Abstract: The stochastic and time-sequence characteristics of the output of distributed generation such as wind power, solar photovoltaic, and the load have brought great challenges to the optimal allocation of distributed generation in the distribution network. Therefore, based on the one year hourly data, the probability distribution model of wind speed, light intensity and load of each hour are established using multi-scenario and time period partition approach. Taking the comprehensive benefit including distribution network operators' revenue and modified voltage quality improvement index as the objective function, a distributed generation planning model is presented based on time-sequence characteristics using chance-constrained programming for each time period. The genetic algorithm is applied to solve the model. The simulation results of the IEEE33 bus distribution system verify the effectiveness of the proposed model, and show that a reasonable distributed generation allocation scheme can achieve significant economical and environmental benefits and improve voltage quality.