Control of tractor-trailer wheeled robots considering self-collision effect and actuator saturation limitations. (15th July 2019)

• Record Type: Journal Article
• Authors: Kassaeiyan, Pouya; Tarvirdizadeh, Bahram; Alipour, Khalil
• Is Part Of: Mechanical systems and signal processing. Volume 127(2019)
• Page Start: 388
• Page End: 411
• Publication Date: 2019-07-15
• Subjects: Model predictive control -- Nonlinear model predictive control -- Trajectory tracking -- Obstacle avoidance -- Tractor-trailer wheeled robot; Structural dynamics -- Periodicals; Vibration -- Periodicals; Constructions -- Dynamique -- Périodiques; Vibration -- Périodiques; Structural dynamics; Vibration; Periodicals; 621
• DOI: 10.1016/j.ymssp.2019.03.016
• Languages: English
• ISSNs: 0888-3270

Abstract

Highlights: Motion of a Tractor-Trailer Wheeled Robot (TTWR) is controlled. The motion control problem is addressed considering the effect of self-collision. Actuators saturation limits are considered during the controller design process. Concurrent trajectory tracking and obstacle avoidance issues are addressed. Suggested model predictive control algorithm for TTWRs is evaluated in experiments. Abstract: A Tractor-Trailer Wheeled Robot (TTWR) is a type of multiplatform robotic systems which contains a tractor towing a (multi) trailer(s). These kinds of mobile robots are nonlinear and underactuated systems exposed to nonholonomic constraints, assuming the pure-rolling condition of the wheels. TTWRs have many applications for transporting various payloads, public transportation, etc. The collision between tractor and trailer not only damages the robot but also may cause instability problems in the system. The self-collision avoidance issue will be noticed in TTWR for the first time in this study and is one of the contributions of this research. To this end, after deriving the kinematics model of the system, Linear Model Predictive Controller (LMPC) and Nonlinear Model Predictive Controller (NMPC) are developed to trajectory tracking control of the system. Considering actuators saturation bounds, which is an essential issue in real-world applications, in the control design process is the other contribution of this study. The developed controllers produce control signals in the feasible bounds of the actuators and also avoid self-collision systematically. Then, trajectory tracking and obstacle avoidance problems are solved simultaneously using NMPC capability. Finally, robustness and effectiveness of the proposed controllers and comparison of the performance of LMPC and NMPC are studied by real-world experimental implementations.