基于GA-Elman神经网络的短期风电功率预测

Short Term Wind Power Prediction Based on GA-Elman Neural Network

风电功率预测研究已经成为风能安全保障中十分重要的一部分,然而,由于风速信号是非平稳的,所以要实现高精度的预测是很困难的。针对风电功率短期预测,采用遗传算法对神经网络的阈值、权值进行优化,提出将遗传算法优化的Elman神经网络混合预测模型应用于酒泉风电基地某风电场短期风电功率预测。以酒泉风电基地某风电场风速、温度和风电功率的历史数据为样本,通过仿真验证表明,该混合预测模型与传统的BP、Elman神经网络模型相比,在进行短期风电功率预测时精度更高,在实际风电功率预测中更具备实用性。

Wind power prediction has become a very important part of wind energy security. However, it is very difficult to achieve high accuracy prediction due to the non-stationary wind speed signal. For short-term wind power prediction, the genetic algorithm is used to optimize the threshold and weight of neural network, and the hybrid forecasting model of Elman neural network based on genetic algorithm is proposed to be applied to the short-term wind power prediction of a wind farm in Jiuquan wind power base. Based on the historical data of the wind speed, temperature, wind power in the Jiuquan wind power base, the simulation is made. The results show that the hybrid prediction model has a higher precision and more practicability in the short-telzn wind power prediction compared with the traditional BP and Ehnan neural network model.