Field Reliability Assessment Of PV Modules Based On Regional Clustering Of Factors Influencing Performance Degradation

Photovoltaic(PV)power generation system mainly consists of PV module,controller,inverter,etc.Among them,PV modules are one of the core components to carry out photoelectric conversion and ensure the continuous operation of the whole system.Its reliability directly affects the PV system’s photoelectric conversion efficiency and service life.Due to the operating conditions of PV modules at different geographic areas differ in varying degrees,there is certain variability between their field reliability and inherent reliability.Therefore,it is of great value and significance to accurately assess the field reliability of PV modules under different service conditions for scientifically guiding the site selection and operation and maintenance decisions of PV power plants.For this reason,this paper takes the PV modules as the research object and conduct an in-depth study in four aspects: regional differences in factors influencing performance degradation,field reliability assessment method,applicable range of field reliability assessment results,and warranty optimization design based on field reliability.The main research work completed in the thesis is as follows.1.Regional differences in the factors influencing module performance degradation were investigated.Firstly,a systematic and comprehensive analysis of the influencing factors in terms of workload and natural wear dimensions was conducted according to the classification criteria developed.Secondly,the theoretical method based on time series quantitatively reveals the differences of the influencing factors in different regions,and then a research method and route based on regional clustering of influencing factors was proposed.Finally,the panel data model and clustering model of the influencing factors were constructed.Besides,the clustering indicators were determined,and the evaluation indices of the relevant clustering algorithms were given as well,which laid the foundation for the solution of the subsequent clustering results.2.A field reliability assessment method based on regional clustering and cumulative damage model was proposed.In order to characterize the effects of multiple environmental stresses on the module performance degradation,the general form of mathematical model reflecting the relationship between performance degradation and environmental stresses was first established based on the cumulative damage model.Second,according to the analysis of the aforementioned influencing factors,five quantifiable environmental factors,including module temperature,module cycling temperature,ambient relative humidity,UV irradiance and PM10 concentration,were incorporated into the model and specific functional expressions were determined too.The influence of other unquantifiable environmental factors was reflected by the remaining terms of the model.Meanwhile,the solution method of the model parameters was given.Finally,considering the regional properties of the model parameters,regional clustering based on the influencing factors was then proposed to determine the applicable range of the parameter values,which ensures the estimation accuracy of the model.3.The issue of expanding the applicable range of field reliability assessment results was studied.First of all,the necessity of studying this problem is analyzed from the perspective of maximizing cost-effectiveness,and a solution based on regional clustering of influencing factors was given.The overall logical flow of the solution was analyzed and elaborated at the same time.Second,on the basis of the regional clustering results with the best consistency in influencing factors of different dimensions,the weights of influencing factors of each dimension were further measured by using mutual information theory and solving partial derivatives of multivariate functional variables.Lastly,the panel data were weighted based on the weight calculation results,and the regional clustering results with the best consistency in the field reliability of each dimension were obtained separately,based on which the final comprehensive clustering result reflecting the information of the two dimensions was obtained by the secondary clustering method.Accordingly,the applicable range of field reliability assessment results can be reasonably determined.4.A warranty optimization method based on field reliability is proposed.At first,assuming the unified pricing and warranty strategy is adopted,a quantitative relationship model reflecting the regional variability of field reliability was established with warranty period and selling price as the decision variables to be optimized and profit as the optimization objective according to the obtained regional clustering result.Sales volume,warranty costs,and research,development,and production costs were considered in the model.Secondly,warranty cost was obtained by estimating the failure ratio of PV modules in different clustering regions for a given warranty period through the reliability assessment model,and then the joint optimization of warranty period and selling price was achieved.Finally,the influence of sales period on the mean time to failure and failure ratio of PV modules under the optimal warranty period was further analyzed,which lays a theoretical foundation for manufacturers to reasonably arrange inventory and logistics planning,etc.The research results not only improve and supplement the existing theory and method of field reliability assessment of PV modules,but also can provide more scientific guidance for engineering practice.