Abstract: In order to cope with the challenges brought by the random fluctuations of the source-end renewable energy and load demand on the operation scheduling and capacity allocation of integrated energy production unit (IEPU), a two-stage stochastic optimization method was proposed in this paper. First, in the bottom operation optimization problem, this paper proposed a minimum cost solution method based on mixed integer linear programming (MILP) by establishing the equipment models and constraints. Secondly, Monte Carlo simulation was used to generate multiple random scenarios to determine the cost expectation of the system under a given capacity configuration condition. Finally, in the top-level capacity allocation optimization problem, this paper took the system capacity as the decision variable, and used genetic algorithm to call Monte Carlo simulation and MILP operation optimization algorithm to realize the optimal capacity allocation that minimized the whole life cycle cost of IEPU system. The optimization results show that the access of gas storage in the bottom operation optimization reduces the light curtailment and carbon emissions by 5.49% and 0.35%, respectively. The capacity of power equipment in the top level considering the uncertainty of source and load will increase by about 20%, which is closer to reality and verifies the effectiveness of the proposed method. Combined with parameter sensitivity analysis, it provides reference for the large-scale design of IEPU system.