Projects:2017s1-170 Formation Control and Obstacle Avoidance for Heterogeneous Multi-Agent Systems (Unmanned Aerial Vehicles and Robots)

Introduction

This project is to build an autonomous heterogeneous multi-vehicle system where several unmanned ground robots (UGV) and unmanned aerial vehicle (UAV) could achieve obstacle avoidance and formation control. In detail, UAV could move forward to any destination by controlling four airscrews on each side of UAV while two UGVs which are run by two wheels on each side would follow UAV as followers. In addition, UGVS could follow the UAV in certain formation, bypassing all obstacles on ground during moving process. After encountering an obstacle in moving, UGV could temporarily jump out of the formation shape and take bypassing this obstacle as the priority. If there are several obstacles on ground, the UGV could also find an optimal path to follow the UAV as the leader.

Motivation

................

System Strucutre

The whole system consists of such major parts: UAV, UGV, Optitrack camera system and a computer connected to above systems through Wifi and a wireless modem to release Wifi signal as communication channel. All equipment is shown in Figure 1.1.

Figure 1.1 Hardware of this system The working mechanism of this system is, firstly, the OptiTrack camera system is used to detect surrounding environment, to identify UAV and UGV and to locate UAV and UGV which would run in such environment. Secondly, locations of these vehicles could be transmitted from the camera system to the controller in Simulink in a host computer, so that the controller could convert these input location data into spiral velocity of four airscrews of UAV and velocity of two wheels in UGV. Lastly, after receiving above velocity data, UAV could move to any expected destination while UGV can follow UAV in certain formation, bypassing any obstacle in motion. The illustration of how this project work is shown in Figure 1.2. The following sections would have a focus on path planning and obstacle avoidance and explore what method could be used to address this issue.