The optimal operation of microgrid can effectively reduce the operating cost and improve the efficiency of energy utilization. Most of the multi-objective energy optimization models of microgrid are solved by transforming the multi-objective problems into single objective problems. Considering the demand response, the dynamic optimization model was established for the grid connected microgrid. To solve the multi-objective optimization model of microgrid, the weight of each objective was determined firstly using analytic hierarchy process. Then the two important objectives were optimized using the non-dominant sorting genetic algorithm(NSGA-Ⅱ).The optimal solutions obtained by optimizations were taken as the candidate solutions. Then the objective function values of the candidate solutions were evaluated .The final solution was determined by the method of multiple attribute decision making based on evidence reasoning. The results show the effectiveness of the proposed method. It provides a new way to solve the problem of multi-objective optimal operation of microgrid.