Research On Intelligent Pattern Identification Method In PV Array Fault Diagnosis

With the photovoltaic power generation technology is more and more widely used,a large number of photovoltaic systems have been put into operation.The actual operation of the photovoltaic power station is harsh in nature and the complex and diverse photovoltaic array faults seriously affect the safe and stable operation and economic benefits of the power station.At present,the monitoring resolution of the DC side of the photovoltaic power station has reached the level of the photovoltaic array,and even the resolution of some intelligent photovoltaic power plants has reached the component level.It is of great significance to realize the intelligent predictive maintenance of photovoltaic power station and improve the operation level of photovoltaic power station.In this paper,a new method of PV array fault diagnosis based on intelligent identification is proposed based on the analysis and research of typical PV array faults.The correctness and effectiveness of the proposed method are verified by simulation and experiment.The main work done in this paper is as follows:(1)The operation characteristics of PV array structure and PV modules are analyzed,and several common faults of PV array are analyzed.The causes of PV array faults are summarized and the theory is put forward for the following methods of PV fault diagnosis based on intelligent recognition basis.(2)A mathematical model of photovoltaic cell and photovoltaic array simulation model suitable for different fault conditions are established.With the aid of photovoltaic array simulation model,the output characteristics of photovoltaic modules under different light intensities and different temperatures are simulated.(3)Aiming at several common fault types(short circuit,open circuit,partial occlusion)of PV array,a new method of PV array fault diagnosis based on FCM algorithm and fuzzy membership algorithm is proposed.The proposed method Diagnostic methods were verified.The results show that the diagnostic method proposed in this paper can effectively discriminate the typical faults of the PV array.The method is simple and reliable and has good diagnostic results.(4)A new diagnosis method based on FCM and PNN is proposed.Based on chapter 4,the fault pattern of PV array is further refined,and the mapping relationship between electrical characteristics and fault features under different fault conditions is established.Several fault types are classified by FCM clustering algorithm,And then through the PNN algorithm to achieve photovoltaic array fault diagnosis.Finally,the system simulation in MATLAB environment is used to verify the correctness and effectiveness of this method.