東華大學圖書館 |

語系: 繁體中文

說明(常見問題)

回圖書館首頁

手機版館藏查詢

登入

回首頁

切換: 標籤 | MARC模式 | ISBD

FindBook

Google Book

Amazon

博客來

Machine Learning Based Approaches to Dynamic Wind Estimation for Unmanned Aerial Vehicles.

紀錄類型:	書目-電子資源 : Monograph/item
正題名/作者:	Machine Learning Based Approaches to Dynamic Wind Estimation for Unmanned Aerial Vehicles./
作者:	Baraka, Ahmed.
面頁冊數:	1 online resource (96 pages)
附註:	Source: Masters Abstracts International, Volume: 84-12.
Contained By:	Masters Abstracts International84-12.
標題:	Mechanical engineering. -
電子資源:	http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=30486911click for full text (PQDT)
ISBN:	9798379731748

Machine Learning Based Approaches to Dynamic Wind Estimation for Unmanned Aerial Vehicles.
Baraka, Ahmed.

Machine Learning Based Approaches to Dynamic Wind Estimation for Unmanned Aerial Vehicles. - 1 online resource (96 pages)

Source: Masters Abstracts International, Volume: 84-12.

Thesis (M.S.)--New Mexico State University, 2023.

Includes bibliographical references

Wind estimation techniques for Unmanned Aerial Vehicles (UAVs) using data-oriented machine learning (ML) algorithms are studied in this thesis. These ML-based methods work without relying on drone dynamics models but by using only the position and Euler angle measurements taken by the UAV inertial measurement units (IMU) and Global Positioning System (GPS). Further, the use of internal state parameters as the input to the algorithms eliminates the reliance on control thrust inputs. This study examines the efficacy of Decision Tree Regression (DT) and Reservoir Computing (using the Echo State Network) compared against the Long-Short Term Memory (LSTM) Neural Networks and the K-Nearest Neighbors (KNN) algorithm. Both simulated data and real flight data were used to develop the four proposed techniques. The results show that ESN was outperformed by the other techniques but offers a flexible, non-linear dynamical model that maintains the sequential time-dependent data. Moreover, since it did not require any training its run time is a fraction of what it would take the LSTM. Further, this allows for future online training trials which would be preferable for the drone application and flexible control system to wind forces. In more instances the KNN outperformed DT but very marginally for the outdoor flight data. DT was better than KNN in instances where the dataset lengths were longer.

Electronic reproduction.
Ann Arbor, Mich. :
ProQuest,
2023

Mode of access: World Wide Web

ISBN: 9798379731748Subjects--Topical Terms:

649730
Mechanical engineering.
Subjects--Index Terms:

Decision treeIndex Terms--Genre/Form:

542853
Electronic books.

Machine Learning Based Approaches to Dynamic Wind Estimation for Unmanned Aerial Vehicles.
LDR:02772nmm a2200397K 4500 001 2363204
005 20231116093823.5
006 m o d
007 cr mn ---uuuuu
008 241011s2023 xx obm 000 0 eng d
020 $a 9798379731748
035 $a (MiAaPQ)AAI30486911
035 $a AAI30486911
040 $a MiAaPQ $b eng $c MiAaPQ $d NTU
100 1 $a Baraka, Ahmed. $3 3703955
245 1 0 $a Machine Learning Based Approaches to Dynamic Wind Estimation for Unmanned Aerial Vehicles.
264 0 $c 2023
300 $a 1 online resource (96 pages)
336 $a text $b txt $2 rdacontent
337 $a computer $b c $2 rdamedia
338 $a online resource $b cr $2 rdacarrier
500 $a Source: Masters Abstracts International, Volume: 84-12.
500 $a Advisor: Sun, Liang.
502 $a Thesis (M.S.)--New Mexico State University, 2023.
504 $a Includes bibliographical references
520 $a Wind estimation techniques for Unmanned Aerial Vehicles (UAVs) using data-oriented machine learning (ML) algorithms are studied in this thesis. These ML-based methods work without relying on drone dynamics models but by using only the position and Euler angle measurements taken by the UAV inertial measurement units (IMU) and Global Positioning System (GPS). Further, the use of internal state parameters as the input to the algorithms eliminates the reliance on control thrust inputs. This study examines the efficacy of Decision Tree Regression (DT) and Reservoir Computing (using the Echo State Network) compared against the Long-Short Term Memory (LSTM) Neural Networks and the K-Nearest Neighbors (KNN) algorithm. Both simulated data and real flight data were used to develop the four proposed techniques. The results show that ESN was outperformed by the other techniques but offers a flexible, non-linear dynamical model that maintains the sequential time-dependent data. Moreover, since it did not require any training its run time is a fraction of what it would take the LSTM. Further, this allows for future online training trials which would be preferable for the drone application and flexible control system to wind forces. In more instances the KNN outperformed DT but very marginally for the outdoor flight data. DT was better than KNN in instances where the dataset lengths were longer.
533 $a Electronic reproduction. $b Ann Arbor, Mich. : $c ProQuest, $d 2023
538 $a Mode of access: World Wide Web
650 4 $a Mechanical engineering. $3 649730
650 4 $a Automotive engineering. $3 2181195
650 4 $a Transportation. $3 555912
653 $a Decision tree
653 $a Wind estimation
653 $a Neural networks
653 $a K-nearest neighbors
653 $a Unmanned aerial vehicles
655 7 $a Electronic books. $2 lcsh $3 542853
690 $a 0548
690 $a 0709
690 $a 0540
710 2 $a ProQuest Information and Learning Co. $3 783688
710 2 $a New Mexico State University. $b Mechanical and Aerospace Engineering. $3 3540154
773 0 $t Masters Abstracts International $g 84-12.
856 4 0 $u http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=30486911 $z click for full text (PQDT)