Research On Fault Early Warning Of Thermal Power Unit Based On Information Fusion Technology

There are many equipments and complex structures in large thermal power units.The thermal parameters indicate the running state of the equipment directly.Any abnormal change of the parameters will have a serious impact on the thermal process and bring serious consequences to the power production and power users.Therefore,it is necessary to monitor and warn the running status of thermal equipment in real time.In this context,the paper aims to investigate several problems of thermal equipment fault early warning based on information fusion technology.The main research work and achievements are as follows:Firstly,a data preprocessing method based on wavelet analysis and density bias sampling is proposed to deal with the large amount of data and noise interference in thermal power units.The original sample data is de-noised by using wavelet analysis method,then the typical sample is selected by density bias sampling(VGDBS)based on variable grid division,and finally,the data of high-adding system is analyzed as an example.It not only eliminates the noise interference,but also reduces the size of the sample data,and provides high-quality data for the subsequent clustering analysis.Then,an improved whale algorithm is proposed to optimize the fuzzy C-means clustering algorithm(IWOA-FCM),which is sensitive to the initial cluster centers and needs to be given the number of clusters in advance for the fuzzy C-means clustering algorithm(FCM).Four clustering validity indexes are used to determine the optimal number of clusters,and then IWOA algorithm is used to optimize the initial cluster center of FCM algorithm.FCM algorithm and IWOA-FCM algorithm are used for cluster analysis of the sample data of high-adding system.The results show that IWOA-FCM algorithm not only has a faster convergence rate,but also has a better clustering effect.Finally,the fault early warning problem of thermal equipment is studied.Thermal process is usually has a certain parameter foreboding in the early stage.In view of the above characteristics,a fault early warning method based on evidence theory and KNN algorithm is proposed.The results show that the method proposed in this paper is reasonable and superior in fault early warning of thermal equipment.