基于PID神经网络的变桨控制系统的辨识

The identification of the wind pitch control system with the PID neural network

PID神经网络是一种多层前向反馈网络,它除了具备传统的多层前向网络的特点,如逼近能力、并行计算、非线性变换等特性外,其隐含层神经元还分别具有比例、微分和积分等动态特性。本文将介绍PID神经网络对风机变桨控制系统辨识应用。

PID neural network is a kind of the multilayer forward and back network. Besides multilayer forward networks traditional merit, such as approach the ability proceed together the calculation nonlinear transformation, its middle layer has the proportional,integral, derivative dynamic characteristic. This article presents the system identification application with PIDNN, which uses for pitch control in the wind turbine.