Artificial Intelligence Based Solar Radiation Predictive Model Using Weather Forecasts

Solar energy has gained attention in the past two decades, since it is an effective renewable source that causes no harm to environment. Irradiation Prediction (SIP) essential plan, schedule, and manage photovoltaic power plants grid-based generation systems. Numerous models have been proposed for SIP literature while such studies demand huge volumes of weather data about target location a lengthy period time. In this scenario, commonly available Artificial Intelligence (AI) technique can be trained over values irradiance as well weather-related parameters temperature, humidity, wind speed, pressure, precipitation. Therefore, current study, authors aimed at developing solar prediction model by integrating big analytics with AI (BDAAI- SIP) using forecasting data. order perform long-term collection data, Hadoop MapReduce tool employed. The operates on different stages. Primarily, preprocessing take place various sub processes conversion, missing value replacement, normalization. Besides, Elman Neural Network (ENN), type feedforward neural network also applied predictive analysis. It divided into input layer, hidden load-bearing output layer. To overcome insufficiency ENN choosing weights layer neuron count, Mayfly Optimization (MFO) algorithm applied. validate performance model, series experiments was conducted. experimental infer outperformed other methods used comparison.