Aiming at the problems of complex components with many processing features and various processing techniques, a method for optimizing operation sequence based on improved genetic algorithm was proposed. In the method, a three-dimensional process matrix was used to store and express the manufacturing information, a process constraint matrix was established as a constraint condition, and a mathematical model of operation sequence optimization was established with the lowest cost as the goal; then the genes were coded hierarchically, the corresponding crossover and mutation operators were selected for different decision-making levels, and the selection level coding was introduced to realize the dynamic decision-making of processing methods and manufacturing resources in the processing workshop; the box parts were used as an example to verify. The results show that by using the improved genetic algorithm, the population quality is improved, the convergence speed of the algorithm is accelerated, and a processing sequence is obtained that meets the process constraints, which verifies the effectiveness of the improved method.