Abstract: In order to avoid the traditional rat swarm optimization（RSO） algorithm from converging slowly and falling into local optimum, we propose a rat swarm optimization algorithm（HRSO） that integrating Circle chaos mapping and vertical and horizontal crossover strategies. The HRSO is applied to a microgrid optimal dispatching model with the objective of minimizing the total generation cost as considering security and stability. The results show that HRSO improves the stability by 12.45% and the convergence speed by 28.17% compared with RSO under four test functions, at the same time the total generation cost of microgrid is reduced by about 8.43%, in the microgrid optimal dispatch model.