Abstract: Microgrid is an effective way to integrate the distributed generations, and the optimal operation of microgrid has become one of the important topics in the research of microgrid. The optimal operation of microgrid is modelled with a consideration of multiple operation indicators such as the economic cost, environmental cost, network loss and node voltage fluctuation, and a balance of the interests of various stakeholders. The elite reverse learning strategy and the worst particle exclusion method are introduced into particle swarm optimization algorithm (Particle Swarm Optimization, PSO) to solve the multi-objective and multi-constraint optimal operation problems of microgrid. In the process of searching, chaotic disturbance is made on the existing optimal particle to enhance the local searching ability, and to improve the ability of particle to jump out of local optimal solution. Under same conditions, the optimal operation model of the microgrid is solved using the original algorithm and the improved algorithm respectively, and the superiority of the improved algorithm is verified by comparing the solution results.

Key words: microgrid, operation optimization, particle swarm optimization, elite reverse learning strategy, the worst particle exclusion, chaotic disturbance