Research And Realization Of Key Technology For Operation And Maintenance Platform Of Thermal Station

At present,the integration of modern information technologies such as mobile Internet,big data analysis and artificial intelligence with the service management of heating enterprises has become a trend,which can promote the intelligent construction of heating enterprise management and service system,and improve the intelligent,efficient and refined management of heating stations.At present most of the heating enterprise did not do it on the heating power station monitoring system for regular maintenance,result in the thermal station operational data lack of reference.Therefore it is of great significance to set up operational data remote monitoring platform and mining the operational data potential information of heating power station so as to play the value of operational data while monitoring the thermal station operational data.Based on the analysis and research of the actual usage and operation and maintenance data characteristics of a thermal power station in Kaifeng city,this paper makes a detailed analysis of the required functions of the operation and maintenance platform of the thermal power station.According to the needs of the thermal station operation and maintenance data monitoring,analysis and visualization of functional requirements,the thermal station operation and maintenance platform design objectives,and the need to complete the content.According to the requirements,the overall architecture,implementation function,data storage and platform interface of the platform are designed in detail.Finally,based on the historical operation and maintenance data of thermal power station,the thermal load prediction is made,and the design of remote monitoring platform for operation and maintenance data of thermal power station is realized.Accurate prediction of heat load can make heat supply enterprises arrange heat supply plan reasonably according to the demand of heat users.The influence of the historical operation and maintenance data of the thermal power station on the realization of the thermal load prediction is comprehensively analyzed and the input sample data of the prediction model is determined.The support vector regression(SVR)prediction model was established,and the control parameters of the prediction model were optimized by using grid search method,genetic algorithm and particle swarm optimization algorithm on the basis of cross validation.A comprehensive comparison of the three algorithms on the performance of the SVR prediction model is made to determine that the performance of particle swarm optimization algorithm is not only high in prediction accuracy,but also short in time,which can be used in the thermal station operation and maintenance platform for thermal load prediction.