An Efficient Trajectory-Planning Method With a Reconfigurable Model for Any Tractor-Trailer Vehicle

Trajectory planning for a tractor-trailer vehicle (TTV) is a tough task even in a static environment because of its complicated kinematic constraints. This article aims to efficiently plan an optimal trajectory with a reconfigurable model for any TTV in the static environment. A reconfigurable modeling method is first proposed to simulate the TTV with an arbitrary number of trailers and hitching mode. Then, the planning task is formulated as a nonlinear optimal control problem. However, it is too difficult to be solved and is time-consuming to find a feasible solution, let alone the optimal one. To this end, a novel guided solving method (GSM) is proposed to solve this problem progressively with great efficiency. GSM is divided into two steps. Step 1 plans rough trajectories for the TTV, taken as an initial guess for the next step. In step 2, the problem is split into a sequence of subproblems, solved by asynchronously expanding the size of both vehicle body and obstacles, speeding up the optimization process. Simulations are carried out in several complex cases, and four competitive algorithms are selected for comparison. The results show that the proposed method outperforms others in the feasibility and efficiency of problem-solving for trajectory planning. Specifically, the proposed method is feasible in all testing cases and saves up to 81.4%, 83.2%, and 55.9% solving time compared with other three methods in the 135° turn scenario, bringing the great improvement of planning efficiency.