Real-time optimal reactive power dispatch using multi-agent technique

Abstract

This paper presents a multi-agent systems (MAS) based approach and adopts a coordinative optimization method to realize real-time global optimal reactive power dispatch (ORPD) for large-scale power systems. The ORPD problem is one of the combinatorial optimization problems with characteristics of multi-objective, -uncertainty, -restriction, -extremum and non-linear property. It becomes more and more complicated with the increasing of power system scale. Distributed solving method is promising to this problem. The main idea of MAS is to generate approximate solutions to very complex problems, such as ORPD, by distributing them to autonomous problem solvers, namely agents. These agents solve local ORPD problems by themselves asynchronously and localize dispatch and control. By coordinating in a communication environment, a solution could be found to the global ORPD problem rapidly. Test results show that the proposed MAS based approach performed well in real-time global ORPD of large-scale power systems.