Doctor of Engineering (DEng)
Biological & Agricultural Engineering
Autonomous vehicles are becoming ever more common and offer many attractive benefits to society. They can operate for long periods of time unattended, operate in environments that may be dangerous to humans, perform time consuming or repetitive tasks and all with greater efficiency and lower costs than humans. For these vehicles to be able to do these things, algorithms need to be designed and optimized that allow them to interact with the real-world environment in safe, effective, and efficient ways.
We designed and built a set of three homogeneous water-based autonomous surface vehicles equipped with appropriate sensors and communications ability along with algorithms designed to allow these vehicles to perform various cooperative tasks using data obtained from the vehicles’ sensors and data shared between the vehicles. These vehicles were designed to be modular, economical, and, where possible, were constructed using off-the-shelf technology with programming designed to take advantage of these systems. When the COVID-19 pandemic put an end to lab and field work the physical vehicles were stored but the research continued utilizing a hybrid hardware-software simulation of the system. Three microcontrollers identical to the devices controlling the physical boats were attached via a Universal Serial Bus (USB) hub to a desktop computer running a simulated environment written in Python™. The three vehicles (microcontrollers) were given tasks including patrolling adjoining areas of the water body delineated by latitude and longitude boundaries while staying within their own boundary and avoiding collisions with the other vehicles. Initial testing was successful with the algorithm able to maintain the vehicles within their boundary >=95% of the time with no collisions. Additional problem types including parallel travel; wind and current challenges; and gradient tracking and relevant algorithms are discussed.
Smith, Daniel Davis, "Communications, Decision-Making, and Interactions of a Multi-Agent Autonomous Vehicle System" (2021). LSU Doctoral Dissertations. 5648.
Constant, W. David