Proximate model predictive control strategy for autonomous vehicle lateral control

Abstract

A proximate model predictive control strategy is proposed applied to autonomous vehicle lateral control. A simple and fast method to compute an approximate optimal solution is developed from the interpolation between the pre-computed optimal solutions, which is used for warm-start on-line optimization to reduce iterations in finding optimal solutions. The proposed proximate model prediction control exhibits proximate optimality in very few on-line iterations, which can become arbitrarily close to its optimality with further iterations. The results of computational and real-time experiments for autonomous vehicle lateral control demonstrate the utility of the proposed proximate model predictive control.