Trajectory Optimization of Meteorological Sampling
Abstract
Swarming involves controlling multiple unmanned aerial systems or UAS in formation through the use of controls and algorithms. Swarm systems may be distributed and not rely on a central controller. As a result, this gives the system the potential to be robust and scalable, allowing for flexibility for the engineers to approach problems differently. Based on a variety of a few models and algorithms, such as artificial potential fields (APFs), agent-based modeling, dynamic data driven application systems (DDDAS), and virtual structures, it may be determined that using a variation of one of these would be the best course of action for formation flight for a swarm of UASs. Choosing the right controller is dependent on what works best for acquiring atmospheric data in a coordinated formation. Current atmospheric data is commonly taken using a weather tower or mesonet. A mesonet is typically a 10m high tower with a pressure, temperature, humidity sensor placed at the top. Deciding which controller can be used to not only take useful atmospheric data, but in many cases replace a mesonet due to mobility and customization is the goal. A wind profile is a transient matter, so using a swarm vs using one drone or a mesonet helps to solve the issues that the latter two run into due to time and space. A swarm can record multiple points at one time due to each agent being a data point representation, whereas a single drone can only account for a single location in time. A swarm using a virtual structure (VS) can cover a variety of amounts of space in a coordinated shape. A meosnet is stationary and only oriented vertically and an uncoordinated group of UAS does not have the capability to operate together. This leaves the capability that a VS swarm has to fill in the gaps or even replace the traditional approaches. An array of sensor packages with mobility, coordinated movement, and endless data points could give the VS swarm the advantage in atmospheric data sampling.
Collections
- OSU Theses [15752]