基于RBFPID的火电机组协调主控系统

COORDINATED MAIN CONTROL SYSTEM OF THERMAL POWER UNIT BASED ON RBFPID

基于Matlab软件设计了RBF神经网络辨识PID(RBFPID)控制策略,并采用OPC通讯方式实现了Matlab与火电机组仿真支撑软件APROS的数据交换,从而可利用已开发的机组仿真培训系统使Matlab中设计的控制算法直接参与过程控制,以验证RBFPID控制策略的实用性。验证结果表明,RBFPID控制策略的控制效果明显优于常规PID加前馈控制策略,且调试方便,具有良好的鲁棒性,能够较好地适应机组运行工况的变化。

Based on MATLAB software, a control strategy of RBF neural network identifying PID （RBFPID） has been designed,and adopting OPC communication mode to realize the data exchange be- tween MATLAB and the emulation supporting software APROS of thermal power units, thereby, the developed emulation and training system of units can be used to make the control algorithms designed in MATLAB to directly participate into process control, so as to verify the practicality of control strat- egy RBFPID. Results of verification show that the control effectiveness of control strategy RBFPID is substantially superior to control strategy of conventional PID plus feed forward,it is convenient to de- bug, boasting good robustness,and can be more suitable to operation condition variation of unit.