In order to reduce therework iteration in the product development process, the optimization analysis of product design process based on reinforcement learning (multi-objective process optimization genetic algorithm based on design structure matrix (DSM) theory) is proposed. By optimizing the task execution sequence, therework in the product development process can be reduced to compress the progress and reduce the cost. The optimization algorithm is an improved genetic (GA) algorithm, in which time and cost are considered in the fitness function. In the selection, crossover and mutation operators, the strategy of maintaining optimal solution is adopted. The simulation results show that the optimization algorithm can reduce the development time by 30% ∼ 40% and the cost by 7% ∼ 20% for product development projects with high task coupling.