Layered management and hybrid control strategy based on hybrid automata and random forest for microgrid

Abstract

A layered management and hybrid control strategy based on hybrid automata and random forest for the microgrid is proposed in this study. It is used to realise energy management and multi-modal coordinated switching of microgrid system to ensure high security, stability and high efficiency. Management and control strategies are implemented through a three-layer microgrid architecture based on multi-agent system theory. An energy management strategy is designed for the microgrid system in the upper layer agent structure. A coordinated control strategy is designed in the middle layer agent structure. A local control strategy is designed into the bottom layer agent structure. In the middle layer coordinated control strategy, the hybrid automata is used to characterise the complex logical features of each underlying unit and trigger the switching of working modes according to the condition change. The random forest algorithm is used to analyse the running state and reduce the number of switching. Finally, the simulation results verify the effectiveness of the proposed control strategy in different scenarios.