二阶加纯滞后模型的闭环在线辨识

Identification of Second-order Plus Dead-time Model from Closed Loop On-line Test Data

本文提出一种在线测取N组闭环系统的数据 ,得出二阶加纯滞后模型参数的辨识方法 ,该方法辨识准确 ,辨识时间短且对正常运行的系统影响较小。大量的仿真实验表明该方法辨识精度高 ,对采样点数不敏感 ,受干扰的影响很小。该辨识方法可以在过程和仪器的自适应控制中采用。

In this paper, an identification method for second-order plus dead-time model is proposed, in which SOPDT model parameters can be estimated from N groups of closed-loop on-line tested data. This method has many advantages, i.e. precise result, shorting the time of identification and decreasing disturbance effect on a normal operational system. A great deal of simulation experiment has been done to show that the estimation error is very small, the method is less sensitive to the number of sampled data and hardly affected by the disturbance. The identification method can be used in adaptive control of process and industrial instrumentation.