Planning of Microgrids Coalition Considering Renewable Energy Prediction based on Machine Learning Algorithm

Abstract

This paper presents an energy management plan to schedule the different generation resources in a smart distribution system composed of several microgrids under normal and abnormal conditions. In this context, after identifying the main characteristics for a management scheme, an appropriate framework is planned and the functions of various management units in a multi-microgrid system are initiated. In the first step, the microgrids plan their generation resources using a new model based on the framework. In the second step, the distribution system operator determines the possible power transfers the microgrids and utilizes the remaining capacity of the microgrid resources for supply the unserved loads in the first step. To determine the available capacity for renewable energy resources such as solar and wind power, the Extreme Learning Machine (ELM) is used to predict the value of solar irradiation and wind power. The proposed plan is implemented on a test system under normal and various abnormal events via realistic case studies.