Examples of Simulation

This page is generated by generate_example_gallery.py from the demo images under src/simulations and updated automatically after merge. Do not edit it by hand — edit the docstring of the simulation script instead.

Localization

EKF vs UKF vs Particle Filter Comparison

Compares EKF, UKF and Particle Filter localization with the same motion model, GNSS observations and noise
Author: Sahruday Patti

EKF vs UKF Comparison

Compares EKF and UKF localization with GNSS observations
Author: Bruno DOKPOMIWA

Extended Kalman Filter Localization

Estimates the vehicle pose by fusing a motion model and GNSS observations with an Extended Kalman Filter
Author: Shisato Yano

Kalman Filter 1D

Estimates a 1 dimensional position with a Linear Kalman Filter using a noisy velocity and LiDAR observation
Estimates a 1 dimensional position and speed with a Linear Kalman Filter using a noisy acceleration and LiDAR observation
Author: Shisato Yano

Particle Filter Localization

Estimates the vehicle pose with weighted samples (particles) and GNSS observations
Author: Sahruday Patti

Unscented Kalman Filter Localization

Estimates the vehicle pose with an Unscented Kalman Filter using sigma points and GNSS observations
Author: Bruno DOKPOMIWA

Mapping

Binary Occupancy Grid Map

Builds a grid map that marks each cell as occupied or free from LiDAR point clouds
Author: Shisato Yano, Bhavesh Lokesh Agarwal

Cost Map

Builds a grid map that assigns a cost to each cell from LiDAR point clouds
Author: Bhavesh Lokesh Agarwal

NDT Map

Builds a map of normal distributions (mean and covariance) per grid cell from LiDAR point clouds
Author: Shisato Yano

Potential Field Map

Builds a potential field map from LiDAR point clouds, with high potential around obstacles
Author: Panav Arpit Raaj

VFH Polar Histogram

Builds a polar obstacle density histogram around the vehicle from LiDAR point clouds
Author: Khushi

Path Planning

ACO

Ant Colony Optimization: ants search the grid and reinforce shorter paths with pheromones
Author: Banaan Kiamanesh

Bidirectional A*

Runs A* search from both the start and the goal until the two searches meet
Author: Erwin Lejeune

Hybrid A*

A* search over continuous vehicle poses using motion primitives that respect the minimum turning radius
Author: Shreyansh Shethia

A*

Grid-based shortest path search guided by a heuristic distance to the goal
Author: Shantanu Parab

Dijkstra

Grid-based shortest path search that expands the lowest-cost node first
Author: Ashish Varma

D*

Incremental replanning when a new obstacle appears on the planned path
Author: Erwin Lejeune

Elastic Bands

Smooths an A* path by treating it as an elastic band pushed away from obstacles
Author: Erwin Lejeune

GA

Genetic Algorithm: evolves candidate paths and smooths the best one with a spline
Author: Banaan Kiamanesh

Gradient Descent

Plans a path by descending the gradient of a potential field toward the goal
Author: Panav Arpit Raaj

Informed RRT*

RRT* that samples only inside an ellipse once a first path is found, to converge faster
Author: Rajat Arora

PRM

Samples random nodes, connects them into a roadmap and searches it with Dijkstra
Author: Erwin Lejeune

PSO

Particle Swarm Optimization: a swarm of candidate paths converges toward the best one
Author: Erwin Lejeune

Q-Learning

Learns a Q-table policy on the grid by reinforcement learning and follows its greedy path
Author: Banaan Kiamanesh

RRT

Grows a random tree from the start until it reaches the goal
Author: Shantanu Parab

Bidirectional RRT*

Grows RRT* trees from both the start and the goal and connects them
Author: Erwin Lejeune

RRT*

RRT that rewires the tree to shorten paths as more samples are added
Author: Erwin Lejeune

Path Tracking

Adaptive Pure Pursuit Path Tracking

Pure Pursuit whose look-ahead distance shortens on sharp turns and lengthens on straights
Author: Erwin Lejeune

LQR (Linear Quadratic Regulator) Path Tracking

Steering control that minimizes a quadratic cost of tracking error and control input
Author: Shisato Yano

MPC Path Tracking

Optimizes future inputs over a prediction horizon with do-mpc and shows the predicted trajectory
Author: Mohit Kumar

MPPI Path Tracking

Samples many input sequences, weights them by cost, and shows all evaluated trajectories
Author: Erwin Lejeune

Pure Pursuit Path Tracking

Steers toward a look-ahead point on the course
Author: Shisato Yano

Rear Wheel Feedback Path Tracking

Steering control based on the lateral and heading errors at the rear wheel
Author: Shisato Yano

Stanley Steering Control Path Tracking

Steering control based on the heading error and the lateral error at the front wheel
Author: Shisato Yano

Perception

LiDAR Obstacle Sensing

Simulates an omni-directional LiDAR that detects moving and static obstacles as point clouds
Author: Shisato Yano

Rectangle Fitting Detection

Fits a rectangle to each obstacle’s point cloud by L-shape fitting
Fits a rectangle to each obstacle’s point cloud by robust least squares
Author: Shisato Yano, Shreyansh Shethia

Point Cloud Nearest Neighbor Search (kd-tree)

Finds the nearest source point for each target point with a kd-tree
Finds all source points within a radius of each target point with a kd-tree
Author: Shisato Yano

Sensor’s Extrinsic Parameters Estimation

Estimates the LiDAR’s mounting position and angle on the vehicle with an Unscented Kalman Filter
Author: Shisato Yano

Course

Cubic Spline Course

Builds a 2D cubic spline course and computes its yaw angle and curvature
Interpolates sample points with a 1D cubic spline
Author: Shisato Yano