The development of active distribution systemScreates the conditionsSfor theShigh penetration ofSdistributed generation. For the positions and capacitiesSproblems ofSvariousStypes of distributed generationSandSenergy storageSsystem in activeSdistributionSsystem, aSmulti-objective optimization model is established, in which the economic costs of active distribution system,S voltage qualitySand the emissions of CO2 is considered. The traditional NSGA-SII algorithmSis improved. Opposition-based learning is introducedSin initial population , and the dominated individuals of the population are selected in a certain proportion to increase the population diversity and improve searching ability of the algorithm. The method of adjusting weight coefficient is used to screen the optional solution out from the Pareto optimal solution set. Finally, by a 33-bus network system the reasonableness and feasibility of the proposed model and algorithm are proved.