Taking the seat vibration system of a certain type of engineering vehicle as an object, using the characteristics of adaptive network-based fuzzy inference system（ANFIS） self-learning and fuzzy reasoning, a semiactive control method of seat suspension vibration based on ANFIS was proposed, which solved the problem of excessive seat vibration in the working process of engineering vehicles. Firstly, the vibration process of the seat suspension of engineering vehicles was analyzed, and the dynamic model of the seat suspension was established. Then, the fuzzy neural control system of seat suspension was designed by adding feed-forward and feedback links. Finally, the control effects of different vibration excitation frequencies, different controller control objects and the input of measured seat signals were observed respectively. The results show that the fuzzy neural control system reduces the vibration energy and frequency of the seat, and the root mean square value of the seat vibration displacement is 52% to 66% of the passive suspension seat.