Application of iterative learning control in tracking a Dubin’s path in parallel parking

Intelligent parking assist systems will soon be available for most vehicles on the market. Many optimal parking trajectories and control strategies have been proposed for reverse parking. However, most of these require intensive computation, causing difficulties in practical use. This paper makes use of a classical path planning method to find the shortest parking path, and establishes the possibility of integrating iterative learning control (ILC) to exploit the capability of learning from experience to track the designed path. The effectiveness of the ILC structure is demonstrated by simulation and experiments. Tracking performance is shown to be much improved by using a simple learning control law.