基于高阶统计量的非线性控制系统性能评估方法

针对一类基于反馈线性化策略的非线性控制系统,提出一种基于高阶矩估计非线性控制器性能检测新方法.非线性闭环控制系统由外部PID控制器、反馈线性化补偿环节和非线性控制对象组成,在模型和对象完全一致时,闭环系统输入输出具有线性特征.在实际应用中,经常遇到反馈线性化补偿环节模型参数与被控对象实际值不一致的情况,上述线性关系也一般不再存在.基于这种直观认识,以闭环控制系统所具有的线性关系为性能参考基准,通过估计系统输出的三元谱和四元谱等统计特征量,构造非高斯性以及非线性等假设推理命题,检验控制系统是否还处于原期望的操作点上,通过大量仿真计算表明了非线性检测新方法的有效性.

Nonlinear Control System Performance Assessment Method Based on the Higher Order Statistics

A controller performance detection method based on the higher order moments estimating nonlinearity is presented for a class of nonlinear systems using the feedback linearization strategy. The nonlinear closed-loop control system consists of an external PID controller,feedback linearization compensator and nonlinear controlled plant,when model/object is in full accord,with each other,the closed-loop system is featured with linearity. In practical applications,there is always an occasion that the feedback linearization compensator model parameters do not match with controlled plant practical values,then this linear relationship does not exist any more. Based on the intuitive understanding,this paper regards the closed-loop control system linear relation as the performance benchmark,by estimating the statistical characteristics of three and four element spectrum of the system outputs, uses the nonlinear and non-Gaussian hypothetical tests to judge whether the control system works in the original expected operating point. Several simulation calculations indicate that the proposed nonlinearity detection method is effective.