Research On The Construction Of High-pressure Heater State Warning Evidence Base

The safe operation of the power grid affects all walks of life and people's life and production.To maintain the security of the power grid is not only to rely on the power grid enterprises,but also to the power plant.The safe and stable operation level of the unit and equipment directly affects the stable operation level of the power grid.The diversified power supply structure and complex operation mode put forward new requirements for the reasonable and effective development of the power side technology supervision.Therefore,it is imminent to study the early warning of the operation state of the power generation equipment.This paper studies the construction of evidence base in the warning system of high pressure heater based on evidence theory.This paper introduces the failures and preventive measures of HP heaters,summarizes the research status of the equipment status monitoring and early warning,and the development status of evidence theory and evidence construction,and derives the feasibility of evidence theory in the early warning field of HP heater.There are errors and random errors in the history data of HP heater.The error correction is done by using the improved observable rate of change test method and the wavelet threshold de noising method based on SURE,and the reliability of the method is verified by simulation experiments.Because the measurement points of the running state of the high pressure heater are many and the amount of data is huge.In order to reduce the operation time and improve the reaction speed,it is necessary to select the measuring point of the running state.By using the principal component analysis method,the selection of the measuring point of the running state is realized.The practicability of the grey correlation entropy method and the principal component analysis method is compared in the simulation experiment,and the application scope of the principal component analysis method is verified.The concept of DS evidence theory is introduced,and data clustering is used to excavate the state information of high pressure heater in the data.For fuzzy C mean clustering algorithm,it is easy to fall into local optimum.Combined with simulated annealing algorithm and genetic algorithm,FCM clustering based on genetic simulated annealing algorithm is formed,which strengthens global search ability.Based on the data clustering results,the stateearly-warning evidence base of high pressure heater is constructed.The concept of DS evidence theory is introduced and data clustering is used to excavate the state characteristics of high pressure heater in the data.Fuzzy C mean clustering algorithm and FCM clustering algorithm based on genetic simulated annealing algorithm are used to get data clustering results.Combined with evidence theory,a state warning database of HP heater is built.