Research On Key Model Of Household Photovoltaic Abnormal Diagnosis Based On Smart Meter

The photovoltaic array composed of photovoltaic modules in series and parallel is the power part of the photovoltaic system.With the widespread application of household photovoltaics in the world,fault identification and diagnosis of photovoltaic systems have become the fundamental guarantee for its continuous and sound development.Most of the existing methods at home and abroad are only suitable for enterprise-level photovoltaic users,but there are few fault identification and diagnosis studies applicable to a large number of household photovoltaics.First of all,this article summarizes and summarizes the current research status of existing methods and technologies from the three aspects of photovoltaic module fault classification,abnormal recognition,and fault diagnosis,sorts out the problems and difficulties in household photovoltaic applications,and provides feedback to households.The research topics,key technologies and development trends of photovoltaic abnormal identification and fault diagnosis are prospected.Secondly,the UI/UP curve is the core power characteristic of photovoltaic modules.Single or compound failures of photovoltaic modules will directly cause the deformation of the UI/UP characteristic curve of the photovoltaic array,thereby affecting the maximum possible output of the array,that is,the array output under MMPT.On the basis of the photovoltaic array engineering model,this paper proposes the equivalent circuit of a typical single fault of photovoltaic modules,and obtains the characteristic deformation curve of the single fault photovoltaic array through simulation,and analyzes the maximum possible output impact of each single fault on the array.Sorting analysis;further,this paper proposes a unified multi-fault simulation model of the array based on a single-fault equivalent circuit,and obtains the characteristic deformation curves of photovoltaic arrays for various combined faults,and also sorts and analyzes the maximum possible output impact in each scenario,So as to grasp the impact of various failures on the output of the arrayThird,the metering smart meter installed by the grid company for the household photovoltaic power generation side is the measurement source of almost all household photovoltaics,and it is also the only available data source for household photovoltaic fault judgment and fault diagnosis;due to the consistency of weather and microclimate,Large-scale household photovoltaics in counties will show a certain spatial correlation.Therefore,this paper proposes an improved ARMA household photovoltaic cluster prediction model based on spatial correlation to form household photovoltaic households who are not able to detect incompatibility in the cluster.Judgment method of abnormality and failure.Finally,for the household photovoltaic households detected in the group whose output is not compatible with the group,this paper proposes a power loss analysis method based on common failures of household photovoltaic systems.Through the simulation of common household photovoltaic systems within the county,the array failures of different capacity systems are summarized.Based on the daily power generation per unit capacity,two types of abnormal user identification criteria,3σ and box diagram,are proposed.For the identified abnormal users,the full-time power loss is calculated on the basis of the photovoltaic output prediction,and the fitting based on the least squares method is performed,and the fault type can be determined according to the degree of fitting and the output at different light intensities.The effectiveness of the method is verified by examples and simulations,which provides a new idea and method for the fault diagnosis of household photovoltaics.