The working environment of the ring main unit(RMU) in large solar photovoltaic power plants is complex and variable

faced with harsh environments such as temperature differences and humidity

it is extremely easy to cause operational failures of the ring grid cabinet

which seriously affects the safe and stable connection of solar photovoltaic to transmission lines. Based on the measured temperature and humidity data inside the RMU

utilizing the advantages of ARIMA and RBF model in linear and nonlinear data processing

a temperature and humidity prediction model with ARIMA-RBF weight combination is constructed to dynamically predict the temperature and humidity inside the RMU. The dynamic prediction of temperature and humidity in the actual loop cabinet of a photovoltaic power station is carried out. The prediction results show that

compared with the single model

the ARMI-RBF weight combination model has higher prediction accuracy and better stability.The combined model gives full play to the processing ability of a single model for different characteristics of data through appropriate weighting strategies

and can better evaluate the temperature and humidity state inside the RMU. It can provide a reference for the establishment of a more universal prediction model

and help to reduce the failure caused by long-term operation of the ring cabinet under ultra-mild and humid environment.