Optimal, stable switching between arcs during low-speed Ackerman path tracking with rate-limited steering

Many applications require an automated front-steered vehicle to follow a desired path. Examples include autonomous mining trucks, tractors, and military target vehicles. We assume the path consists of a sequence of circular arcs, which are tangent where they meet. In previous work we showed that when a linear lag was used to model the steering servo, the time constant directly determined the optimal time to switch to the next arc in the sequence. In this paper we consider a more realistic steering servo model in which the rate is limited. We find that the optimal way to switch from one arc to the next involves performing half of the steering on the previous arc and half on the arc which follows. Experiments with an automated tractor verify the analysis.