基于神经网络的火电机组负荷预见控制方法及其仿真研究

Load preview control method of a thermal power unit based on neural network an its simulating study

利用多个自适应神经元构造了火电单元机组负荷预见控制系统 ,研究了用神经网络实现多变量系统的预见协调控制 ,神经网络输入量应该怎样选取。仿真结果证明了这种控制方案的正确性。与原有的负荷最优预见控制方法相比 ,该方法不依赖于对象的精确模型 ,也不涉及权重矩阵的选取。

This paper research the way of using adaline neural neurons construct power unit load preview control system and put forward how to choose the input of neural network in order to realize preview & harmony control of multi-variable system by neural network.Compared to original method of load optimal preview control,this project don't depend on object's precise model and don't involve the selection of weighting matrix.