed forecasting. The proposed model consists of four ANNs (artificial neural networks) with optimum weight coefficients based on MOBA (multi-objective bat algorithm). MOBA overcomes the defect that only one criterion can be achieved by single objective optimization algorithms. In addition, data decomposition and de-noising are also incorporated in the data pre-processing stage. Ten-minute wind speed data from three datasets in Penglai, China, were selected for multi-step ahead forecasting to evaluate the effectiveness of the developed combined model. The experimental results indicate that the combined model outperforms other comparison models for generating forecasts in terms of forecasting accuracy and stability.

Wind speed forecasting is a crucial issue in the wind power industry. However, the disadvantage of the existing wind speed forecasting models is that they often ignore similar fluctuation information between the adjacent WTGs (wind turbine generators), which leads to poor forecasting accuracy. This paper proposes a hybrid wind speed forecasting model to overcome this disadvantage. Specifically, grey correlation analysis is applied to select useful fluctuation information from the adjacent and observed WTGs, and the chosen fluctuation information is fed into the v-SVM (v-support vector machine), which offers good capability in nonlinear fitting, to perform wind speed forecasting of the observed WTGs. Meanwhile, to reduce the impacts of the model parameters on the final forecasting performance, CS (cuckoo search) is used to tune the parameters in the v-SVM. The results from two case studies show that the proposed model, which considers the fluctuation information of the adjacent WTG, offers greater accuracy than the other compared models. As concluded from the results of three accuracy tests, the performances of v-SVM and e-SVM (e-support vector machine) show no significant difference, and the CS algorithm is more efficient than the PSO (particle swarm optimization) for tuning of the parameters in the v-SVM.

With the increasing importance of wind power as a component of power systems, the problems induced by the stochastic and intermittent nature of wind speed have compelled system operators and researchers to search for more reliable techniques to forecast wind speed. This paper proposes a combination model for probabilistic short-term wind speed forecasting. In this proposed hybrid approach, EWT (Empirical Wavelet Transform) is employed to extract meaningful information from a wind speed series by designing an appropriate wavelet filter bank. The GPR (Gaussian Process Regression) model is utilized to combine independent forecasts generated by various forecasting engines (ARIMA (Autoregressive Integrated Moving Average), ELM (Extreme Learning Machine), SVM (Support Vector Machine) and LSSVM (Least Square SVM)) in a nonlinear way rather than the commonly used linear way. The proposed approach provides more probabilistic information for wind speed predictions besides improving the forecasting accuracy for single-value predictions. The effectiveness of the proposed approach is demonstrated with wind speed data from two wind farms in China. The results indicate that the individual forecasting engines do not consistently forecast short-term wind speed for the two sites, and the proposed combination method can generate a more reliable and accurate forecast.

Short-term wind speed forecasting plays a major role in wind energy plant operations and the integration of wind power into traditional grid systems. This paper proposes a hybrid model, which is composed of the EWT (empirical wavelet transform), PACE (partial auto-correlation function) and GPR (Gaussian process regression) method, for short-term wind speed prediction. In this proposed approach, the EWT is employed to extract meaningful information from the wind speed series by designing an appropriate wavelet filter bank, and the GPR simulates the internal uncertainties and dynamic features of the wind speed time-series using inputs identified by the PACF. The hybrid GPR model can offer point predictions and interval estimations of future wind speed. Additionally, this study adopts a moving window approach in the prediction process to deal adequately with the training data set, thereby adapting to the time-varying characteristic of the wind speed. The proposed hybrid model was validated with real mean half-hour wind speed data and hourly wind speed data. The computational results show that the suggested hybrid model favorably improves point wind speed forecasts in comparison with other models and provides satisfactory interval wind speed prediction.

Forecasting wind speed is a crucial issue in the wind power industry, and wind energy is gradually coming to be viewed as one of the most promising alternative energy sources because of its cleanliness and renewability for large-scale commercial production. With the aim of developing accurate tools for forecasting wind speed, this paper presents a novel hybrid intelligent forecasting model based on Least Square Support Vector Machine and the Markov model. Prior to conducting the wind speed forecasting, the original wind speed series is processed using the C–C method, which is a phase space reconstruction method, to automatically determine the input form. Subsequently, the LSSVM (least squares support vector machine) model, which is optimized using the PSOGSA (partical swarm optimization combined with gravitational search algorithm) algorithm, is employed to forecast wind speed. Additionally, the Markov model is developed to make error corrections with the state ranges determined using the FCM (fuzzy C-means) model. The proposed model takes the average forecasting values between the outputs generated by LSSVM and error corrected LSSVM model as the final forecasting results. The simulation results indicate that the proposed model can outperform the other models discussed in this paper with respect to forecasting accuracy and can be a suitable tool for forecasting wind speeds.

Electrical load forecasting has always played a key role in power system administration, planning for energy transfer scheduling and load dispatch. For electrical load forecasting, due to the fact that combined model has the capacity to effectively calculate the seasonality and nonlinearity shown in the electrical load data, absorb the merits and avoid the limitations of the individual models, a new combined model is presented. In this model, the data pre-analysis is used to reduce the interferences from the data, meanwhile cuckoo search is firstly used to optimize weight coefficients of the combined model. To evaluate the forecast performance of the proposed combined model, half-hourly electricity power data from February 2006 to 2009 for the State of New South Wales, August 2006 to 2008 for the State of Victoria and November 2006 to 2008 for the State of Queensland, Australia, were used in this paper as a case study. The experimental results show that the proposed combined model is superior to the individual forecasting models regarding forecast performance.

Due to energy crisis and environmental problems, it is very urgent to find alternative energy sources nowadays. Solar energy, as one of the great potential clean energies, has widely attracted the attention of researchers. In this paper, an optimized hybrid method by CS (Cuckoo Search) on the basis of the OP-ELM (Optimally Pruned Extreme Learning Machine), called CS-OP-ELM, is developed to forecast clear sky and real sky global horizontal radiation. First, MRSR (Multiresponse Sparse Regression) and LOO-CV (leave-one-out cross-validation) can be applied to rank neurons and prune the possibly meaningless neurons of the FFNN (Feed Forward Neural Network), respectively. Then, Direct strategy and Direct-Recursive strategy based on OP-ELM are introduced to build a hybrid model. Furthermore, CS (Cuckoo Search) optimized algorithm is employed to determine the proper weight coefficients. In order to verify the effectiveness of the developed method, hourly solar radiation data from six sites of the United States has been collected, and methods like ARMA (Autoregression moving average), BP (Back Propagation) neural network and OP-ELM can be compared with CS-OP-ELM. Experimental results show the optimized hybrid method CS-OP-ELM has the best forecasting performance.

This study proposes a hybrid forecasting approach that consists of the EWT (Empirical Wavelet Transform), CSA (Coupled Simulated Annealing) and LSSVM (Least Square Support Vector Machine) for enhancing the accuracy of short-term wind speed forecasting. The EWT is employed to extract true information from a short-term wind speed series, and the LSSVM, which optimizes the parameters using a CSA algorithm, is used as the predictor to provide the final forecast. Moreover, this study uses a rolling operation method in the prediction processes, including one-step and multi-step predictions, which can adaptively tune the parameters of the LSSVM to respond quickly to wind speed changes. The proposed hybrid model is demonstrated to forecast a mean half-hour wind speed series obtained from a windmill farm located in northwestern China. The simulation results suggest that the developed forecasting method yields better predictions compared with those of other popular models, which indicates that the hybrid method exhibits stronger forecasting ability.

Wind energy is regarded as a worldwide renewable and alternative energy that can relieve the energy shortage, reduce environmental pollution, and provide a significant potential economic benefit. In this paper, a hybrid method is developed to properly and efficiently forecast the daily wind speed in Hainan Province, China. The proposed hybrid forecasting model consists of outlier detection and a bivariate fuzzy time series, which provides a more powerful forecasting capacity of daily wind speed than that of traditional single forecasting models. To verify the developed approach, daily wind speed data from January 2008 to December 2012 in Hainan Province, China, are used for model construction and testing. The results show that the developed hybrid forecasting model achieves high forecasting accuracy and is suitable for forecasting the wind energy of China's large wind farms.

For accurate electricity demand forecasting, this paper proposes a novel approach, MFES, that combines a multi-output FFNN (feedforward neural network) with EMD (empirical mode decomposition)-based signal filtering and seasonal adjustment. In electricity demand forecasting, noise signals, caused by various unstable factors, often corrupt demand series. To reduce these noise signals, MFES first uses an EMD-based signal filtering method which is fully data-driven. Secondly, MFES removes the seasonal component from the denoised demand series and models the resultant series using FFNN model with a multi-output strategy. This multi-output strategy can overcome the limitations of common multi-step-ahead forecasting approaches, including error amplification and the neglect of dependency between inputs and outputs. At last, MFES obtains the final prediction by restoring the season indexes back to the FFNN forecasts. Using the half-hour electricity demand series of New South Wales in Australia, this paper demonstrates that the proposed MFES model improves the forecasting accuracy noticeably comparing with existing models.

Electric load forecasting is an important task in the daily operations of a power utility associated with energy transfer scheduling, unit commitment and load dispatch. Inspired by the various non-linearity of electric load data and the strong learning capacity of support vector regression (SVR) for small sample and balanced data, this paper presents an adaptive fuzzy combination model based on the self-organizing map (SOM), the SVR and the fuzzy inference method. The adaptive fuzzy combination model can effectively count for electric load forecasting with good accuracy and interpretability at the same time. The key idea behind the combination is to build a human-understandable knowledge base by constructing a fuzzy membership function for each homogeneous sub-population. The comparison of different mathematical models and the effectiveness of the presented model are shown by the real data of New South Wales electricity market. The obtained results confirm the validity of the developed model.

Wind energy has been well recognized as a renewable resource in electricity generation, which is environmentally friendly, socially beneficial and economically competitive. For proper and efficient evaluation of wind energy, a hybrid Seasonal Auto-Regression Integrated Moving Average and Least Square Support Vector Machine (SARIMA–LSSVM) model is significantly developed to predict the mean monthly wind speed in Hexi Corridor. The design concept of combining the Seasonal Auto-Regression Integrated Moving Average (SARIMA) method with the Least Square Support Vector Machine (LSSVM) algorithm shows more powerful forecasting capacity for monthly wind speed prediction at wind parks, when compared with the single Auto-Regression Integrated Moving Average (ARIMA), SARIMA, LSSVM models and the hybrid Auto-Regression Integrated Moving Average and Support Vector Machine (ARIMA–SVM) model. To verify the developed approach, the monthly data from January 2001 to December 2006 in Mazong Mountain and Jiuquan are used for model construction and model testing. The simulation and hypothesis test results show that the developed method is simple and quite efficient.

This paper presents a neural network based on adaptive resonance theory, named distributed ART (adaptive resonance theory) & HS-ARTMAP (Hyper-spherical ARTMAP network), applied to the electric load forecasting problem. The distributed ART combines the stable fast learning capabilities of winner-take-all ART systems with the noise tolerance and code compression capabilities of multi-layer perceptions. The HS-ARTMAP, a hybrid of an RBF (Radial Basis Function)-network-like module which uses hyper-sphere basis function substitute the Gaussian basis function and an ART-like module, performs incremental learning capabilities in function approximation problem. The HS-ARTMAP only receives the compressed distributed coding processed by distributed ART to deal with the proliferation problem which ARTMAP (adaptive resonance theory map) architecture often encounters and still performs well in electric load forecasting. To demonstrate the performance of the methodology, data from New South Wales and Victoria in Australia are illustrated. Results show that the developed method is much better than the traditional BP and single HS-ARTMAP neural network.