Search

Wind power forecasting can effectively improve the energy utilization efficiency of a power system and ensure its stable operation. In this study, a novel hybrid multistep prediction model, including the complete ensemble empirical mode decomposition with adaptive noise (CEEMDAN), variational mode decomposition (VMD), the kernel principal component analysis (KPCA), an enhanced arithmetic optimization algorithm (ENAOA), a bidirectional long short-term memory (BILSTM) neural network, and error correction, was designed for short-term wind power forecasting. First, the collected original wind power data were decomposed into multiple intrinsic mode functions (IMFs) through a secondary decomposition composed of the CEEMDAN and VMD, which eliminated the interactions between different components to achieve denoising. Second, the KPCA was adopted to reduce the dimensionality of the multiple IMFs by extracting the principal components, effectively reducing the complexity of the multidimensional IMF data and improving the forecasting efficiency of the proposed prediction model. Subsequently, an ENAOA was proposed based on the Sobol sequence, adaptive T-distribution, and random walk strategy to optimize the BILSTM parameters. Finally, multiple preprocessed components were predicted by the optimized BILSTM, after which error correction was performed to obtain the final prediction results, which could further reduce the forecast error of the designed prediction model. Based on two sets of data collected from a wind farm in northwest China, the simulation results of 1-step, 4-step, 7-step, and 10-step forecasting revealed that compared with other incomplete models, the various algorithms adopted in the hybrid forecasting model reduced the prediction errors to different degrees, significantly enhanced the wind power prediction performance, and validated the effectiveness and feasibility of the proposed model. • Original wind power data is preprocessed through CEEMNAN and VMD. • KPCA is adopted to extract principal components of several IMFs. • Propose an improved ENAOA optimization algorithm through three improvements. • Error correction is employed to further increase the forecast accuracy. • The hybrid multi-step forecasting model greatly reduces the prediction error. [ABSTRACT FROM AUTHOR]

Accurate prognostic for the remaining useful life (RUL) of lithium-ion batteries (LIBs) is extremely crucial to the stable operation and timely maintenance of a battery system. Nevertheless, battery lifespan is difficult to measure due to the capacity regeneration in non-linear and unstable degradation trend. To increase the prediction accuracy, the Time Varying Filter-based Empirical Mode Decomposition (TVF-EMD) is innovatively introduced to decompose the original capacity data into subseries. Meanwhile, the complexities of the subseries are measured by the Box-counting dimension (BCD). Moreover, Fast Ensemble Empirical Mode Decomposition (FEEMD) is exploited to further decompose the most complex subseries. Additionally, Bidirectional Gated Recurrent Unit (BiGRU) is established for (sub-)subseries prognosis. The prediction performance is further strengthened by an error correction method (ECM). Eventually, the effectiveness of the proposed prognosis framework is verified on two battery datasets. The experimental results illustrate that the maximum root mean squared error (RMSE), mean absolute error (MAE) and mean absolute percentage error (MAPE) of the proposed hybrid framework are merely 1.917, 0.434 and 0.706% respectively. Compared with two decomposition methods, MAE can be reduced by at least 22.73%, and a reduction of not less than 7.4% in RMSE is achieved. • A new hybrid methodis proposed for the remaining useful life forecasting. • The secondary decomposition method employed can enhance the composition ability. • The box-counting dimension is utilized to choose the most complicated sub-layers. • An error correction scheme is proposed to further enhance prediction performance. [ABSTRACT FROM AUTHOR]

The effective utilization of wind energy is a positive development trend in renewable energy that provides an effective impetus to alleviate global warming. However, the complex volatility and extreme suddenness of wind speeds make it difficult to accurately predict wind speeds. To overcome this challenge, a novel short-term wind speed prediction model based on double decomposition, piecewise error correction, Elman neural network and the autoregressive integrated moving average model is proposed. In this model, the fuzzy entropy is applied to judge the complexity of each subseries generated by variational mode decomposition. To address error correction, a piecewise error correction method that needs to extract only a part of the error sequence as a training set is proposed. Three experiments utilizing 6 Datasets and 14 compared models are conducted to verify the effectiveness of the proposed model. The results show that (i) the hybrid strategy based on double decomposition and piecewise error correction can improve the final prediction accuracy of different single models; (ii) the proposed hybrid model can effectively handle both linear and nonlinear problems; (iii) the root mean squared error of the proposed model for Dataset1-6 in seven-step forecasting are 0.1243, 0.1444, 0.0233, 0.1015, 0.1846 and 0.2513, respectively. • The double decomposition strategy is applied to wind speed prediction. • A novel piecewise error correction method has been developed. • Fuzzy entropy is used to judge the complexity of subsequences. • The proposed model is verified using 6 wind speed datasets in multi-step forecasting. [ABSTRACT FROM AUTHOR]