Abstract: Traction control system (TCS) based on road adhesion coefficient estimation is designed for the rear drive tractor. Firstly , to track the time-varying nonlinear system in the classical Kalman filter, a fuzzy control theory and an attenuated memory filter are introduced into untraced Kalman filter, and the fuzzy forgetting factor unscented Kalman filter (FFUKF) is proposed to estimate road adhesion coefficient, so as to improve the tracking performance of the algorithm. Traction control includes torque control and braking control. Then, for the TCS torque control, the road adhesion coefficient and the vehicle acceleration when the slip rate of the driving wheel is near the target slip rate are used to calculate the target base torque, respectively. According to the vehicle state and chattering parameters, the extension set is divided into the classical domain, extension domain and non-domain with the extension control theory, and the dynamic weight coefficient is obtained by the correlation function of extension set. The base torque is designed by extension fusion of the target base torques calculated by the above methods. Next, the error between the actual slip rate and the target slip rate is used as the input, and the target feedback torque is obtained by using the fuzzy self-tuning PI controller. In terms of braking control, the PI pressure control and adhesion difference pressure are designed on two typical road surfaces, respectively. Finally, the test results show that the proposed FFUKF algorithm can track the tire-road friction coefficient more quickly, and the proposed TCS control strategy can effectively restrain the excessive slip of the driving wheels to control the slip rate of drive wheel in the best range, so that improving vehicle dynamic performance significantly.