Multiagent-Based Autonomous Energy Management System With Self-Healing Capabilities for a Microgrid

摘要

Due to the inherent intermittent nature of renewable energy sources (RES), the control and management of microgrids become complex and require modern control techniques and management strategies to cope up with the changes in the dynamics of the system. This paper focuses on improving the dynamic performance of a microgrid by developing different energy management and control strategies employing a four-layer multiagent concept. The first layer of the proposed system is a forecasting and estimation layer and is responsible for providing the real-time forecasts for RES output power, load, and battery state of charge. The second layer is a control and action layer, assigned with the duties of taking the control decisions and actions accordingly by collecting information from different agents deployed in the system. The third layer is a real-time monitoring and control layer that provides the real-time information of the microgrid to the second layer. The fourth layer is a fault detection and action layer that facilitates the self-healing capability in the proposed microgrid. The performance and the applicability of the proposed model is tested by simulating the faults at different locations in the system.