Abstract: An inaccurate control input matrix affects the control performance in the active disturbance rejection control (ADRC) system of a substation inspection robot. In this study, an ADRC scheme for an inspection robot based on a data-driven method is proposed to address this problem. First, the discrete time dynamic model of the inspection robot is established. Second, an extended state observer is used to observe parameter errors, unmodeled dynamics, friction, and other uncertainties. A nonlinear ADRC controller is devised on the basis of the dynamic model of the inspection robot. Then, the data-driven model-free adaptive control (MFAC) method is utilized for trajectory tracking control by considering the ADRC system and the robot system as a whole. Results reveal that the proposed method effectively combines the ADRC and MFAC methods. It also achieves the working mechanism with complementary advantages of the two control methods. Lastly, the effectiveness of the proposed method against traditional ADRC and MFAC methods is verified via simulations.