Tikhonov Regularization Based Control Allocation for Underactuated Input-Affine Systems

This paper proposes a novel control allocation method for underactuated input-affine nonlinear mechanical systems, based on the Tikhonov regularization. A hardware-in-the-loop experiment is conducted using an unmanned aerial manipulator to corroborate the effectiveness of the proposed method. The results demonstrate that, in comparison with the classic Moore-Penrose pseudo-inverse control allocation method, the proposed Tikhonov regularization-based control allocation is computationally faster and attenuates additive noise terms within the generalized input vector. Besides, it reduces the residual error between the desired and applied generalized input when the input coupling matrix is ill-posed.