Research On Remaining Lifetime Prediction And Maintenance Strategy Of Photovoltaic Power Generation System Based On Wiener Process

For the past few years,photovoltaic power stations have inevitably ushered in a high frequency period of equipment failure and aging.For the sake of avoiding the occurrence of faults and cascading faults,improving the service efficiency of the power station and reducing the operation and maintenance costs,it is of great significance to study the lifetime prediction methods and maintenance strategies of photovoltaic power generation systems.Based on the above research background,the work content of this dissertation as follows.(1)Establish the performance degradation model of photovoltaic modules.Aiming at the problem that the performance degradation rate of photovoltaic modules is low and the degradation process shows the unstable and fuzzy characteristics of attenuation over time,a new power degradation model of PV modules on the basis of Wiener stochastic degradation process is proposed.Firstly,the parameters in the model are randomly variable to portray the non-monotonicity of the photovoltaic modules degradation process,as well as the time uncertainty and individual variability of the module degradation process.Then,the degradation state of the modules are analyzed based on the performance degradation model.The degradation model lays the basis of subsequent remaining useful lifetime prediction and maintenance strategy research.(2)Study on the remaining lifetime of photovoltaic modules.To address the problem of the performance degradation rate of photovoltaic modules is low and it is difficult to collect long-term data to confirm the degradation path and lifetime,a method is proposed to construct a remaining useful lifetime model on the basis of the Wiener stochastic degradation model and to estimate the remaining useful lifetime of photovoltaic modules.Firstly,the remaining lifetime model of the photovoltaic modules is constructed based on the power degradation model of the photovoltaic module of the Wiener process.Then,based on the degradation trajectory of photovoltaic modules,the model parameters are updated adaptively in real time by using the state monitoring data in the service degradation process combined with bayesian update and expectation maximization algorithm.Finally,the remaining lifetime of photovoltaic module at any service moment is predicted based on the remaining lifetime probability function and distribution function update.The arithmetic analysis indicates that the proposed method improves the prediction precision of the remaining lifetime of photovoltaic modules.(3)Research on maintenance strategy of photovoltaic array.Aiming at the degradation process of photovoltaic arrays during service,a time-domain rolling majorization dynamic maintenance tactic based on array remaining lifetime prediction under incomplete maintenance behavior is proposed.Firstly,the residual degradation amount and degradation rate factor are used to portray the maintenance effect of incomplete maintenance behavior repairing non-new,and the remaining lifetime prediction model considering the maintenance effect is established.Then,a dynamic predictive maintenance tactic model is set up by regard the lowest average maintenance spend of the maintenance update cycle as objective function,and the optimal maintenance threshold and the optimal predictive maintenance moment are obtained.Finally,the predictive maintenance behavior execution time is determined by updating the remaining life under incomplete maintenance.An example is given to verify the economy and effectiveness of the dynamic maintenance strategy model.