uav_motion

This repository generates a minimum-snap polynomial trajectory and controls a quadrotor with PX4 framework. It depends on ethz-asl/mav_trajectory_generation and Jaeyoung-Lim /mavros_controllers.

VIDEO DEMO

Give uav_motion a set of high-level waypoints, and it will do the remaining.

Installation

1. ethz-asl/mav_trajectory_generation

cd ~/catkin_ws/src
git clone https://github.com/catkin/catkin_simple.git
git clone https://github.com/ethz-asl/eigen_catkin.git
git clone https://github.com/ethz-asl/eigen_checks.git
git clone https://github.com/ethz-asl/nlopt.git
git clone https://github.com/ethz-asl/glog_catkin.git
git clone https://github.com/ethz-asl/mav_comm.git
git clone https://github.com/ethz-asl/yaml_cpp_catkin.git
git clone https://github.com/ethz-asl/mav_trajectory_generation.git
cd ~/catkin_ws/
catkin build

2. Jaeyoung-Lim/mavros_controllers

cd ~/catkin_ws/src
git clone https://github.com/Jaeyoung-Lim/mavros_controllers.git
cd ~/catkin_ws/
catkin build

3. ZhiangChen/uav_motion

cd ~/catkin_ws/src
git clone https://github.com/ZhiangChen/uav_motion.git
cd ~/catkin_ws/
catkin build

Getting Started

This gives you an example of using this package in gazebo.

1. Launch a quadrotor with px4 and mavros in gazebo

roslaunch px4 mavros_posix_sitl.launch

2. Define your keyframe waypoints in waypoint_generator.py and run the following commands

roslaunch uav_motion uav_motion.launch
rosrun uav_motion waypoint_generator.py

ROS Nodes

1. trajectory_generator

Publishers:

• "path_segments": mav_planning_msgs::PolynomialTrajectory ¹
• "path_segments_4D": mav_planning_msgs::PolynomialTrajectory4D ¹

¹ Either will be remapped to "trajecotry" depending on if yaw is included in key waypoints.

Subscribers:

• "/reference/pose": geometry_msgs::PoseStamped
• "waypoints": uav_motion::waypointsGoal ²

² It is a ROS action server.

Parameters:

• "mav_v": maximum velocity
• "mav_a": maximum acceleration
• "mav_ang_v": maximum angular velocity
• "mav_ang_a": maximum angular acceleration
• "current_ref_pose_as_start": the current reference pose will be included as the start point of trajectory if true

2. trajectory_sampler

Publishers:

• "reference/flatsetpoint": controller_msgs::FlatTarget
• "reference/yaw": std_msgs::Float32

Subscribers:

• "path_segments": mav_planning_msgs::PolynomialTrajectory ¹
• "path_segments_4D": mav_planning_msgs::PolynomialTrajectory4D ¹

Parameters:

• "dt": trajectory sampling rate

3. geometric_controller

Publishers:

• "/command/bodyrate_command" -> "/mavros/setpoint_raw/attitude": mavros_msgs::AttitudeTarget

Subscribers:

• "reference/flatsetpoint": controller_msgs::FlatTarget
• "reference/yaw": std_msgs::Float32

More information about this node can be found in Jaeyoung-Lim /mavros_controllers.

4. waypoint_generator.py

Publisher:

• "waypoints": uav_motion.msg.waypointsAction ³

³ It is a ROS action client.

Parameter Tuning for Real UAV

Most of the parameters have been well tuned. However, when having a UAV with different mass, airframe, and battery, you need to pay extra attention to the following parameters in geometric_controller:

/geometric_controller/Kp_x
/geometric_controller/Kp_y
/geometric_controller/Kp_z
/geometric_controller/enable_gazebo_state
/geometric_controller/enable_sim
/geometric_controller/normalizedthrust_constant
/geometric_controller/normalizedthrust_offset

The first three parameters Kp_z, Kp_y, and Kp_z determines how responsive the UAV is following the reference pose. For example, if Kp_z is too large, there might be an overshoot problem on z-axis. enable_gazebo_state and enable_sim have to be set to false for real UAVs. The last two parameters normalizedthrust_constant and normalizedthrust_offset are determined by UAV mass, airframe, etc. When your UAV has a larger mass than Iris, you may need to increase normalizedthrust_offset. Otherwise, the UAV can't reach the desired elevation. When the battery is low, you may also need to increase normalizedthrust_offset.