This thesis addresses a four–wheel independent steering and driving (4WISD) mobile robot, in which the steering and driving of all four wheels are independently actuated, and proposes a velocity controller that accurately tracks the commanded body v...
This thesis addresses a four–wheel independent steering and driving (4WISD) mobile robot, in which the steering and driving of all four wheels are independently actuated, and proposes a velocity controller that accurately tracks the commanded body velocity (translational and rotational). A 4WISD mobile robot is an over–actuated system with separate steering and driving motors for each wheel, and thus has many degrees of freedom with strongly interconnected wheel motions, which calls for sophisticated control design. If these characteristics are not properly taken into account, excessive tire slip and actuator overload can easily occur. In particular, when the four wheels do not share a common instantaneous center of rotation (instantaneous center of rotation, single ICR), the vehicle cannot realize an ideal rigid–body motion, so the key challenge in controller design is how to satisfy and maintain the single–ICR condition.
To tackle this problem, the thesis first derives a 4WISD kinematic model that describes the relationship between the body–frame translational/rotational velocity and each wheel velocity. The wheel velocities are then decomposed into a component that satisfies the common ICR condition and a component that violates it; the latter is defined as a slip–vector. Using this decomposition, trajectory tracking error, input variation, and the degree of ICR violation are quantified within a unified cost function, and by penalizing the slip–vector the vehicle motion is driven to satisfy the single–ICR condition. The proposed model predictive controller (MPC) is thus designed so that trajectory tracking performance, reduction of actuator effort, and suppression of slip are achieved simultaneously. Step–response, pure–rotation, and trajectory–tracking experiments conducted on a 4WISD mobile robot platform demonstrate that the proposed controller satisfies the single–ICR condition under various driving scenarios, while ensuring stable and efficient tracking of the commanded velocity and significantly reducing slip compared with a conventional ROS–based velocity controller.