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

Abstract

Abstract This paper presents an energy management plan to schedule of different generation resources in a smart distribution system comprised of multi-microgrids in normal and abnormal condition. In this regard, after detecting the main characteristics for a management scheme, a suitable framework is planned and the functions of various management entities in a multi-microgrids system are initiated. The microgrids schedule their generation resources in the first step using a new model based on the framework. In the second step, distribution system operator determines the possible power transmits between the microgrids and uses the remained capacities of microgrids’ resources for supplying the unserved loads in step I. In order to determine the amount of available capacity for renewable energy resources such as solar and wind energy, Extreme Learning Machine (ELM) is used to predict the value of solar radiation and wind power. The proposed plan is implemented on a test system under normal and different abnormal events via realistic case studies.