A Multi-steering Trailer System: Conversion into Chained Form Using Dyanmic Feedback

In this paper, the kinematic model of an autonomous mobile robot system consisting of a chain of steerable cars and passive trailers with axle-to-axle hitching is examined and converted into a multi-input chained form using dynamic state feedback. Some of the methods which have been proposed for steering two-input chained form systems are generalized to multi-chained systems, and then applied to two example multi-steering trailer systems. 1. INTRODUCTION In this paper the motion planning problem is solved for a multi-steering trailer system; that is, a car-like mobile robot pulling a combination of n passive trailers and m?1 steerable car-like robots. The controls available to the system are the velocity of the lead car and the steering velocities of all m car-like robots. This system can be thought of as a generalization of a standard n-trailer system, the motion planning problem for which was considered and solved in (SSrdalen, 1993; Tilbury et al., 1993b). Mobile robot systems of this kind are of interest in practical applications; part of the motivation for this work came from work on the re truck (Bushnell the authors have been told anecdotally about the construction of such n-trailer systems with multi-steering for use in nuclear environments (Canudas de Wit, 1993) and also for baggage and cargo handling (Giralt, 1993). These mobile robot systems can be modeled as having one constraint on each axle; namely, that the wheels are allowed to roll but not to slip. These constraints are nonholonomic or non-integrable and do not reduce the reachable conng-uration space of the mobile robot. Although the present system appears at rst glance to be a straightforward extension of the systems considered earlier, the main motivation in writing this paper is to exhibit its added richness and complexity. As before, the motion planning planning problem is solved by rst converting the system into chained form. Unlike the previous examples, this system will require states to be added to the system before the transformation