一种闭环对象辨识方法及其在生产中的应用

Closed Loop Object Identification Method and Its Application

针对一些难达稳态过程的模型不易辨识的问题,提出了一种对象模型参数和状态方程初值同时辨识的新方法。该方法可不进行装置实验,仅使用装置历史数据就可进行在线辨识,且适用于系统处于非稳定状态的阶跃响应数据。闭环控制系统辨识仿真和实际应用表明,应用该方法辨识出的对象特性进行PID参数设计,效果明显,并且使控制系统具有很好的抗干扰能力和鲁棒性。

To deal with the design of control loop parameter problem on large disturbance processes and large dead-time processes of practical industrial control, a novel identification method--MPSEIVI（Model parameter and state equation initial values identification simultaneous） is proposed. Process model parameters and initial states can be identified at the same time by using the method. Simulation and practical application shows that the proposed method is advantageous to the anti-interfere and the robustness. And it gives a new way for the optimization of the process system.