Research On Assessment Method Of Control Loop Performance Based On Distribution Estimation

Control loop performance assessment(CPA)plays an important role in system operation.Through the performance assessment,the specific operating conditions of each control loop can be understood and the analysis of the control loop can be realized.Then improve the poor performance of the control loop in time to eliminate potential problems.For non-Gaussian linear control systems with random noise disturbance,the traditional CPA methods for Gaussian system will no longer be applicable.This article reviews the development history and research status of CPA as a starting point,summarizes various assessment scenarios and analyzes the advantages and disadvantages of each method.On the basis of this,it was found that the shape of the probability density function(PDF)determines the criteria for assessing the non-Gaussian system's control loop performance.Therefore,this paper starts with the PDF estimation of non-Gaussian systems and uses the CPA assessment of non-Gaussian control systems as its application background.Under this framework,this article mainly completes the following three aspects of work:1.The paper analyzes the stochastic system with non-Gaussian noise disturbance,points out the limitation of the MV-based CPA method and the irregularity of the calculation of the minimum entropy benchmark.It shows that the estimation of probability density function based on the minimum entropy is an indispensable step.2.By analyzing the existing problems and defects of the traditional estimation of distribution algorithm(EDA)in the identification process,the corresponding improvement measures are put forward.Simulation comparison experiments show that the improved algorithm is feasible and has high identification accuracy.3.On the basis of 2,this paper applies it to the performance assessment of linear non-Gaussian system control loop,and proposes a new type of performance assessment method based on distribution function:a CPA method based on minimum rational entropy control(MREC).Simulations verify the effectiveness of the proposed algorithm.