Photovoltaic Fault Diagnosis Based On GA-AO Hybrid Optimization Algorithm And BiLSTM

With the increasing prominence of energy and environmental problems,photovoltaic power generation technology has become more and more mature,and the monitoring and fault diagnosis of photovoltaic arrays have become more prominent with the wide application of photovoltaic power generation technology.Aiming at the shortcomings of the existing photovoltaic fault diagnosis methods,this paper develops a photovoltaic fault diagnosis model based on genetic algorithm and Aquila optimizer combine algorithm and bidirectional long short-term memory network.First of all,considering the establishment of the mathematical model of the photovoltaic power station,it plays an important role in the operation analysis of the unit.Based on the introduction of photovoltaic power generation,the photovoltaic modules,booster circuit,inverter circuit and filter circuit are introduced in detail.The mathematical model of photovoltaic power generation system lays the foundation for the later research on photovoltaic fault diagnosis.In order to optimize the relevant hyperparameters of the bidirectional long-short-term memory network,for the genetic algorithm and the Aquila optimizer algorithm,it is easy to fall into the local optimum and the convergence speed is too slow.the optimized population that satisfies the transition conditions is used as the initial population input of the Aquila optimizer,and the combine optimization algorithm is obtained.The standard test function is used to test the effectiveness of the algorithm.The proposed algorithm can greatly improve the speed and accuracy of solving the optimal value of the function.Finally,modeling and simulation are carried out for the six faults of open circuit,short circuit,partial shadow,shadow shading,photovoltaic dust accumulation and photovoltaic hot spot that often occur in photovoltaic power plants.And analyze the output characteristics of the photovoltaic array before and after the photovoltaic fault,and then obtain the characteristic parameters of the six faults according to the simulation experiment and input the GA-AO-BiLSTM fault model.And compared with the PSO-BiLSTM,GA-BiLSTM and AO-BiLSTM fault models,BiLSTM fault model accuracy is 99.44%.By building an experimental platform to simulate photovoltaic faults,and inputting the acquired fault parameters into the trained fault model,it is verified that this fault diagnosis model is superior to other models.The proposed method can accurately and effectively improve the fault diagnosis performance of photovoltaic substations.There are 59 figures,9 tables and 53 references in this paper.