自适应变异粒子群优化BP的短期风电功率预测模型

针对反向传播(BP)神经网络预测模型在风电预测中预测精度低、输入变量多的问题,提出一种基于互信息的自适应变异粒子群优化BP的短期风电功率预测模型.首先,采用互信息筛选出原始数据中与输出功率相关度较大的影响因素,减少冗余信息;然后,引入具有自适应惯性权重系数和变异因子思想的粒子群算法对预测模型进行优化.结果表明:与传统预测模型相比,该预测模型具有收敛速度快、预测精度高等特点.

Short-Term Wind Power Prediction Model of Adaptive Mutation Particle Swarm Optimization BP

Aiming at the problem that back propagation(BP)neural network prediction model has low prediction accuracy and large input variables in wind power prediction, a short-term wind power prediction model based on mutual information of the adaptive mutation particle swarm optimization BP is proposed. Firstly, the mutual information is used to screen out the influencing factors of the original data with high correlation in the output power, and the redundant information is reduced. Then, the particle swarm algorithm based on the idea of adaptive inertia weight coefficient and variation factor is introduced to optimize the prediction model. The results show that compared with the traditional prediction model, the proposed model has the characteristics of fast convergence and high prediction accuracy.