The Inclusion of Realistic Winds in a Simulated Environment for the Study of Wind-Unmanned Aircraft System (UAS) Interactions
Abstract
Rapidly advancing technologies in graphics, gaming, and artificial intelligence are driving the development of synthetic environments (SEs). Accurate physics is a key component in SEs to ensure that simulations represent reality as closely as possible. This is particularly important for engineering systems, or for planning operations in regions of high complexity, such as urban environments. We have developed a simple simulation platform for testing interactions between unmanned aircraft systems (UASs) and winds in urban environments. This platform is built on Unreal Engine (i.e., a gaming engine) and cityscapes of real-world cities adapted from Open Street Map. The wind fields are simulated using lattice Boltzmann methods within a geometrically representative city model. From these wind fields, we can simulate wind-UAS interactions, visualize wind streamlines, and identify regions of likely high wind speed and/or turbulence. We employ remote procedure calls to handle the large size of volumetric wind data, storing and managing these data separately from the simulation environment. Within this simulation platform, we demonstrate how a hyper-local knowledge of winds can be used to model significant deviations in UAS trajectories for a preplanned flight through Chicago. The resulting information could be used to plan UAS operations balancing safety, area coverage, power usage, location avoidance, and total operational time. Improved UAS-wind interaction models would also allow the testing of UAS suitability for specific locations and conditions.
Document Details
- Document Type
- Technical Report
- Publication Date
- May 27, 2022
- Accession Number
- AD1170051
Entities
People
- Brent Kraczek
- Carl Lederman
- Kelly Kirk
- Yansen Wang
Organizations
- United States Army Research Laboratory