Due to the fact that the traditional fuzzy controller of the suspension system expressively depends on expert experience, and the fact that the control strategy and damper model of the suspension system are verified separately, we put forward a novel control strategy for semi-active suspension systems based on deep neural fuzzy system. The dynamic model of semi-active suspension systems is established, and by using adaptive neuro fuzzy inference system, a variable universe Takagi-Sugeno fuzzy controller is designed. Then, combining the bench experimental data and adaptive neuro fuzzy inference system, a non-parametric model of continuous damping control damper is first proposed. Finally, a semi-active suspension controller based on deep neural fuzzy system is constructed by using serial ideas. The effectiveness of the proposed deep neural fuzzy system control strategy is verified by simulation experiments under different working conditions. The results show that the proposed deep neural fuzzy system control strategy has strong adaptability to various working conditions, which can effectively improve the ride comfort and handling stability of vehicles.