基于辅助模型的量化控制系统辨识方法

Auxiliary model-based identification method for quantized control systems

针对具有通信约束的量化控制系统模型,在采用随机重复性试验测量信息的技术上,提出了基于辅助模型的量化系统参数辨识方法.首先分析了在随机重复性试验方法下量化系统的模型特征并给出了分两步辨识的策略.分析表明,在上述模型里系统具有时变的估计误差,推导了进行参数辨识所满足的持续激励条件,并给出了基于辅助模型的多新息量化辨识递推算法.接着研究了所给出辨识算法的收敛性分析,得到了系统参数估计误差上界的计算式,最后将方法推广到一类Hammerstein非线性系统量化辨识问题上.数字仿真验证了该算法及结论的有效性.

An auxiliary model-based identification method for quantized systems subjected to communication constraints is introduced, based on the technique of repetitive stochastic empirical measurements. The model characteristics of the quantized system are analyzed, and a two-step identification strategy is presented. It is shown that the quantized system based on repetitive stochastic empirical measurements involves time-varying estimation error. The persistent exciting condition for parameter identification is derived. The auxiliary model-based quantized multi-innovation recursive algorithm for quantized systems is also given. Convergence analysis of the auxiliary model-based algorithm provides the method for computing the upper bound of parameter identification error. It is demonstrated that under certain conditions, the recursive algorithm is consistently convergent. Finally, this identification method is extended to a class of Hammerstein nonlinear quantized systems. Simulation results show the effectiveness of the conclusions.