Research On Parameter Identification And Fault Identification Of Photovoltaic Module Model

Against the backdrop of the energy crisis and the growing problem of environmental pollution,the installed capacity of photovoltaic power generation systems is increasing explosively.As the core of the PV system,the management and maintenance of PV modules are of great significance for the safe operation of PV plants and power systems.Therefore,this article proposes a method to identify the model parameters of PV modules and establishes a fault diagnosis model based on the model parameters of PV modules.The specific research of the thesis is as follows:Firstly,since the parameter are closely related to the output characteristics of solar PV models,this paper briefly introduces the single diode model of photovoltaic modules,and then on this basis,the effects of photogenerated current,diode reverse saturation current,series resistance,parallel resistance and diode ideal factor on the external output characteristics of photovoltaic modules are studied,which lays a foundation for the following research.Secondly,in order to efficiently extract the model parameters of PV modules,an identification method based on dynamic elite leader M ulti-Verse optimizer algorithm is proposed in this paper.In DLMVO,an adaptive strategy of control parameters based on population evolution rate and aggregation rate is introduced to balance the exploration and exploitation of the algorithm,avoiding the search to fall into local optimum;A dynamic elite leader-based variation strategy is proposed to enhance the probability of variation success and improve the speed of merit search.DLMVO was applied to the parameter extraction of two different PV modules and validated using 6 algorithms for comparison,and module parameters under different operating conditions were also identified.The experimental results show that DLMVO has good applicability,accuracy,robustness and convergence.Finally,the Ⅰ-Ⅴ characteristics of four typical faults of PV modules,namely,shadow shading,short circuit,open circuit and aging,are analyzed in depth,and a probabilistic neural network fault diagnosis model with model parameters as input layer vectors and module faults as output layer vectors is established by combining the effects of model parameters on module Ⅰ-Ⅴcharacteristics and the results of model parameter identification under faults.The proposed fault diagnosis model is used to conduct diagnostic tests on the collected experimental simulation data,and the test results verify the effectiveness of the method.