Multilayer perceptron based control of hybrid power system for DC homes

The current development of electric power systems brings about the need to deal with increasingly multifaceted interactions of various technical components and relevant actors in order to integrate more comprehensive spectrum of different aspects into a single system. Thus, here is a proposed system of an intelligent control system using multi perceptron method of supervised learning, where the system will be continuously gathering a training set of time, output of various sensors, load, etc and finally learn how to switch between Grid Connected Hybrid Power System for DC Homes, where system is dealt with utility grid, battery and solar PV to power DC load in homes. The performance of the system is then compared by varying the number of samples. Finally the system is learnt and the needful switching between three power sources, utility grid, battery and solar PV is efficiently done based upon time, weather and load, which are inputs to the system.