In order to solve the problem of restrictive application of active suspension in electric vehicles due to energy consumption, using permanent magnet(PM) linear motor as the actuator of active suspension system, a dynamic model was developed to investigate the relationship between vehicle's dynamic performance and energy regenerative capability. The active suspension LQG controller was designed based on the optimal control theory, and the controller design parameters were optimized by analytic hierarchy process(AHP) and particle swarm optimization(PSO), which improved the vehicle dynamic performance and energy regenerative power. In order to achieve the state identification and mode switching, a novel multi-mode switch control strategy was proposed. The innovation of the proposed control strategy is the introduction of the comfort factors which depend on the driver's choice and the detailed identification of the driving state of the vehicle, so as to realize the strategy switching under different modes. The simulation results show that the proposed multi-mode switch control strategy is significantly better than conventional active suspension control mode, and achieves a comprehensive and intelligent improvement of dynamic performance and energy regenerative capability in vehicle. This study provides guidance for the suspension feed-energy control strategy.