A multi-mode coupling drive system has been designed and controlled to improve the dynamic characteristics and fuel economy of plug-in hybrid electric vehicles, which also can make full use of theconfigured superiority of centralized drive systems and distributed drive systems and avoid theirstructural defects. The configuration evolution process, working mechanism and drive modes of themulti-mode coupling drive system are introduced. The powertrain model is established for the targetvehicle. Based on Charge Depleting-Charge Sustaining energy management strategy, an Electric VehicleCharge Sustaining energy management strategy is developed. The Improved Real-valued Genetic Algorithm is used to optimize the system structural and control parameters, it can help prioritize the drivemodes which are based on the proposed energy management strategy. While ensuring the vehicle dynamics, the best energy allocation is achieved. The results show that comparing with a series distributeddrive hybrid system and the intelligent Multi-Mode Drive (i-MMD) hybrid system under the NEDCcondition, the 100-km fuel consumption of the optimized multi-mode coupling drive system is reducedby 16.52% and 15.40%. Respectively, it further proves the superiority of the drive system in improvingvehicle economy.