Real - Time Trajectory and Velocity Planning for Autonomous Vehicles

Path planning algorithm integrated with a velocity profile generation-based navigation system is one of the most important aspects an autonomous driving system. In this paper, real-time path solution to obtain feasible and collision-free trajectory proposed for navigating car on virtual highway. This achieved by designing incorporate planner finding optimal path, ensuring safe comfortable motion along obtained path. The was validated Unity 3D Highway-Simulated Environment practical while maintaining acceleration constraints. vehicle drives at maximum specified until interrupted vehicular traffic, whereas then, planner, based various constraints provided simulator using µWebSockets, decides either decelerate or shift more secure lane. Subsequently, splinebased generation results in continuous smooth trajectories. employs analytical method trapezoidal generate velocities traveling precomputed To provide control, s-like considered that uses cubic spline generating ramp-up ramp-down portions curve. constraints, which are derived from road limitations physical systems, explicitly considered. Depending upon these higher module requirements (e.g., velocity, stopping), appropriate segment deployed. profiles all use-cases generated verified graphically.