Approach for short-term wind power prediction via kernel principal component analysis and echo state network optimized by improved particle swarm optimization algorithm

In recent years, short-term wind power forecasting has proved to be an effective technology, which can promote the development of industrial informatization and play an important role in solving the control and utilization problems of renewable energy system. However, the application of short-term wind power prediction needs to deal with a large number of data to avoid the instability of forecasting, which is facing more and more difficulties. In order to solve this problem, this paper proposes a novel prediction approach based on kernel principal component analysis and echo state network optimized by improved particle swarm optimization algorithm. Short-term wind power generation is affected by many factors. The original multi-dimensional input variables are pre-processed by kernel principal component analysis to determine the principal components that affect wind power. The dimension of principal component is less than the original input data, which reduces the complexity of modeling. The convergence and stability of the echo state network can be improved by using the principal component of the input variable. The advantage is to reduce the input variables, eliminate the correlation between the input variables, and improve the prediction performance of the prediction model. Furthermore, an improved particle swarm optimization algorithm is proposed to optimize the dynamic reservoir parameters of echo state network. Compared with other state-of-the-art prediction models, the case studies show that the proposed approach has good prediction performance for actual wind power data.