对目前风电系统的数学建模作了较系统的阐述,对该领域已有的建模方法进行了比较分析,指出常用的机理建模法存在着精确性低和普遍性差的问题。引人神经网络建模的方法。采用BP网络和Elman网络基于现场数据对风电场进行了辨识建模,并分别在训练时间、记忆功能、精度和稳定性等4个方面做了比较。
The current mathematical modeling methods of wind power system are introduced and compared in this paper. Mechanism modeling method commonly has such problems as poor accuracy and low universality, while modeling based on artificial neural network can overcome these shortcomings. This paper adopts BP network and Elman network to identify and model based on abundant actual data of wind farm and compares respectively in four aspects - training time, memory, accuracy and stability.