With the large-scale application of photovoltaic modules

the update and maintenance of photovoltaic modules become a problem that cannot be ignored.In the use of photovoltaic modules

heat spots caused by failure

aging caused by daily use

and damage caused by external forces such as weather factors will cause abnormal operation.Firstly

the states of photovoltaic modules are roughly divided into normal

aging state

open circuit fault

short circuit fault

local shade.In order to make repair plan or replacement plan of photovoltaic module more accurately

it optimizes the traditional neural network fault analysis strategy.Secondly

by introducing the feature importance analysis of the random forest algorithm

it adds the fill coefficient

the slope of the U-I curve and the power parameters to improve the accuracy of the neural network in fault judgment.At the same time

it uses the genetic algorithm to optimize the BP neural network algorithm and solve the problem that the accuracy of accuracy is poor and it is easy to fall into the local optimal.Finally

the experimental results show that after introducing these parameters

the accuracy of fault type determination is improved.The error between the theoretical estimate and the actual result is reduced by 5%～11%

and the optimized diagnosis accuracy is increased by 2.4% compared with the traditional BP neural network fault diagnosis model.