• Disable PCL_CONVERTER ground plane filtering

• Make node for state estimation
• Add function for splitting node into multiple smaller sections
• Add variable sampling domain

This package contains node for implementing the motion planner in a mobile robot. Package contains the following nodes:

• obstacle_tracker: track pedestrian trough world and esimate its velocity
• rrt: Generate a motion plan using RRT with closed-loop prediction
• state_estimator: generates state estimation message
• simulator: contains a Toyota Prius Gazebo simulator
• car_msgs: contains all messages used by the package

ROS Melodic with Visualization stack

mkdir ~/catkin_ws/src
cd ~/catkin_ws/src

git clone -b newcostfunction https://github.com/vdBerg93/motionplanner.git
git clone -b four_persons https://github.com/vdBerg93/pedsim_ros.git
git clone https://github.com/bbrito/lmpcc_msgs.git
git clone -b rrt https://gitlab.tudelft.nl/bdebrito/lmpcc_obstacle_feed.git 
git clone -b rrt_integration https://gitlab.tudelft.nl/bdebrito/lmpcc.git

catkin_make

Two scenarios can be tested:

1. pedwalking.launch
2. pedcrossing.launch

The following code will start the simulations:

roslaunch rrt #scenario#
roslaunch lmpcc_obstacle_feed lmpcc_obstacle_feed.launch
roslaunch lmpcc lmpcc.launch
roslaunch rrt rrt.launch

Explain what these tests test and why

Give an example

Explain what these tests test and why

Give an example

Add additional notes about how to deploy this on a live system

• Berend van den Berg - Initial work - vdBerg93