What algorithms are used for path planning in autonomous vehicles?

Path planning generates safe, feasible trajectories. Algorithms include: A* and Dijkstra - graph search for global route planning; RRT (Rapidly-exploring Random Trees) - sampling-based for complex environments; Polynomial trajectory generation - smooth curves meeting kinematic constraints; Lattice planners - precomputed motion primitives combined on-the-fly; Model Predictive Control (MPC) - optimization-based trajectory generation considering vehicle dynamics. Planning operates at multiple levels: Route planning (minutes, km-scale), behavioral planning (seconds, maneuver decisions), and motion planning (milliseconds, trajectory execution). Planners must consider vehicle dynamics, comfort constraints, traffic rules, and sensor uncertainty.