基于神经网络的风力发电控制系统

A Wind Power Control Systems Based on Neural Network Theory

提出一种用于风力发电控制系统的人工神经网络。首先将系统的控制问题转化为适合人工神经网络处理的识别问题,然后对样本数据进行适当的预处理,最后采用BPN神经网络结构,选择一种可降低网络灵敏度的新学习算法。仿真结果表明,所提出的含有两个隐层的神经网络模型具有较强的适应性。

An artificial neural network theory for the control of wind power systems is proposed. The control problem is converted to a suitable way for the neural network theory to recognize and deal with first. And then a pre-dealing to the sampled data is done. Last the BPN neural network structure is adopted, and a new study method which lowers the network's sensitivity is selected. The simulation result shows that the presented model of the control system has very high suitability.