Sequential Convex Programming Methods for Real-Time Optimal Trajectory Planning in Autonomous Vehicle Racing

Optimization problems for trajectory planning in autonomous vehicle racing are characterized by their nonlinearity and nonconvexity. Instead of solving these optimization problems, usually a convex approximation is solved instead to achieve high update rate. We present real-time-capable model predictive control (MPC) planner based on nonlinear single-track Pacejka’s magic tire formula racing. After formulating the general nonconvex problem, we form using sequential programming (SCP). The state art convexifies track constraints linearization (SL), which method relaxing constraints. Solutions relaxed problem not guaranteed be feasible problem. propose restriction (SCR) as convexify SCR guarantees that resulting solutions show recursive feasibility restricted MPC evaluated scaled version Hockenheimring simulation. results yields faster lap times than SL, while still being real-time capable.