基于神经网络补偿的主汽温多模型控制

Multi-model control system of superheated steam temperature based on neural network compensation

针对过热汽温对象的特性参数随机组运行工况的变化而有较大变化的特点,设计了基于神经网络补偿的多模型切换的串级控制系统.基于不同工况点设计的固定控制器可根据运行工况进行切换,补偿控制器在此基础上补偿控制作用的不足.仿真结果表明,这种控制方案实现了系统全工况运行的满意控制.

The dynamic behavior parameters of superheated steam temperature vary widely indifferent conditions. In this paper, a new cascade control system based on neural network compensation and multi-model switch control, static controller associated with each model can be switched according to different load scopes. And neural network controller gives.compensating control effect. The simulation results indicate that the posed control system is efficient over the entire operating conditions.