Abstract: Wheel-legged robots are inevitably in contact with the environment during movement. Due to the fixed parameters of the target impedance model, the traditional method of impedance isolation force control often puts up a poor robustness under an unknown information environment (stiffness and position) or time-varying. Considering traditional impedance control, an adaptive impedance controller was designed based on Lyapunov stability theorem. The method was arranged to indirectly adjust the impedance parameters with position compensation, make the steady-state error of the system to be zero and improve the adaptability to the unknown and changeable environment. Vibration isolation force control experiments were carried out for the robot under the condition of moving on different deceleration belts. Results show the superiority of the adaptive impedance control.