Power Tube Dynamic Redundancy Control And Fault Diagnosis Of Three-phase Photovoltaic Grid Connected Inverter

With the rapid increase of global energy demand,the stock of traditional fossil energy has declined sharply and the problems such as the greenhouse effect have become increasingly prominent.Finding clean,high-reserve new energy sources has become the focus of attention and research object.Among them,photovoltaic power generation has become the fastest-growing new energy application technology due to its advantages of low pollution and convenience.However,based on statistical data,as the core of photovoltaic systems,the reliability of photovoltaic grid-connected inverters is low,and the power tubes in them are prone to failure,which affects the power generation efficiency and reliability of photovoltaic systems.Therefore,studying how to improve the reliability and fault diagnosis methods of grid-connected inverters is of great significance for improving the efficiency of photovoltaic power generation systems and reducing losses.Firstly,a three-phase photovoltaic grid-connected system simulation model is designed and built.The results show that the output of the simulation model meet the requirements of the GB/T37408-2019 standard.The necessity of fault diagnosis and the feasibility of fault diagnosis based on three-phase output current can be proved by analyzing and studying the output current of three-phase inverters and the junction temperature characteristics of IGBT modules after different IGBT open-circuit faults,and summarize the fault characteristics.Secondly,a dynamic redundancy control method for a three-phase two-level inverter is proposed by using the inverter redundancy topology.In this paper,a four-arm reversing switch cycle switching scheme is designed,and a dynamic redundant control strategy is realized by redistributing three-phase IGBT control signals.An inverter based on a dynamic redundancy control strategy is design,and simulation is performed to verify the effectiveness of the dynamic redundancy control strategy in suppressing peak IGBT junction temperature.Then,based on the optimized extreme learning machine（ELM）and the improved gray wolf algorithm（IGWO）,an IGWO-ELM identification model for multi-tube open-circuit faults in inverters is proposed.This paper compares the normalized kurtosis and normalized pulse index of the inverter fault output current,and determines the seven-dimensional eigenvector of the open circuit fault[I_a-kur,I_b-kur,I_c-kur,I_a-f,I_b-f,I_c-f,α];Then,a seven-dimensional eigenvector of the open-circuit fault is used as the input,and the open-circuit fault number of the inverter is used as the output to construct a limit learning machine identification model for the inverter power tube fault.Then,the improved Grey Wolf algorithm（IGWO）is introduced to optimize the initial value of the extreme learning machine（ELM）.The validity of the IGWO-ELM identification model is verified by simulation and experiments.Finally,for the non-redundant bridge arm inverters,non-dynamic redundant control inverters and dynamic redundant control inverters,the corresponding reliability models are constructed to calculate the failure rate of the components and the reliability of the inverter system and to verify that the system reliability of the dynamic redundant control inverter will be improved.This paper defines thermal safety to quantitatively characterize the thermal safety of IGBT modules and inverters,to verify that the thermal safety of dynamic redundant control inverters will be improved.Based on the January illumination and ambient temperature data of a certain place,the working junction temperature of the inverter IGBT module is simulated,and the cumulative damage and remaining life are calculated,verifying that the dynamic redundant control inverter will have better durability.