Skip to content

Behavioural planning and trajectory generation in a highway scenario

Notifications You must be signed in to change notification settings

bdschrisk/CarND-Path-Planning-P1

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

11 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

CarND-Path-Planning-Project

Self-Driving Car Engineer Nanodegree Program

Simulator. You can download the Term3 Simulator BETA which contains the Path Planning Project from the releases tab.

Overview

In this project, a model is constructed to allow a car to drive on a simulated highway. A number of factors are at play while the vehicle is in motion, including; keeping close to the speed limit and avoiding collisions with other cars.

Originally in the project I intended to use a Hybrid A* planner to find the shortest path in a discretised state or snapshot of the highway. Following research into highway planning and control, this would have been difficult and costly with compute resources. Instead, I have opted for a physics based approach. Using the equations of motion, the car is able to perform a number of key features:

  • Brake within a set safety distance while minimising jerk
  • Accelerate using an S-curve akin to real life systems
  • Detect potential collisions within a certain timeframe

To facilitate the acceleration and braking, a spline was contructed for discrete velocities. Acceleration is performed using an S curve from zero velocity and max acceleration, to max velocity and zero acceleration. This works for any velocity "moment" and generates the required acceleration when the target velocity is higher than the current. Conversely, when braking, equations of motion are used to calculate the required braking distance and time required to pull up to a safe travelling distance. Similarly, a deceleration curve is generated using these negative velocity moments.

The trajectory planner then solves for each waypoint generated, using these acceleration curves for maintaining safe driving.

Collision detection is performed by iteratively projceting each state forward in time and calculating possible collisions. These are then used (or lack-of) to perform lane shifts.

Finally, when performing a lane shift, a cross product of the transition lane and the current is used to find the "average" velocity. This aids in speed matching when changing lanes.

The map of the highway is in data/highway_map.txt

Each waypoint in the list contains [x,y,s,dx,dy] values. x and y are the waypoint's map coordinate position, the s value is the distance along the road to get to that waypoint in meters, the dx and dy values define the unit normal vector pointing outward of the highway loop.

The highway's waypoints loop around so the frenet s value, distance along the road, goes from 0 to 6945.554.

Basic Build Instructions

  1. Clone this repo.
  2. Make a build directory: mkdir build && cd build
  3. Compile: cmake .. && make
  4. Run it: ./path_planning.

Here is the data provided from the Simulator to the C++ Program

Main car's localization Data (No Noise)

["x"] The car's x position in map coordinates

["y"] The car's y position in map coordinates

["s"] The car's s position in frenet coordinates

["d"] The car's d position in frenet coordinates

["yaw"] The car's yaw angle in the map

["speed"] The car's speed in MPH

Previous path data given to the Planner

//Note: Return the previous list but with processed points removed, can be a nice tool to show how far along the path has processed since last time.

["previous_path_x"] The previous list of x points previously given to the simulator

["previous_path_y"] The previous list of y points previously given to the simulator

Previous path's end s and d values

["end_path_s"] The previous list's last point's frenet s value

["end_path_d"] The previous list's last point's frenet d value

Sensor Fusion Data, a list of all other car's attributes on the same side of the road. (No Noise)

["sensor_fusion"] A 2d vector of cars and then that car's [car's unique ID, car's x position in map coordinates, car's y position in map coordinates, car's x velocity in m/s, car's y velocity in m/s, car's s position in frenet coordinates, car's d position in frenet coordinates.

Details

  1. The car uses a perfect controller and will visit every (x,y) point it recieves in the list every .02 seconds. The units for the (x,y) points are in meters and the spacing of the points determines the speed of the car. The vector going from a point to the next point in the list dictates the angle of the car. Acceleration both in the tangential and normal directions is measured along with the jerk, the rate of change of total Acceleration. The (x,y) point paths that the planner recieves should not have a total acceleration that goes over 10 m/s^2, also the jerk should not go over 50 m/s^3. (NOTE: As this is BETA, these requirements might change. Also currently jerk is over a .02 second interval, it would probably be better to average total acceleration over 1 second and measure jerk from that.

  2. There will be some latency between the simulator running and the path planner returning a path, with optimized code usually its not very long maybe just 1-3 time steps. During this delay the simulator will continue using points that it was last given, because of this its a good idea to store the last points you have used so you can have a smooth transition. previous_path_x, and previous_path_y can be helpful for this transition since they show the last points given to the simulator controller with the processed points already removed. You would either return a path that extends this previous path or make sure to create a new path that has a smooth transition with this last path.

Dependencies

About

Behavioural planning and trajectory generation in a highway scenario

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published