主蒸汽压力的无模型自适应预测控制

针对火电机组主蒸汽压力被控对象的大惯性、大迟延等特点,采用伪梯度向量的概念,对被控对象进行动态化处理,得到预测模型.提出了1种新型的无模型自适应预测控制(MFAPC)算法,将MFAPC算法用于主蒸汽压力串级控制系统的主调节器中,副调节器采用常规PI控制器.仿真结果表明,MFAPC-PI串级控制系统较常规PID-PI串级控制系统响应速度快、抗干扰性能强,对系统参数时变具有较好的鲁棒性.

MODEL FREE ADAPTIVE PREDICTIVE CONTROL FOR MAIN STEAM PRESSURE SYSTEM

According to the characteristic of main steam controlled object such as large inertia and delay,an improved model free adaptive predictive control(MFAPC) scheme is designed by using the pseudo-partial-derivative dynamically to linearize a controlled system and build a predictive model.The new control method is applied to the main steam pressure cascade control system which is a typical controlled process with large time constant and large time-delay parameters in a thermal power plant.The main control loop is implemented by the MFAPC algorithm,and the inner control loop by a conventional PI controller.Simulation results convince that,compared to the traditional PID-PI cascade control system,the MFAPC-PI system has higher response speed,stronger anti-jamming performance,and better robustness for time-variable of system parameters.