Research On Of Performance Assessment Method Of Non-Gaussian Control System

At present,most researches are based on the assumption that interference obeys Gaussian distribution in the field of performance assessment.In actual industrial production processes,non-Gaussian phenomena also exist in large numbers.With the deepening of industrialization and informatization,the Gaussian assumption of disturbance in the past has not been able to meet the demand.Therefore,it is necessary to develop a performance assessment method suitable for non-Gaussian systems.Based on the study of traditional Gaussian system performance assessment methods,this paper extends the performance evaluation methods of non-Gaussian control systems.As follows the main work of this paper is:1.The minimum variance benchmark is extended,and the minimum entropy benchmark is used to assess the performance of the non-Gaussian cascade control system.In order to verify the advantages of the minimum entropy benchmark,it is compared with the minimum variance benchmark.2.Based on the generalized minimum variance benchmark and the minimum entropy benchmark,combined with the characteristics of the two,the generalized minimum entropy benchmark is proposed to evaluate the performance of the non-Gaussian single-loop system,analyze the comparison results,and observe whether the advantages of the two benchmarks can be combined.3.Research on the subspace identification algorithm and improve it to make the subspace identification algorithm suitable for identifying nonGaussian systems,which lays the foundation for the performance assessment of non-Gaussian state-space models with improved LQG benchmarks.The simulation results verify that the proposed improved method has a certain degree of reliability and accuracy in the performance assessment of nonGaussian systems,and improved the subspace identification algorithm,which provides an identification tool for the subsequent work of improving the LQG benchmark.