Abstract: In the assembly process of marine diesel engines, there are common defects such as frequent occurrence of dynamic disturbance events and insufficient monitoring of production status, resulting in the deviation of the production process from the scheduling plan and the inability to respond to disturbance events quickly. The technology is applied to the production scheduling process of the marine diesel engine assembly workshop, realizing the real-time interaction between the physical assembly workshop and the virtual assembly workshop, which can quickly respond to disturbance events and reduce the deviation of the scheduling plan. A production scheduling framework of marine diesel engine assembly workshop based on digital twin is constructed, and the assembly scheduling problem is described and modeled. The dynamic scheduling process is expounded in detail, and a multi-objective discrete ant lion optimization algorithm based on Pareto rule is proposed, which optimizes the two scheduling objectives of maximum completion time and assembly team load, and ensures the real-time and accuracy of scheduling.Finally, the validity of the scheduling mode is verified by deployment in a certain type of diesel engine assembly workshop.