Development of a Self-adaptive based Renewable Energy Management in Smart Micro Grid

Abstract

A self-adaptive artificial intelligence (AI) algorithm is a combination of self-adaptive data-analytics methods with machine learning AI algorithms. In this work, the renewable energy is focusing on solar energy, and the purpose of using a self-adaptive artificial intelligence algorithm is to balance the trade-off between renewable energy (RE) and non-renewable energy (NRE). Hence, this energy management system will integrate optimally the RE sources with the help of AI algorithms. In addition, it can reduce the use of NRE, i.e. the national grid energy and minimize RE wastage in a hybrid system. It is a smart microgrid because the user can schedule the load and monitor the power usage through an Internet of Things (IoT) based system. In this proposed work, the boiler is a time schedule load which only operates when reaches the predetermined working hours. Besides, the user manage to monitor the real-time data of the environment condition, load usage, and power consumption through a smartphone android application. The proposed self-adaptive AI algorithm successfully manage to predict the usage of the power sources, either RE of solar sources, or NRE of either battery or national grid. It is proved in the experiment by considering two conditions, i.e. hot weather and rainy weather. Moreover, an alarm system is also added to notify the user to switch off the load when there is more than half an hour of no human motion detection.