This study attempts to improve the vibration isolation performance of a vehicle suspension system with a magnetorheological damper (MRD) under complex driving conditions. Structure parameter uncertainty, disturbance of the driving process, and response time delay of MRD are all addressed. Firstly, experiments of MRD were carried out in a damping force testing machine to identify the parameters of the MRD adjustable Sigmoid model by the Levenberg-Marquardt optimization algorithm. Then, the parameter identification is verified by comparing experimental and simulation data. Secondly, the state space equations of the suspension system are derived by Newton’s second law. The transfer function from the bounded disturbance input to the control output is obtained based on H∞ control theory. To make the Infinite norm of the system transfer function less than a certain value, three control strategies are proposed: variable structure control (VSC), disturbance rejection control (DRC), and delay tolerance control (DTC). Thirdly, considering these issues together to weaken the effect of disturbances on vehicle driving conditions, a fuzzy cooperative control (FCC) strategy is proposed based on the linear matrix inequality (LMI) theory. Simulation results demonstrate that FCC semi-active vehicle suspension systems conduct effective vibration isolation performance while responding to multiple external disturbances.