This paper, the impact of time-delay feedback control on vibration energy collection and damping performance is discussed. A time-delay control system of active suspension based on a 1/4 vehicle nonlinear model is designed. The linear motor is used as the actuator of active suspension. The variable learning factor particle swarm optimization algorithm is used to optimize the optimal control parameters of active suspension with time delay. The characteristic root clustering method is proposed to analyze the stability of the active suspension control system with time delay. Finally, the energy recovery and vibration reduction of passive suspension, active suspension without time-delay, and active suspension with time-delay are simulated. The simulation results show that the energy-feeding efficiency of the active suspension with time-delay control under harmonic and random excitation is 114.54% and 20.09% higher than that of the active suspension without time-delay control. Finally, a test bench is built to verify the correctness of the simulation results. The experimental and simulation results error is within 15%, which meets the engineering requirements. The proposed method can significantly improve the energy collection capacity.