Research On The Fault Identification Of PV Arrays By Integrating The Analysis Of Electrical Characteristics And The Identification Of Model Parameters

Under the background of "the dual carbon goal" and the promotion of new power system,the installed capacity of photovoltaics is growing explosively,and it is of great significance to accurately and reliably identify the fault types of photovoltaic arrays to improve the stability of photovoltaic systems and reduce O&M costs.Starting from the external and internal characteristics of photovoltaic arrays,this paper conducted in-depth research on the output characteristics of each fault in the array,proposed a method for identifying the parameters of the photovoltaic array model,and established a fault identification model of the photovoltaic array which integrates the characteristics of external electrical and internal model parameters.Firstly,the single diode mathematical model of the PV cell was analyzed,and the explicit and implicit mathematical model of the PV array was established by combining the characteristics of the PV array composition and the Lambert W function.A 4×4 PV array simulation model and test circuit are built using MATLAB/Simulink to analyze the output characteristics of the PV array under different temperatures and irradiance.At the same time,four typical faults of PV arrays,namely short circuit,open circuit,abnormal ageing and shadow shading,as well as several complex faults that can be solved in the actual operation and maintenance process,such as low performance modules,ash accumulation,hot spot effect and mismatch loss,were simulated and analyzed to obtain the characteristics of changes in electrical parameters under various faults,which laid a certain theoretical foundation for the identification of faults in PV arrays later on.Secondly,aiming at the problem of parameter identification of photovoltaic array model.this paper improves on the traditional differential evolution algorithm,and proposes a fuzzy adaptive differential evolution algorithm for parameter identification of photovoltaic array model.In FADE,the DE/current-to-SP-best/1 mutation strategy is proposed,and more elite individuals are used to guide the search direction;Introduce parameter adaptive control strategy to reduce the dependence of algorithms on parameter settings;A fuzzy selection strategy is proposed to use the eliminated individuals as the next generation of random mutation mothers to improve the global search ability.In the same environment,the FADE algorithm and other different algorithms are used to solve the parameters of the PV array model,and the accuracy and effectiveness of the improved method for the identification of the parameters of the array model are verified by simulation and experimental data,which provides a basis for the extraction of model parameter features under each fault of the PV array.Finally,on the basis of the analysis of the electrical characteristics and model parameter characteristics corresponding to various faults of the photovoltaic array,the two types of fault characteristics are combined as the basis for fault identification.A PV array fault identification model based on particle swarm optimization random forest algorithm(PSO-RF)is proposed,and the effectiveness of the proposed PSO-RF model for PV array fault identification is verified by using the collected data sample set.At the same time,the PSO-RF model is used to verify the electrical characteristics and model parameter feature datasets of photovoltaic arrays respectively,and the results show that the fault identification by fusing the electrical characteristics and model parameter characteristics of photovoltaic arrays is more accurate than that of a single fault feature.