Graduate Thesis Or Dissertation

 

Improving the accuracy of in-situ lower ABL wind measurements using sUAS Public Deposited

Downloadable Content

Download PDF
https://scholar.colorado.edu/concern/graduate_thesis_or_dissertations/sf268532d
Abstract
  • Wind measurement is a challenging problem as wind is a spatiotemporal vector field and obtaining a set of measurements with desired spatial and temporal sampling is not always possible. Small unmanned-aircraft systems (sUAS) are an emerging technology offering new capabilities for in-situ sensing in the lower atmospheric boundary layer (ABL) with cost, operational and safety advantages over larger airborne wind-measurement systems including the potential to use swarms of sUAS for denser sampling. Analysis of contemporary sUAS-based wind-measurement systems shows that the wind-measurement error is not dominated by a single source. Errors introduced by the relative-wind sensor, the airframe velocity estimate, the airframe attitude estimate, and the airframe attitude-rate estimate are all significant. This work focuses on improving sUAS velocity estimation and attitude estimation. State estimation for contemporary sUAS is based on sensors that are small, light and inexpensive but have much poorer performance than larger, navigation-grade sensors used in manned aircraft. Careful choice of a sensor fusion algorithm can allow for sUAS velocity-estimation accuracy on the order of 1 cm/s and attitude-estimation accuracy on the order of 0.1_ when using inexpensive consumer grade sensors. With a high-quality relative-wind sensor these levels of state-estimation accuracy allow for wind-measurement with accuracy on the order of 1 cm/s. However, some key issues in sUAS state estimation, particularly for sUAS ying in wind, have heretofore received little attention. Wind-gust-induced motion can have a significant effect on sUAS state estimation and many sensor-fusion algorithms proposed for use with sUAS have poor performance when subjected to wind-gust-induced motion. Speci_c approaches to sUAS attitude and velocity estimation are proposed based on analytic results and testing of contemporary sensors. In particular it is shown that an extended Kalman filter estimating attitude and gyroscope bias drift rate, and using time-differenced GPS velocity measurements to estimate translational acceleration, can provide the desired attitude-estimate accuracy with contemporary sUAS suitable sensors even in the presence of strong winds and turbulence. Simulation of this algorithm's performance when used on a sUAS ying in turbulent conditions show as much as an order-of-magnitude improvement in performance compared to other algorithms presented in the literature. Analysis of sUAS velocity estimation shows a strong sensitivity to the performance of the GPS receiver's velocity-measurement accuracy. Small UAS generally use inexpensive commercial-o_-the-shelf (COTS) GPS receivers which often have higher error levels when experiencing accelerating motion as is typical for sUAS flying in wind. A particular COTS GPS receiver that outputs raw satellite-channel-measurement data is examined. Analysis of test data was used to develop a method of using the COTS GPS raw-measurement data to produce a velocity estimate with the desired accuracy. Desire for an additional sensor suitable for use in either validating or improving SUAS attitude-estimation accuracy motivated the development of an optical reference-vector sensor system capable of making measurements with an accuracy better than 0.1 degrees while operating outside in full daylight at ranges in excess of 100 meters. Field tests were conducted using a typical sUAS, the optical reference-vector sensor system and contemporary sUAS-state-estimation sensors. Analysis of flight test data supports the assertion that sUAS-based wind-measurement systems with accuracy on the order of 1 cm/s are possible with contemporary sensors.
Creator
Date Issued
  • 2015
Academic Affiliation
Advisor
Committee Member
Degree Grantor
Commencement Year
Subject
Last Modified
  • 2019-11-14
Resource Type
Rights Statement
Language

Relationships

Items