Simple Lidar Simulator

📂 What's in this repo

• simulator/
  • simulator_launch.py : Python launch file. Aggregates all .sdf/.stl models and opens the various .yaml for gazebo simulation
  • /filter.py : Python Ros2 node capable of removing excessive LiDAR rays (v_axis only) to achieve non-uniform rays distributions
• params.yaml : File with all customisable parameters see below...
• models/<file_name>.stl : Car and cones .stl models
• sdf/
  • world_template.sdf : Contains the template for the gazebo simulation world
  • world.sdf : Contains the track for the gazebo simulation world with cones
• tracks/race_track.csv : Contains the shape of the track with coordinates

📦 Prerequisite packages

What we need are ros2 humble, gazebo fortress, ros-gz-bridge, colcon, pandas and numpy.

• Ubuntu 22.04 [link]
• ros2 humble [link]
• gazebo fortress [link]
• Colcon [link]

this command will install both numpy and pandas systemwise, creating a virtual environment and installing them with pip is reccomended.

sudo apt-get install ros-humble-ros-gz-bridge python3-colcon-common-extensions python3-numpy python3-pandas -y

⚙️ How to build & Run

mkdir -p ~/ros2_ws/src
cd ~/ros2_ws/src
git clone https://github.com/leonardonels/lidar_simulator_project.git
cd ~/ros2_ws
colcon build --packages-select lidar_simulator_project --symlink-install
source install/setup.bash

To move the car click on the plugins dropdown list in the top right corner of gazebo (vertical ellipsis), select the Key Publisher.

ros2 launch lidar_simulator_project simulator_launch.py

🧮 Parameters

The following code explains what every parameter does:

# LiDAR v_rays filter
filter:
  vertical_zones: [
            {'start': 0.0, 'end': 0.25, 'downsample': 4},  # Upper 25% of rows, keep 1/4
            {'start': 0.25, 'end': 0.75, 'downsample': 1},  # Middle 50%, keep all
            {'start': 0.75, 'end': 1.0, 'downsample': 4},  # Lower 25%, keep 1/4
        ]
  # use [{'start': 0.0, 'end': 1.0, 'downsample': 1}] to keep all LiDAR uniform rays

# Vehicle parameters
vehicle:
  abs_pose: [-2.4, 0, 0, 0, 0, -0.7]  # Absolute position and rotation
  linear_speed_f: 0.5 # Frontal car speed

# Lidar parameters
lidar:
  pose: [0.26, 0, 0.475, 0, -0.1570796, 0]  # Lidar position and rotation relative to the car
  hz: 20 # Lidar frequency
  h_samples: 500 # Number of Horizontal Rays
  h_min_angle: 2.356194   # 135° (convert from rads to degrees)
  h_max_angle: 3.926991   # 225° (convert from rads to degrees)
  h_resolution: 0.1 # Param that multiply h_samples
  v_samples: 128 # Number of Vertical Rays
  v_min_angle: -0.3926991 # -22.5° (convert from rads to degrees)
  v_max_angle: 0.3926991  # 22.5° (convert from rads to degrees)
  v_resolution: 0.1 # Param that multiply v_samples
  min_range: 0.01 # Lidar min depth range (THIS IS METERS!)
  max_range: 80.0 # Lidar max depth range (THIS IS METERS!)
  range_resolution: 0.01 # Param that multiply min_range and max_range
  noise_mean: 0.0 # Mean of the gaussian noise applied to Lidar
  noise_std: 0.05 # Std. Dev. of the gaussian noise applied to Lidar

# Cones parameters
cones:
  min_dist: 1.5 # Minimum distance between cones (THIS IS NOT METERS!)
  scale: 10 # Map coordinates scale
  track_width: 3 # Width of the track (THIS IS NOT METERS!)

Todo List

• Fine-Tuning lidar settings
• implement filter presets to achieve faster and smoother operations
• Implement a more accurate steering mode
• Add ros2 movement control
• Apply the true model for the car
• Improve performance

🏎️ Made by Leonardo Nels - 2025