VERIFICATION OF ANN SOLAR RADIATION PREDICTION ALGORITHM FOR REAL-TIME ENERGY SIMULATION

Real-time building simulations are necessary to evaluate the performance of building systems in real-time. This paper presents an Artificial Neural Network algorithm that predicts global solar radiation only using readily available weather data such as temperature and humidity. EnergyPlus weather converter program is used to calculate diffuse and direct normal radiations from the predicted global solar radiation and to generate the EPW file needed for EnergyPlus simulation. Simulations are conducted using two weather files, 1) measured 2018, and 2) predicted 2018 using ANN output. This study evaluates the accuracy of using predicted solar radiation in the building simulation process.