两阶段闭环辨识算法在加热炉控制中的应用

Two-stage closed identification algorithm and its application

在闭环条件下,系统控制输入信号通过反馈环节与不可测噪声相关,故用开环辨识算法处理闭环数据时所得参数估计常常是有偏的。为克服参数估计的有偏性,采用了基于开环转换的两阶段闭环辨识算法,并分析了中间辨识模型与最终辨识结果的关系。在此基础上,得到了中间辨识模型的选择依据,即在相同运算量下,选用ARX模型比选用FIR模型所得辨识结果更好。并应用于一台加热炉,应用结果验证了该辨识算法的有效性及模型选择的合理性。

The result of the identification under closed-loop experimental conditions is bias when it gets processed by the open loop identification algorithm.To overcome the bias of estimated parameters,an indirect identification method is used.It is a two-stage closed-loop identification method based on an open loop transformation.An analysis of the relation between both transition model and identification result is made,thus yielding the choice criterion of the transition model.In the same operational quantity condition,the result is better by the ARX model than by the FIR model.An application example is given by using a heating furnance as an applied study.The application result shows the validity of the identification algorithm.