The object of this study is a mobile reconfigurable robot with a hybrid wheel-track locomotion system capable of transforming the kinematic structure of its chassis to overcome high vertical obstacles. The relevance of this work is driven by the need to ensure the reliable operation of transformation mechanisms while using low-cost components. This paper addresses the problem of synthesizing a control system that is robust to hardware limitations (dead zones, backlash) and actuator nonlinearities. A novel active vertical scanning method ("Stop-and-Scan") is proposed, which allows for compensating for the physical limitations of standard ultrasonic sensors, such as wide beam width and specular reflection. The method is based on analyzing the distance gradient during the vertical displacement of the sensor, ensuring the accurate determination of obstacle geometric parameters. Particular attention is paid to the control of linear actuators with worm drives, which are characterized by significant dry friction. A control law using high gains in saturation mode has been synthesized, allowing for the compensation of the actuator dead zone without compromising stability. The results of mathematical modeling and field tests confirm that the proposed approach minimizes steady-state positioning error and improves the overall trafficability of the robot.
Que et al. (Tue,) studied this question.