Fault Diagnosis Of Photovoltaic System Based On Principal Component Optimized Neural Network

Solar energy is indispensable in the supply of electricity.The position of photovoltaic modules in photovoltaic systems is self-evident,and the problem of failures cannot be underestimated.Multi-level inverters have excellent results in high-voltage and high-power fields such as start-up adjustment of variable frequency motor speed,integration of grid energy storage,and static reactive power compensation equipment.Among them,the neutral point clamped three-level photovoltaic inverter is the most comprehensive.It is convenient and efficient,with low switching attenuation and stable voltage waveform.In this paper,the photovoltaic array and inverter switch tube are open,and specific faults are identified to ensure a reliable sy stem.This paper first builds the mathematical model and simulation structure of the photovoltaic array,and uses deep neural network algorithm to realize the fault feature extraction and diagnosis of the photovoltaic array on this basis.Next,the output voltage signal collected in the simulation experiment of the three-level photovoltaic inverter is processed by the principal component analysis optimization algorithm,and the characteristic parameters of the output voltage are extracted by matrix transformati on and reduced dimension,and the principal components under each fault condition are obtained.Analyze the optimization results,and obtain the dominant factors that cause the open-circuit failure of the three-level inverter.The BP neural network and genetic-neural network are used to diagnose faults in NPC inverters.The overall optimization features of the genetic algorithm overcomes the deficiencies of the BP neural network like easily fail into minimum value and slow learning speed,and improves the accuracy of fault diagnosis.Finally,rely on LabVIEW software,it constructs an inverter fault offline diagnosis platform based on the above algorithm.This platform can display different types of open circuit faults and output waveforms of the inverter,which has certain practical value and economic significance.