基于闭环SOPIM的LQG性能评价研究

Research on Performance Assessment with LQG Benchmark Based on Closed-loop SOPIM

控制系统性能评价对过程工业具有重要意义。在实际应用中,传统线性二次高斯(Linear Quadratic Gaussian,LQG)性能评价要求已知过程参数模型,但是系统模型的精确辨识计算量大,且可靠性低。针对上述问题,本文对基于闭环子空间正交投影辨识方法(Subspace Orthogonal Projection Identification Method,SOPIM)的LQG性能评价进行研究,并通过仿真实验证实该方法的有效性。

It is very important for process industry to evaluate the performance of control system. In practice, the traditional performance assessment with linear quadratic Gaussian （LQG） demands a known process parametric model, but it is very difficult to identify the process model accurately. Therefore, this paper will study the performance assessment with LQG benchmark based on closed-loop subspace orthogonal projection identification method （SOPIM）. The utility of the method in performance assessment is demonstrated through a simulated example.