Research On Performance Evaluation And Parameter Optimization Of Thermal Control System Based On Mahalanobis Distance

Thermal power units still occupy a dominant position in Chinese power system,and the increase of installed capacity of thermal control system requires the continuous improvement of factory automation level.But with the long-term operation of the system and the existence of various interference factors,the performance of the system is declining,which has an impact on the safety and economy of the whole plant.Therefore,it is an inevitable trend to build a reasonable and effective performance evaluation and optimization system of the control system.This paper takes the superheated steam temperature control system as the research object,and explores the method of performance evaluation and parameter optimization based on the conventional operation data.The main research contents are as follows:(1)Taking the three-stage water spray cooling control system of 600 MW oncethrough boiler as the research object,the application of inverse distance interpolation algorithm in the calculation of the whole process model of the controlled object is studied,and the whole process simulation model of Simuink is built.The practice shows that the establishment of the model has a certain degree of credibility.(2)In order to solve the problem that the traditional evaluation algorithm depends on the system model,a performance evaluation method of multivariable control system based on improved Mahalanobis distance is proposed.Based on the conventional operation data of the control system,the method constructs the benchmark of performance evaluation through Hurst index optimization to determine the data set representing the optimal performance of the system,improves the calculation method of Mahalanobis distance,analyzes the correlation between the samples to be tested and the best performance data set,calculates the α-MD performance index,and uses the membership function to divide the performance grade.The simulation results show that compared with the algorithm before the improvement,this method can optimize the performance evaluation benchmark,more comprehensively quantify the deviation between the sample set to be tested and reference sample,and greatly improve the accuracy of system performance evaluation.(3)By establishing the LSTM time prediction model,monitoring the change of theα-MD index,and giving the regulation suggestion of the controller parameters.Using the simulation data of the third stage superheated steam temperature control loop of600 MW unit under 37% operating conditions,the evaluation and optimization are carried out.The experimental results show that the LSTM time prediction model has higher prediction accuracy,can effectively guide the process of the controller parameter optimization,and effectively improves the performance of the loop in the process control system.(4)The performance evaluation platform based on α-MD index is designed by Gen System,the loop information of the control system and the evaluation results are displayed to the user through a visual interface,and the practical engineering application verifies that the platform has certain practical significance.