Solar Radiation Prediction Using Temporal Gaussian Process Regression

Solar energy is an important source of renewable energy that can be harnessed using a range of evolving technologies such as solar heating, solar photovoltaic, solar thermal energy, solar architecture and artificial photosynthesis. The harnessed solar energy can be used in a wide range of applications like electricity production, fuel production, agriculture planning, water heating, transport, etc. The prediction is focusing in the Southern part of India and the solar light will be available from 8 to 9 months in a year in this region. So to utilize the solar energy in an efficient way the prediction is done. To predict the availability of solar energy a machine learning Temporal Gaussian Process Regression(TGPR) method has been used. It provides better result and also more robust when compared with the existing methods using ELM, SVM, etc. The predicted values could be used to measure and analyze the amount of energy that could be generated and in turn to identify the suitable solar based devices that can be installed in different locations.