Multi Agent-Based Decentralized Control for Improving the Efficiency and Voltage Regulation in Smart Microgrid

Abstract

In this paper, a multi agent-based decentralized control strategy for the power management of smart grid connected microgrid (MG) is proposed to reduce distribution loss and improve the voltage regulation, which is implemented through a Multi Agent System (MAS) framework. MG consists of a two-layer infrastructure. The first layer is the electrical distribution network and the second layer is the communication network connecting the agents which could be any component in the system. Agents can collect present states of distributed generators (DGs) and loads when communication lines are added between two layers. In the proposed power management, a set of control laws for agents are derived by considering distance information between agents. Thus, output powers supplied by each DG are determined according to control laws in a way that distribution losses are reduced. Moreover, reactive power supply by DGs is managed in a way that voltage deviation from nominal value at each bus is minimized. Moreover, a theorem is proved that ensures the total output power from DGs equals the load demand of MG. In addition, control law for proposed approach is tested on a modified version of the IEEE 13 bus test system which is modelled in MATLAB-Simulink (TM). The simulation results in the modified test system are presented to demonstrate the effectiveness of the proposed control strategy.