Generating Optimized and Smooth Path for Two-Body Vehicle Reverse Motion: Initial Guess for NMPC Solver

This paper presents an optimization-based planning and control for intricate maneuvers of two-body vehicle in limited space. Reverse motion planning for n-body vehicles operating in restricted space environments poses significant challenges due to complex kinematics and limited maneuverability. For example, a truck-trailer system may need to navigate between obstacles to enter a loading zone or parallel parking in parking lot and complex maneuvers of connected rovers for moving in unknown terrains in interplanetary exploration projects. The presented approach finds global path-planning solutions for n-body systems using kinematics-based planner along with the optimization-based path smoothing and implementation of multistage nonlinear model predictive control to generate a smooth and feasible path for two-body vehicle. In the proposed method, the generated smooth path can be used as is, or as an initial guess for an NMPC solver, helping to avoid local minima in the nonconvex problem space.