燃煤电厂协调控制系统故障诊断与容错控制应用研究

RESEARCH ON FAULT DIAGNOSIS AND FAULT TOLERANT CONTROL OF THE COORDINATE CONTROL SYSTEM IN COAL-FIRED POWER PLANT

基于多模型自适应滤波技术和线性二次型最优控制提出了电厂锅炉汽机协调控制系统的故障诊断和容错控制方案。建立了包含协调控制系统正常和可能发生的各种故障模式的系统模型集,并根据系统模型集以及系统实时输入输出数据进行多模型自适应滤波,从而实现故障诊断;然后根据故障诊断结果以及针对每个模型设计的线性二次型最优基本控制器,实现控制律的在线重组和主动容错控制。最后将上述故障诊断和容错控制方法应用于某300MW燃煤电厂协调控制系统中,仿真结果表明算法的有效性。

Based on multiple-model adaptive filter technology and linear quadratic gaussian (LQG) control strategy, a scheme of fault diagnosis and fault tolerant control for the boiler-turbine coordinate control systems is presented. At first, the normal model and the fault models of the control system are built. Consequently, with the aid of these models and the real-time input output data in the system is studied, the fault diagnosis is realized by using the multiple-model adaptive filter. Following the result of the fault diagnosis and the basic LQG controllers for each model, an on-line reconfiguration of the control law is implemented to obtain the active fault tolerant control. Finally, the fault diagnosis and fault tolerant control scheme is put in use for a 300MW coal-fired boiler turbine unit and the effectiveness is shown by a set of simulation results.