Research On Performance Assessment Method Of Thermal Power Unit Control System Based On Hurst Exponent

The performance evaluation of industrial control systems is an important part of the National Industry 4.0 plan.How to establish a complete and effective automatic performance evaluation and real-time monitoring system has become a concern of people.In the actual industrial process,the long-term online operation of the equipment will inevitably lead to a decline in performance quality.Online diagnostics of the controller using routine operational data,at the same time,improving the effectiveness of each adjustment is the focus of field operators.In order to improve the control quality of the thermal power generation control system and for the dependence of the traditional performance assessment methods on the controlled object mathematical model,a performance assessment method based on Hurst exponent and Deep learning is proposed.The method is based on the normal operation data of the thermal power unit control system.First the Hurst exponent is used to analyze the data and classify the system performance level.Different controller parameter optimization schemes are given for different performance levels.Then use the LSTM neural network to establish a time series prediction model,monitor the adjustment effect of the Hurst index in real time and make recommendations on the original adjustment range.And the operation data under different working conditions is generated by Matlab/Simulink simulation,and the exponent is used for evaluation and optimization.Finally,the method is applied to industrial actual process to verify the feasibility and effectiveness of this method.The experimental results show that the Hurst exponent can accurately measure the adjustment strength of the controller and guide the optimization and adjustment of the controller parameters;The time series model established by the LSTM neural network has small error and accurate prediction,and can achieve optimal controller performance with a minimum number of adjustments.At the same time,performance evaluation software is designed by using GenSystem platform and Golden monitoring configuration platform,and thermal power unit control loop information,operation data and evaluation results and suggestions of the Hurst exponent algorithm are displayed on the visual interface.The application of the platform in engineering practice also proves that the algorithm is effective and accurate.