dyn_small_obs_avoidance

Params

Common:

~map_frame (string, default: map)

SLAM map reference frame, ENU convention.

~uav_frame (string, default: base_link)

UAV body frame, forward-left-up convention.

Feature extract for mapping

Takes raw pointcloud and does basic pre-processing, publishes 3 processed pointclouds for mapping consumption. (TODO: Some of these params likely dont need to be exposed.)

~blind (double, default: 0.5)

Radius of the circle around the LiDAR to ignore (units same as LiDAR pointcloud).

~inf_bound (double, default: 10.0)

? Ignore points past this boundary.

~N_SCANS (int, default: 6)

Number of LIDAR scans. AVIA default is 6, VLP16 is 16.

~group_size (int, default: 8)

~disA (double, default: 0.01)

~disB (double, default: 0.1)

~p2l_ratio (double, default: 225.0)

~limit_maxmid (double, default: 6.25)

~limit_midmin (double, default: 6.25)

~limit_maxmin (double, default: 3.24)

~jump_up_limit (double, default: 170.0)

~jump_down_limit (double, default: 8.0)

~cos160 (double, default: 160.0)

~edgea (double, default: 2.0)

~edgeb (double, default: 0.1)

~smallp_intersect (double, default: 172.5)

~smallp_ratio (double, default: 1.2)

~point_filter_num (int, default: 1)

FAST-LIO mapping

Takes processed pointcloud and produces registered pointcloud. (TODO: Some of these params likely dont need to be exposed.)

~proc_cloud_topic (string, default: /laser_cloud_flat)

Topic to subscribe, produced by feature extract for use in map creation.

~reg_cloud_topic (string, default: /cloud_registered)

Topic to publish resulting registered cloud.

~imu_topic (string, default: /mavros/imu/data)

Topic to subscribe for IMU data.

~odom_topic (string, default: /mavros/odometry/out)

Topic to publish odom data registered to map.

~map_file_path (string, default: )

~max_iteration (int, default: 10)

~dense_map_enable (bool, default: true)

~fov_degree (double, default: 75.0)

~filter_size_corner (double, default: 0.3)

~filter_size_surf (double, default: 0.2)

~filter_size_map (double, default: 0.2)

~cube_side_length (double, default: 20.0)

Path-planning

Plans paths through map produced by the SLAM node (FAST LIO package), where the map is in pointcloud form. Essentially a wrapper for path-searching node.

~cloud_topic (string, default: /cloud_registered)

Referenced pointcloud, generated by SLAM node.

~odom_topic (string, default: /mavros/odometry/out)

Topic to publish odometry.

Path-searching

Uses kinodynamic A* search. An instance of path search is created by the path planning node.

~max_tau (double, default: 0.6)

~init_max_tau (double, default: 0.8)

~max_vel (double, default: 2.0)

~max_acc (double, default: 2.0)

~w_time (double, default: 10.0)

~horizon (double, default: 100.0)

~resolution_astar (double, default: 0.1)

~time_resolution (double, default: 0.8)

~lambda_heu (double, default: 5.0)

~vel_margin (double, default: 0.0)

~allocate_num (int, default: 100000)

~check_num (int, default: 1)

~optimistic (bool, default: true)

Acknowledgments/Original Repo

This repo was forked from https://github.com/hku-mars/dyn_small_obs_avoidance.git which states the following: This repository is used for UAV dynamic small obstacles avoidance. It is a complete system for lidar-based UAV, including FAST-LIO slam, time-accumulated KD-Tree mapping and kinodynamic A* search modules. It is able to avoid dynamic small obstacles (down to 20mm diameter bars) by running at 50Hz.

Related paper: "Avoiding dynamic small obstacles with onboard sensing and computating on aerial robots", available on arxiv now https://arxiv.org/abs/2103.00406.

Related video: https://youtu.be/pBHbQ_J1Qhc

Thanks for FAST-PLANNER(Zhou, Boyu and Gao, Fei and Wang, Luqi and Liu, Chuhao and Shen, Shaojie. Robust and efficient quadrotor trajectory generation for fast autonomous flight), FAST-PLANNER.