Abstract: In order to realize the path tracking of the on-off valve-controlled crawler mobile platform that cannot be closed-loop speed regulated, considering the low speed of the platform and the ability to turn in place, a bang-bang path tracking control algorithm with a boundary layer without the bottom wheel speed control is proposed based on pure pursuit algorithm. Based on the geometric and kinematic relations, a target point tracking model of crawler mobile platform is established. Taking the heading angle deviation as the switching function, the method controls the forward and reverse rotation of the crawlers on both sides through the bang-bang algorithm, and introduces the boundary layer thickness parameter to reduce the switching frequency of the solenoid valve. According to the Lyapunov stability theory, it is proved that all states of the system outside the boundary layer are stable, but the heading angle deviation inside the boundary layer is unstable at the origin. Moreover, the smaller the look ahead distance is, the faster the heading angle deviation diverges. Based on the Matlab/Simulink and Recurdyn joint simulation platform, the rectangular target path is selected, and the simulation verifies the path pursuit effect of the bang-bang algorithm with different look ahead distances, and compares it with the pure pursuit algorithm of closed-loop speed regulation. A prototype test platform is built to conduct ground tests on the bang-bang path tracking control algorithm. The simulation and test results show that the bang-bang path tracking control algorithm can control the on-off valve-controlled crawler mobile platform to track straight or broken-line paths. Moreover, the tracking accuracy of the algorithm at corners is higher than that of the pure pursuit algorithm path of closed-loop speed regulation. The steady-state error is less than 9 cm, which meets the requirements of autonomous navigation walking of coal mine robots.