Real‐time resilient microgrid power management based on multi‐agent systems with price forecast

Abstract

Abstract Microgrids have emerged to diversify conventional electric generation using small‐scale distributed generation. Large efforts have been put into designing control strategies to optimise the power schedules of microgrids, however, verification that such control systems also are reliable in terms of stability during normal operation and fault conditions is needed. This study presents a hierarchical distributed control system that fulfils these conditions for an AC microgrid. The stability maintained by proposed controller, considering the large signal model, is analysed with the use of Lyapunov's direct method. Resilient control distribution is achieved by the implementation of suitable forecast models and fault‐tolerance mechanisms to avoid single points of failure. The resilience of the control system is verified with the use of graph theory. The stable and resilient operation of the proposed control system is tested by a real‐time microgrid model implemented with an OPAL‐RT real‐time simulator, combined with a communication network built with Raspberry Pis, testing the control system presented under normal and faulty conditions. Simulation results show a stable operation in terms of voltage and frequency in both conditions, resilient operation is shown for the faulty condition case. Additionally, cost minimisation performance is included to validate optimal power management capabilities.

Keywords

• AC microgrid
• artificial neural network
• auto‐regression
• Lyapunov stability
• Markov chain Monte Carlo
• multi‐agent system