Wind Power Time Series Research Articles

Wind Power Forecasting using A Heterogeneous Ensemble of Decomposition-based NNRW Techniques

Accurate and reliable wind power forecasting plays a vital role in the operation and management of power systems. Hence, it has become necessary to research and develop a high-accuracy wind power forecasting model. However, owing to highly nonlinear and non-stationary patterns of wind power time-series, creating a wind forecasting model capable of predicting such series accurately is both complicated and challenging. Aiming at this challenge, this paper introduces a new decomposition-based hybrid model based on multiple decomposition techniques, neural network with random weights (NNRW), and linear combiner. In our approach, the original time-series is decomposed into a collection of sub-series by different decomposition techniques. Each sub-series is modeled and predicted separately using NNRW. The predicted signals of each decomposition model are then reconstructed independently. Finally, all of the reconstructed results are integrated by the combiner using a linear combination method. The predictive performance of the proposed method was compared with other state-of-the-art techniques in over 12 wind power time-series. The experimental results show that the predictive performance of the proposed method remarkably outperforms the other competitors, proving the developed model to be effective, efficient, and practicable.

Less Information, Similar Performance: Comparing Machine Learning-Based Time Series of Wind Power Generation to Renewables.ninja

Driven by climatic processes, wind power generation is inherently variable. Long-term simulated wind power time series are therefore an essential component for understanding the temporal availability of wind power and its integration into future renewable energy systems. In the recent past, mainly power curve-based models such as Renewables.ninja (RN) have been used for deriving synthetic time series for wind power generation, despite their need for accurate location information and bias correction, as well as their insufficient replication of extreme events and short-term power ramps. In this paper, we assessed how time series generated by machine learning models (MLMs) compare to RN in terms of their ability to replicate the characteristics of observed nationally aggregated wind power generation for Germany. Hence, we applied neural networks to one wind speed input dataset derived from MERRA2 reanalysis with no location information and two with additional location information. The resulting time series and RN time series were compared with actual generation. All MLM time series feature an equal or even better time series quality than RN, depending on the characteristics considered. We conclude that MLM models show a similar performance to RN, even when information on turbine locations and turbine types is unavailable.

Prediction of offshore wind farm power using a novel two-stage model combining kernel-based nonlinear extension of the Arps decline model with a multi-objective grey wolf optimizer

The development of the wind power market has led various countries to begin shifting the construction of wind farms from land to offshore. Accurately predicting the short-term wind power of an offshore wind farm is significant for preventive control and scheduling. This paper proposes a novel two-stage hybrid model to predict short-term wind power. First, complete ensemble empirical mode decomposition with adaptive noise is used to preprocess the raw data and make it smoother. Second, an improved grey-box model is utilized to predict the wind power, where a multi-objective grey wolf optimizer is used to optimize the kernel-based nonlinear extension of the Arps decline model to ensure both prediction stability and accuracy. The historical wind power data for offshore wind farms in Belgium are taken as a case study, and the results indicate that the proposed model has higher prediction accuracy and stability than those of six benchmark models. Three critical issues related to this work are also considered. The following conclusions are reached: (1) Utilizing a denoising method and multi-objective optimizer effectively improves the prediction accuracy and stability of the original model. Specifically, the multi-objective optimizer improves the prediction performance; (2) The proposed model performs best at predicting wind power in autumn; (3) Because of the great variation in dimensions for the wind power time series, the mean absolute percentage error and root mean squared percentage error are not suitable as error evaluation indicators.

The Hybridization of Ensemble Empirical Mode Decomposition with Forecasting Models: Application of Short-Term Wind Speed and Power Modeling

In this research, two hybrid intelligent models are proposed for prediction accuracy enhancement for wind speed and power modeling. The established models are based on the hybridisation of Ensemble Empirical Mode Decomposition (EEMD) with a Pattern Sequence-based Forecasting (PSF) model and the integration of EEMD-PSF with Autoregressive Integrated Moving Average (ARIMA) model. In both models (i.e., EEMD-PSF and EEMD-PSF-ARIMA), the EEMD method is used to decompose the time-series into a set of sub-series and the forecasting of each sub-series is initiated by respective prediction models. In the EEMD-PSF model, all sub-series are predicted using the PSF model, whereas in the EEMD-PSF-ARIMA model, the sub-series with high and low frequencies are predicted using PSF and ARIMA, respectively. The selection of the PSF or ARIMA models for the prediction process is dependent on the time-series characteristics of the decomposed series obtained with the EEMD method. The proposed models are examined for predicting wind speed and wind power time-series at Maharashtra state, India. In case of short-term wind power time-series prediction, both proposed methods have shown at least 18.03 and 14.78 percentage improvement in forecast accuracy in terms of root mean square error (RMSE) as compared to contemporary methods considered in this study for direct and iterated strategies, respectively. Similarly, for wind speed data, those improvement observed to be 20.00 and 23.80 percentages, respectively. These attained prediction results evidenced the potential of the proposed models for the wind speed and wind power forecasting. The current proposed methodology is transformed into R package ‘decomposedPSF’ which is discussed in the Appendix.

Asymmetric GARCH type models for asymmetric volatility characteristics analysis and wind power forecasting

Wind power forecasting is of great significance to the safety, reliability and stability of power grid. In this study, the GARCH type models are employed to explore the asymmetric features of wind power time series and improved forecasting precision. Benchmark Symmetric Curve (BSC) and Asymmetric Curve Index (ACI) are proposed as new asymmetric volatility analytical tool, and several generalized applications are presented. In the case study, the utility of the GARCH-type models in depicting time-varying volatility of wind power time series is demonstrated with the asymmetry effect, verified by the asymmetric parameter estimation. With benefit of the enhanced News Impact Curve (NIC) analysis, the responses in volatility to the magnitude and the sign of shocks are emphasized. The results are all confirmed to be consistent despite varied model specifications. The case study verifies that the models considering the asymmetric effect of volatility benefit the wind power forecasting performance.

Wind Power Short-Term Forecasting Hybrid Model Based on CEEMD-SE Method

Accurately predicting wind power is crucial for the large-scale grid-connected of wind power and the increase of wind power absorption proportion. To improve the forecasting accuracy of wind power, a hybrid forecasting model using data preprocessing strategy and improved extreme learning machine with kernel (KELM) is proposed, which mainly includes the following stages. Firstly, the Pearson correlation coefficient is calculated to determine the correlation degree between multiple factors of wind power to reduce data redundancy. Then, the complementary ensemble empirical mode decomposition (CEEMD) method is adopted to decompose the wind power time series to decrease the non-stationarity, the sample entropy (SE) theory is used to classify and reconstruct the subsequences to reduce the complexity of computation. Finally, the KELM optimized by harmony search (HS) algorithm is utilized to forecast each subsequence, and after integration processing, the forecasting results are obtained. The CEEMD-SE-HS-KELM forecasting model constructed in this paper is used in the short-term wind power forecasting of a Chinese wind farm, and the RMSE and MAE are as 2.16 and 0.39 respectively, which is better than EMD-SE-HS-KELM, HS-KELM, KELM and extreme learning machine (ELM) model. According to the experimental results, the hybrid method has higher forecasting accuracy for short-term wind power forecasting.

Echo state network based ensemble approach for wind power forecasting

Accurate and reliable wind power forecasting is of great significance for the economic and safe operation of electric power and energy systems. However, due to the negative factors such as inaccurate numerical weather prediction and frequent ramping events, previous wind power prediction methods are difficult to meet actual needs for practical applications. To this end, this paper originally proposes a new wind power prediction method with high accuracy based on echo state network. This method is a hybrid of wavelet transform, echo state network and ensemble technique. Wavelet transform is used to decompose raw wind power time series data into different frequencies with better outliers and behaviors. Then, we adopt the echo state network to automatically learn the nonlinear mapping relationship between input and output in each frequency. Later, ensemble technique is applied to deal with the model misspecification and data noise problems, which are common in wind power prediction, thereby reducing the uncertainty of wind power forecasting and improving prediction accuracy. Finally, we use wind power data from real wind farms in Belgium and China to verify the feasibility and effectiveness of the proposed method. The simulation results show that the proposed method is superior to other benchmark algorithms in prediction accuracy, which indicates that the method has high potentials for practical application in real electric power and energy systems.

A dynamic wavelet-based robust wind power smoothing approach using hybrid energy storage system

In this paper, a new robust dynamic-wavelet-enabled approach is proposed for wind power smoothing by using the hybrid energy storage system (HESS) consisting of batteries and super-capacitors. The developed approach is able to decompose wind power time series into self-adaptively optimized wavelet parameters without violating the physical constraints. The latter includes those for power injection regulation, state-of-charge (SOC) of HESS, and allowable charge/discharge depth. By doing that, batteries and super-capacitors can be coordinated in an optimal manner, yielding high efficiency. To address wind power uncertainty, a box-type uncertainty set that describes the probability of wind power prediction error is developed. The uncertainty set is further leveraged by the robust model predictive control (MPC) strategy and the robust coefficient to assess the trade-off between robustness and economic benefits. The advantages of the method are validated through the realistic wind farm data and the comparisons with other approaches. The results indicate that the proposed method can be implemented online to determine the robust smoothing strategy for HESS, yielding the highly-qualified integration of renewable energy into the power grid.

Markov chain‐based wind power time series modelling method considering the influence of the state duration on the state transition probability

Due to the inherent uncertainties of wind power, its large-scale integration strongly impacts the planning and operation of power systems. To investigate these impacts, a stochastic model is required to more accurately capture the wind power's characteristics. This study proposes an improved Markov chain (MC)-based time series (TS) modelling method for the stochastic generation of synthetic wind power TS. First, a self-adaptive state division strategy is proposed to objectively classify historical data into several typical states. This strategy combines a state optimisation clustering model with a random-variable-modelling-oriented filter parameter optimisation method. Then, a three-dimensional state transition probability matrix (STPM) is proposed and constructed to generate synthetic wind power state TS. In contrast to the previous STPMs, the proposed STPM can capture the changing pattern of the transition probability against the state duration. Finally, the fluctuation quantity and noise are separately and sequentially added to the generated state TS, as an improvement over previous fluctuation characteristic addition methods, to obtain the final synthetic wind power TS. The results show that the proposed method outperforms previous MC-based TS modelling methods in reproducing historical characteristics, such as the transition and fluctuation characteristics, and does not increase the STPM construction algorithm's time complexity.

Chaotic wind power time series prediction via switching data-driven modes

To schedule wind power efficiently and to mitigate the adverse effects caused by wind's intermittency and variability, an advanced wind power prediction model is proposed in this paper. This model is a combined model via switching different data-driven chaotic time series models. First, inputs of this model come from the reconstructed data based on the chaotic characteristics of wind power time series. Second, three different data mining algorithms are used to construct wind power prediction models individually. To obtain a regime for switching optimal models, a Markov chain is trained. Then, weights of different data-driven modes are calculated by the Markov chain switching regime, and used in the final combined model for wind power prediction. The industrial data from actual wind farms is studied. Results of the proposed model are compared with that of non-reconstructed input data, traditional data-driven models and two typical combined models. These results validate the superiority of proposed model on improving wind power prediction accuracy.

Wind and solar power probability density prediction via fuzzy information granulation and support vector quantile regression

Uncertainty among wind and solar power affects the stability of power systems. In order to fully describe the uncertainty of wind and solar power, a probability density prediction model is proposed to predict the probability density function of wind and solar power. According to the wind and solar power time series, the original data is processed with fuzzy information granularity to eliminate the fluctuation and uncertainty of data. The Lagrange function is constructed by a support vector quantile regression model to get the quantile of wind and solar power at different points. The conditional quantiles are combined with the Epanechnikov kernel function to acquire complete probability density curves of forecasting results. In order to evaluate the performance of the output results, this paper analyzes the accuracy of the prediction results using point prediction error, prediction interval coverage probability and average bandwidth. The experimental data of wind and solar power with same temporal and spatial resolutions are taken into account. The results show that the method can effectively describe the uncertainty of wind and solar power, and also provide technical support for the safe and stable operation of the power system.

Estimation of the Daily Variability of Aggregate Wind Power Generation in Alberta, Canada

This paper describes a hierarchy of increasingly complex statistical models for wind power generation in Alberta applied to wind power production data that are publicly available. The models are based on combining spatial and temporal correlations. We apply the method of Gaussian random fields to analyze the wind power time series of the 19 existing wind farms in Alberta. Following the work of Gneiting et al., three space-time models are used: Stationary, Separability, and Full Symmetry. We build several spatio-temporal covariance function estimates with increasing complexity: separable, non-separable and symmetric, and non-separable and non-symmetric. We compare the performance of the models using kriging predictions and prediction intervals for both the existing wind farms and a new farm in Alberta. It is shown that the spatial correlation in the models captures the predominantly westerly prevailing wind direction. We use the selected model to forecast the mean and the standard deviation of the future aggregate wind power generation of Alberta and investigate new wind farm siting on the basis of reducing aggregate variability.

Capacity credits of wind and solar generation: The Spanish case

This paper analyses the capacity credits (CCs) of renewable photovoltaic (PV), concentrated solar power (CSP) and wind technologies in the Spanish power system. This system has steadily increased the share of renewables, reaching a penetration level of over 30%. The predictions made by ENTSO-e suggest that this level will increase to 50% by 2030. Therefore, different scenarios are studied in this paper to investigate the evolution of renewable integration and assess the corresponding contributions to reliability. The assessment is performed using a sequential Monte Carlo (SMC) method considering the seasonality of renewable generation and the uncertainties related to renewable sources, failure issues and the maintenance of thermal-based units. The baseline for SMC is provided by historical annual time series of irradiance and wind power data from the Spanish system. In the solar case, these time series are transformed into power time series with models of CSP and PV generation. The former includes different thermal storage strategies. For wind generation, a moving block bootstrap (MBB) technique is used to generate new wind power time series. The CC is assessed based on the equivalent firm capacity (EFC) using standard reliability metrics, namely, the loss of load expectation (LOLE). The results highlight the low contribution of renewables to power system adequacy when the Spanish power system has a high share of renewable generation. In addition, the results are compared with those of similar studies.

Quantifying benefits of wind power diversity in New Zealand

Wind integration studies often focus on the capacity value of wind power without considering Unit Commitment and Economic Dispatch or resolving requirements for ancillary services. Here, a novel method for simulating wind power time series with sufficient temporal span to support capacity studies and temporal resolution to support UCED studies is developed. Wind speed time series (WSTS), with 6 h temporal and 0.7 × 0.7 degree spatial resolutions, are extracted from the ECMWF-interim reanalysis, interpolated, scaled, and imputed so that they are representative of a point wind speed measurement with a 5 min temporal resolution. Imputation is made using a wavelet multi-resolution analysis approach that ensures temporally consistent correlations while accounting for heteroskedasticity. WSTS are transformed to power using wind power plant power curves, low-pass filters, and a Markov Chain model of operational efficiencies. The wind power model is validated using a set of measurements made at wind power plants (WPPs) in New Zealand and used to simulate power time series for 2 GW portfolios of WPPs representing compact, disperse, diverse, and business-as-usual portfolios. Metrics for dependability, variability, and predictability are applied to quantify the benefits of spatial diversification.

Wind Power Short-Term Prediction Based on LSTM and Discrete Wavelet Transform

A wind power short-term forecasting method based on discrete wavelet transform and long short-term memory networks (DWT_LSTM) is proposed. The LSTM network is designed to effectively exhibit the dynamic behavior of the wind power time series. The discrete wavelet transform is introduced to decompose the non-stationary wind power time series into several components which have more stationarity and are easier to predict. Each component is dug by an independent LSTM. The forecasting results of the wind power are obtained by synthesizing the prediction values of all components. The prediction accuracy has been improved by the proposed method, which is validated by the MAE (mean absolute error), MAPE (mean absolute percentage error), and RMSE (root mean square error) of experimental results of three wind farms as the benchmarks. Wind power forecasting based on the proposed method provides an alternative way to improve the security and stability of the electric power network with the high penetration of wind power.

An Improved ELM Model Based on CEEMD-LZC and Manifold Learning for Short-Term Wind Power Prediction

The nonlinear characteristics of wind power series and random fluctuation characteristics of wind resources have a harmful effect on stability of wind power prediction. This paper proposes a novel hybrid wind power short-term prediction model for improving precision and stability of wind power prediction. Firstly, the non-stationary wind power time series is decomposed by complete ensemble empirical mode decomposition - Lempel-Ziv complexity (CEEMD-LZC). Secondly, local linear embedding (LLE) is used to reduce the dimension of meteorological data with maintaining the essential structure of meteorological data. The integrated meteorological data predigested by LLE are treated as input of the prediction model. Finally, ant lion optimization (ALO) is utilized to optimize the weights and biases of extreme learning machine (ELM) network nodes to improve ELM prediction performance. The results of case studies show that the proposed model has a good forecasting effect and is more advantageous in 48h-ahead wind power forecasting.

A Fuzzy Adaptive Probabilistic Wind Power Prediction Framework Using Diffusion Kernel Density Estimators

The inherent uncertainty in predicting wind power generation makes the operation and control of power systems very challenging. Probabilistic measurement of wind power uncertainty in the form of a reliable and sharp interval is of utmost importance, but construction of such high-quality prediction intervals (PIs) is difficult because wind power time series are nonstationary. In this paper, a framework based on the concept of bandwidth selection for a new and flexible kernel density estimator is proposed. Unlike previous related works, the proposed framework uses neither a cost function-based optimization problem nor point prediction results; rather, a diffusion-based kernel density estimator (DiE) is utilized to achieve high-quality PIs for nonstationary wind power time series. Moreover, to adaptively capture the uncertainties of both the prediction model and wind power time series in different seasons, the DiE is equipped with a fuzzy inference system and a tri-level adaptation function. The proposed framework is also founded based on a parallel computing procedure to promote the computational efficiency for practical applications in power systems. Simulation results demonstrate the efficiency of the proposed framework compared to well-known conventional benchmarks using real wind power datasets from Canada and Spain.

Short-term Wind Power Prediction Method Based on Wavelet Packet Decomposition and Improved GRU

Accurately predicting wind power in wind farms is of great significance to ensure the safe and stable operation of the power system. A novel short-term wind power forecasting method (WPD-GRU-SELU) based on wavelet packet decomposition (WPD) and improved gated recurrent unit (GRU) is proposed. Firstly, this method uses WPD to decompose the time series of wind power into several sub-sequences with different frequencies. Then the sub-sequences of different frequency components are predicted by using the improved GRU neural network, which uses the scaled exponential linear units(SELU) as the activation function to squash the hidden states to calculate the output. Finally, the output datum of GRU neural networks are reconstructed to obtain the complete wind power predicting results. Experiments illustrate that the WPD-GRU-SELU model have a more accurate forecast to the short-term wind power prediction compared with other RNN models.

Short-term Wind Power Prediction Model Based on Encoder-Decoder LSTM

We propose a long short-term memory (LSTM) network based encoder-decoder (E-D) model for wind power prediction (WPP). The LSTM-based E-D model is constructed as an auto-encoder for mapping the wind power (WP) time-series into a fixed-length representation, state of the trained E-D LSTM. Then, the representation concatenated with weather forecasting information is used as a new input to another multiple LSTM network to make WPP. Real data collected from a wind farm with capacity of 50 MW of Shan Xi province were used to verify the conclusions. Results illustrate that the proposed method improves the model generalization ability and lowers misspecification risk by utilizing the WP time relationship through auto-encoding (AE) process. Combining extracted representation with weather forecasting information further improves the prediction accuracy.

Photovoltaic–wind joint power probability model based on multiple temporal and spatial scale

With an increasing number of renewable energy integrated into the power system, the impact of intermittent and random wind power and photovoltaic power on the power system is becoming larger and larger. To study the probabilistic correlation between wind power and photovoltaic power on different temporal and spatial scale, a joint probability model needs to be established. On the basis of the third-order Gaussian mixture model, a photovoltaic-wind joint power probability model is proposed in this study. By analysing the statistical probability characteristics of the measured power output data, this method avoids the error accumulation caused by the individual modelling of the wind and photovoltaic generators and considers the joint correlation between wind power and photovoltaic power. On the basis of this model, the corresponding time series of wind power and photovoltaic power is calculated. The results indicated that the coastal joint power probability model and the inland joint power probability model are quite different on the same time scale, the parameters of joint probability model on different time scales are close to each other. The effectiveness of the proposed method is verified by the measured data.