Autonomous Vehicle Front Steering Control Computation Saving

In autonomous vehicle trajectory tracking, lane-keeping control often relies on high-order robust controllers designed with multiple performance requirements encoded via weighting functions. Such designs typically entail a significant computational cost that can overload processors already devoted to demanding tasks, such as computer vision. This work introduces an interlacing strategy for the implementation of a given robust state-space controller, with the aim of reducing its computational burden while preserving acceptable closed-loop behavior. The approach operates directly on a state-space realization and is extendable to high-order MIMO controllers, considering both diagonal (modal) and balanced realizations combined with different input–output update schemes. The method is illustrated through simulations under conditions representative of an automotive test-track circuit, indicating that substantial computational savings can be achieved at the expense of a moderate deterioration of the closed-loop response.