Skip to content

Planning Library for Autonomous Vehicles

Overview

The Planning Library provides advanced tools for trajectory planning and optimization in autonomous vehicles. It includes components for lane following, safety corridor planning, trust region optimization, and utilities to convert high-level routes into executable trajectories. Designed for efficiency, the library integrates optimization techniques and vehicle dynamics models.

Features

  • Lane Following:
  • Implements trajectory planning to maintain a target route.
  • Safety Corridor Planning:
  • Plans trajectories within predefined safety corridors.
  • Trajectory Optimization:
  • Leverages Model Predictive Control (MPC) with Nonlinear Programming for optimal trajectory generation.
  • Trust Region Optimization:
  • Solves cooperative multi-agent planning problems using trust region solvers.
  • Planning Helpers:
  • Utilities for waypoint filtering, point shifting, and trajectory conversion.

Included Modules

Lane Follow Planner

File: lane_follow_planner.hpp - Implements lane-following behavior by generating trajectories that align with the centerline of a given route. - Considers vehicle dynamics, curvature, and control limits.

OptiNLC Trajectory Optimizer

File: optinlc_trajectory_optimizer.hpp - Performs trajectory optimization using Model Predictive Control (MPC). - Optimizes trajectories by solving nonlinear constrained optimization problems. - Features customizable weights for objectives like lateral error, heading error, and steering input.

OptiNLC Trajectory Planner

File: optinlc_trajectory_planner.hpp - Converts high-level routes into optimized trajectories using MPC. - Supports curvature-based velocity profiling and piecewise polynomial representations of routes.

Planning Helpers

File: planning_helpers.hpp - Provides utility functions for planning, such as: - Filtering waypoints in front of the vehicle. - Shifting points laterally for safety or desired offsets. - Converting waypoints into executable trajectories.

Safety Corridor Planner

File: safety_corridor_planner.hpp - Plans trajectories within a defined safety corridor. - Uses polynomial representations for corridor boundaries. - Implements constraints to ensure the vehicle remains within safe zones.

Trust Region Solver

File: trust_region_solver.hpp - Solves multi-agent planning problems using trust region optimization. - Models each agent's trajectory and controls, ensuring convergence through gradient-based optimization.