Study On Fault Diagnosis And Health Monitoring Technology Of Photovoltaic Power System

With the further increase of energy demand and continuous development of photovoltaic industry,solar energy has become an important source of energy,and the importance of photovoltaic power system in the energy supply is gradually increasing.The online monitoring and fault diagnosis of photovoltaic power system has become a key factor to ensure the safety of the photovoltaic power system.Due to the non-linear characteristic,the photovoltaic power system can't be described by a simple mathematical model.In this thesis,the fault diagnosis and health monitoring technology is studied with the research object of the photovoltaic power system and the technical background of data monitoring and data mining.First of all,this thesis introduced the structure and composition of photovoltaic power system.The key technologies and main principles of photovoltaic power system were introduced from three aspects:photovoltaic array,maximum power tracking algorithm and photovoltaic inverter control.Then a set of distributed photovoltaic power unit simulation model was established,the operating status of photovoltaic arrays was simulated under different environmental conditions,then the effects of different irradiance and temperature on the power generation of photovoltaic power system were analyzed.Secondly,with the advantages in high-dimensional non-linear pattern recognition,support vector machine(SVM)was applied to the photovoltaic power fault diagnosis.The improved particle swarm optimization(IPSO)algorithm was proposed to optimize the parameters of SVM.This method can promote the convergence speed and accuracy of the fault diagnosis algorithm and avoid the particle swarm algorithm optimization's falling into the local optimum.Based on the established photovoltaic power system simulation model,a variety of faults in the photovoltaic array got trained and predicted,then the rapidity,reasonability and effectiveness of the proposed algorithm were proved.Finally,this thesis proposed a health analysis method of photovoltaic power system based on the variable weight analytic gierarchy process(AHP).This method was aimed at evaluating the health status of the photovoltaic power generation system with the real-time data and the stored historical data from the photovoltaic power monitoring system.This health analysis method,which realized the quantification of the degree of fault,is the most intuitive manifestation of the operating status of photovoltaic power system.