An all-terrain-controller for over-actuated wheeled mobile robots with feedforward and optimization-based control allocation

Model-based chassis control system for an over-actuated planetary exploration rover

“…The rover has three passive bogies and six active steering joints, all of which are rotational degrees of freedom (DoF), leading to a vector q ∈ ℝ (9×1) of all DoF angles. In [5], the homogeneous transformation matrices, including the position vectors r r c i |r from the rover CoM to the contact points, were derived with the help of the Denavit-Hartenberg (DH) convention. The relative velocity of the ith contact frame can be computed by differentiation as…”

“…with the longitudinal F c i ,x and lateral contact force F c i ,y from equations ( 15) and ( 16), the steering resistance torque T res,z and the wheel inertia about the drive and steering axis I w i ,y and I w i ,z , respectively. Note that the output y w i contains the x-and y-force resolved in the non-steered sframe in order to comply with the contact forces that are used in the force torque balance in equation (5) with…”

Performance analysis of control allocation using data‐driven integral quadratic constraints

“…The control allocation scheme typically exploits the additional degrees of freedom from over-actuation to handle actuator constraints or minimize some performance criteria like energy consumption. This makes control allocation very attractive in various application fields such as road and all-terrain vehicles, [1][2][3] aerospace, [4][5][6] marine vehicles, 7,8 manipulators, 9 or bio-mechanical muscles. 10 However, if the virtual inputs cannot be adequately realized, for example, due to saturation of multiple actuators, closed-loop performance is usually degraded and in the worst case, stability issues may even arise.…”