Design And Implementation Of Multi-UAV Odor Source Localization System

Odor source localization has important applications in the prevention and control of biochemical terrorist attacks,hazardous material spills,fire and explosion accidents.In order to improve the efficiency of odor source localization,some scholars have tried to use robots to search for odor sources autonomously.As traditional ground robots are susceptible to terrain limitations and individual robots have a limited search range,Multi UAV scent source localization has higher efficiency.However,there are few studies on Multi UAV odor source localization,and they mostly stay in the simulation stage.In practical applications,there are three technical difficulties: first,the collision avoidance problem between UAVs and between UAVs and obstacles;second,the communication problem within the system;and third,how to design a stable UAV system and an efficient and reliable ground station management software.To address the above problems,the thesis designs a Multi UAV odor source localization system from practical applications,realizes collision avoidance in the cluster odor source localization task by using the artificial potential field method,and also designing a communication mechanism based on ground station data forwarding to reduce the communication volume of UAVs.The main work of this paper is as follows.(1)System scheme design.The functional requirements of the odor source positioning system are analyzed,and the scheme of the UAV system,communication network,and ground station system are determined.To address the problems of high information redundancy and high communication bandwidth requirements of the traditional full link communication topology,a communication mechanism based on ground station data forwarding is designed to reduce the communication volume of UAVs.(2)Design of cluster odor source location algorithm.In the natural environment,odor molecules released from an odor source form a feather-like distribution pattern under the action of air turbulence called an odor plume.Robot odor source localization mainly contains three stages: plume discovery,plume tracking,and odor source confirmation.For the plume discovery task,this paper designs a route pre-planning algorithm based on an improved rectangular wave zigzag traversal algorithm to achieve cooperative plume discovery search of multiple UAVs.For the smoke plume tracking task,a particle swarm algorithm-based Multi UAV smoke plume tracking algorithm is designed;collision avoidance strategies between UAVs and between UAVs and obstacles are designed using the artificial potential field method throughout the odor source localization process.(3)The software and hardware platform design of the Multi UAV odor source positioning system.The UAV platform was built using the open source flight control PIXHAWK;the UAV software was designed through a multi-process and modular approach,and the ground station software was designed using a hybrid programming approach of QML and C++.(4)Experimental validation.Through the joint commissioning of the ground station and the UAV,the functions of Multi UAV odor source positioning system were realized,the functions of the system were tested,and the algorithm verification was completed.The experimental results show that the system designed in the thesis can locate the odor source while effectively avoiding collisions between UAVs and between UAVs and obstacles.