This paper addresses the motion-accuracy and stability bottlenecks of planar mobile robot platforms and presents an integrated control scheme for a four-wheel differential-drive chassis. Wheel odometry is used as the primary localization source, while an EKF fuses multi-source sensor measurements to suppress pose drift caused by wheel slip and speed fluctuations. At the planning level, quasi-uniform B-spline trajectories are introduced to generate smooth, curvature-continuous reference paths that also satisfy obstacle-avoidance constraints. At the execution level, a trapezoidal velocity profile imposes acceleration limits so that the chassis can start, accelerate, and decelerate smoothly. A real-time tracking controller then converts the planned trajectory into wheel-speed commands, achieving millimeter-level (RMSE of 4.97 mm) path-tracking accuracy in a constrained laboratory environment. Furthermore, an analysis of the system’s scalability in larger operational areas and highly complex environments is presented. Experiments show that the proposed framework delivers stable localization, smooth tracking, and practical obstacle-avoidance capability in indoor environments, providing a reliable motion-control and dead-reckoning basis for scalable service robots and AGVs.
Huang et al. (Tue,) studied this question.