Abstract: Aiming at the balance optimization problem of group product parallel production line with product correlation, a phased research idea was adopted to propose an optimization scheme for process and employee assignment by successively considering process complexity and employee skill level differences. First, by taking the process precedence relation and equipment attributes as the constraints and the production cycle time, time balance index and complexity smoothing coefficient as the optimization objectives, the optimization model was established, so that the processes can be reasonably allocated to the workstations. Then, according to the characteristics of group products and the skill requirements of each workstation, the employee assignment model was established to match employees with different skills to workstations, so as to reduce the difference in the beat between parallel production lines and achieve balanced production of group products. Genetic algorithm and quasi-annealing algorithm were used to solve the process assignment problem and the employee assignment problem respectively, and the effectiveness of the optimization idea and method was verified with the example.