基于残差方差的系统阶次和参数辨识算法  

Simultaneous Recursive Identification Algorithm of System Order and Parameters Based on Residual Variance Method

作　　者：赵永历[1] 仲恒[1] 李定远[1] 

机构地区：[1]上海大学机电工程与自动化学院,上海200072

出　　处：《工业控制计算机》2014年第4期130-131,133,共3页Industrial Control Computer

摘　　要：系统辨识的研究一般是将系统的阶次辨识和参数估计分开的,但实际应用过程中这两个问题又是紧密相关的。有的模型阶次辨识过程是伴随着模型的参数估计,因此可以对这类阶次辨识方法同参数估计的方法进行融合和扩展。针对输出误差系统,借助辅助模型推导出基于残差方差递推算法,利用该算法辨识出了模型的阶次和参数,减少了传统系统辨识过程的计算量和辨识时间。The order identification and parameter estimation are usually separated in system identification,but the two issues are closely related in practical application.Some model order identification process is accompanied by the model parameter estimation,so these methods could be integrated and extended with parameter estimation methods.For the output error model, the residual variance recursive method is put forward with the aid of auxiliary-model.The system order and parameter estimation is got simultaneously with smaller computational complexity by this method.

关 键 词：残差方差 阶次辨识 参数估计 辅助模型 输出误差模型 

分 类 号：N945.14[自然科学总论—系统科学]