Wind Power Prediction of Kernel Extreme Learning Machine Based on Differential Evolution Algorithm and Cross Validation Algorithm

Ning Li,Wentao Ma,Fuxing He,Xiaoping Zhang,Ruotong Wang

doi:10.1109/access.2020.2985381

Abstract

As fossil fuel is being depleted, the percentage of wind power capacity in total electricity generation is increasing. In order to improve the absorption capacity of wind power, wind power prediction has been introduced. Aiming at the disadvantage of low prediction accuracy and unstable model of traditional extreme learning machine (ELM), a kernel extreme learning machine based on differential evolution (DE) and cross validation optimization method is proposed to predict short-term wind power generation. Firstly, the average mean square error (MSE) verified by k folding and cross validation is adopted as the error function of the model to improve the stability and generalization performance of the model. Secondly, differential evolution algorithm is used to optimize the regularization coefficient and kernel width of the kernel extreme learning machine with cross validation and improve the precision of model is 8.34%. Finally, compared with the application of extreme learning machine with genetic algorithm and cross validation to a wind farm prediction case in northwest China, the experimental results show that the convergence rate of this method is twice that of genetic algorithm (GA) optimization algorithm, and the accuracy is higher.

Highlights

At present, the basic trend of global energy transformation is to realize the transition from fossil energy system to lowcarbon energy system, and the ultimate goal is to enter the era of sustainable energy dominated by renewable energy
This paper focuses on developing a wind power prediction approach based on a novel Kernel extreme learning machine (KELM) algorithm
Due to its simplicity and flexibility, the traditional KELM algorithm has been optimized by many scholars, including data preprocessing, error function optimization, parameter optimization, network model improvement, etc

Summary

INTRODUCTION

The basic trend of global energy transformation is to realize the transition from fossil energy system to lowcarbon energy system, and the ultimate goal is to enter the era of sustainable energy dominated by renewable energy. There are many methods to optimize ELM algorithm, which have been widely used in various fields, such as wind power prediction It can be divided into different categories, such as data preprocessing, error function optimization and parameter optimization. Kernel extreme learning machine (KELM), as an error function optimization method, has been rarely studied in applications of wind power prediction. Literature [31] adopts the method of leaving one out cross validation to improve the stability and generalization performance of the model, which cannot be realized in the case of processing wind power big data. 3) the kernel width and regularization coefficient of the model optimized by DE algorithm are used for parameter optimization to improve the prediction accuracy of wind power, and the accuracy of the proposed model is verified by comparing and analyzing the GA optimization algorithm. The optimization of its parameters optimization and algorithm has become a research hotspot

STRUCTURE OF THE KELM MODEL

CVKELM MODEL BASED ON DIFFERENTIAL EVOLUTION ALGORITHM

COMPARISON OF KELM RESULTS WITH OTHER ALGORITHMS

CONCLUSION