SIF-DWTRL: solar irradiation forecasting using discrete wavelet transform and regression learning

In current scenario, the non-renewable resources of energy are continuous shifting towards renewable sources of energy owing to the rising global warming, environmental pollution and depleting reserves of fossil fuels. The most promising renewable source of energy happens to be solar energy whose share in continually increasing among energy sources which are leveraged. Currently, many industries are consuming a huge amount of solar energy for their applications, thus there is a need of establishment of solar power plants with substantial capacities. In order to fulfill this objective, one of the key challenges is estimating the amount of solar irradiation which can be achieved by solar irradiation forecasting which is one of the crucial area of research that direct implicants into generation of solar power. Therefore, in order to increase the generation rate of solar power, in this paper, we are proposing a new technique named as Solar Irradiation Forecasting using Discrete Wavelet Transform and Regression Learning (SIF-DWTRL). Here, the prior to employing regression learning to solar data, the raw data is pre-processed using the discrete wavelet transform to filter out local disturbances and facilitate the pattern recognition process. The obtained results show that the proposed approach attains a mean absolute percentage error of 1.16% and a forecasting accuracy of 98.43%. The regression achieved in 0.948 indicating the fact that the proposed approach attains a high and reliable forecasting accuracy.