Abstract: In order to obtain the accurate load sharing design conditions of the coaxial surface gear shunt system, the gear matching and multi-objective optimization design methods were studied. Using the centralized mass parameter method, a 45-degree-of-freedom bending-torsional coupling dynamic model of the system was established considering the manufacturing error, torsional stiffness of the double coupling, meshing stiffness, damping and other factors. Considering the factors of system load sharing and lightweight, a design example of gear matching that satisfies the dynamic load sharing characteristics of the system is solved, and a multi-objective optimization design model considering the load sharing characteristics and quality is established, and the non-dominant sequencing genetic algorithm is used to solve the problem. The results show that the manufacturing error is the biggest factor affecting the dynamic load sharing characteristics of the system, and a good dynamic load sharing layout can be obtained through the establishment of gear matching conditions, which improves the load sharing performance by 5.4%. The optimized system load sharing performance is 3.3% and 8.5% higher than that before and after the establishment of gear matching conditions, respectively. The research results can provide a theoretical basis for the design of system dynamics and stability.