Research On Soft Fault Diagnosis Method Of PV Inverter

With the development of power electronic technology,power electronic equipment becomes more complex.The fault diagnosis of power electronic circuits is more important.As a typical power electronic circuit,photovoltaic inverter circuit has been paid attention by researchers.Taking NPC three-level inverter circuit as an example,this paper studies a soft fault diagnosis method caused by component parameter changes.The specific work is as follows:First of all,this paper selects NPC three-level photovoltaic inverter circuit as the research object.By analyzing the working principle of NPC three-level inverter circuit,it sets the type of capacitor soft fault,and then uses MATLAB software to build a simulation model for the inverter circuit.Secondly,to solve the problem that it is difficult to distinguish the soft fault features caused by the degradation of capacitance parameters,a soft fault feature extraction method based on WOA-VMD wavelet energy is studied.Firstly,the whale algorithm(WOA)is used to optimize the center frequency and decomposition level of the variational modal decomposition algorithm(VMD),determine the appropriate center frequency and decomposition level and decompose the fault signal,then extract wavelet energy from the decomposed modal component to construct the feature vector,and finally use support vector machine(SVM)to classify the fault.The experimental results show that,compared with the traditional wavelet and VMD Shannon entropy methods,the soft fault feature extraction method based on WOA-VMD wavelet energy has great advantages in diagnosis time and accuracy.Finally,because the kernel parameters and penalty factors of SVM classifier have a certain impact on the diagnostic performance,it is difficult to select appropriate parameters.Therefore,a soft fault diagnosis method based on IWOA optimized support vector machine(SVM)is studied.Firstly,the advantages and disadvantages of whale algorithm are analyzed.The traditional whale optimization algorithm is improved by using feedback mechanism and nonlinear factor,and IWOA is tested on the benchmark function;Secondly,the improved whale algorithm(IWOA)is used to optimize the kernel parameters and penalty factors in support vector machine(SVM);Finally,the IWOA-SVM diagnostic model is established,and the models of different algorithms(IOWA,WOA,PSO)combined with different classifiers(SVM,ELM,BP)are compared.The experimental results show that the soft fault diagnosis method based on IWOA-SVM has a better effect on the soft fault diagnosis of photovoltaic inverter circuit.