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

Identification of microgravity distu...

Smith, Andrew David.

FindBook

Google Book

Amazon

博客來

Identification of microgravity disturbances from space acceleration measurement system (SAMS) data using a neural network approach.

紀錄類型:	書目-電子資源 : Monograph/item
正題名/作者:	Identification of microgravity disturbances from space acceleration measurement system (SAMS) data using a neural network approach./
作者:	Smith, Andrew David.
面頁冊數:	203 p.
附註:	Source: Dissertation Abstracts International, Volume: 64-01, Section: B, page: 0390.
Contained By:	Dissertation Abstracts International64-01B.
標題:	Engineering, Mechanical. -
電子資源:	http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=3077001
ISBN:	0493974725

Identification of microgravity disturbances from space acceleration measurement system (SAMS) data using a neural network approach.
Smith, Andrew David.

Identification of microgravity disturbances from space acceleration measurement system (SAMS) data using a neural network approach. - 203 p.

Source: Dissertation Abstracts International, Volume: 64-01, Section: B, page: 0390.

Thesis (Ph.D.)--The Pennsylvania State University, 2002.

In order to support the individual microgravity science experiments and the microgravity missions, the Space Acceleration Measurement System (SAMS) (Baugher et al., 1993) has been developed by the NASA Lewis Research Center. The amount of data collected by a SAMS unit during a shuttle mission is in the several gigabytes (GB) range. In order to assess experimental results and to develop future experiments and facilities, a knowledge base must be created for carrier acceleration environments and effects of various disturbances such as crew activities, thruster firings, glove box fans, compressors, etc. Using the large amount of existing data, a cause and effect relationship between the type of disturbance and the acceleration environment for the shuttle and, eventually, the space station should be developed. In this thesis, neural networks have been used to achieve this objective.

ISBN: 0493974725Subjects--Topical Terms:

783786
Engineering, Mechanical.

Identification of microgravity disturbances from space acceleration measurement system (SAMS) data using a neural network approach.
LDR:03416nmm 2200325 4500 001 1865960
005 20041220103641.5
008 130614s2002 eng d
020 $a 0493974725
035 $a (UnM)AAI3077001
035 $a AAI3077001
040 $a UnM $c UnM
100 1 $a Smith, Andrew David. $3 1953380
245 1 0 $a Identification of microgravity disturbances from space acceleration measurement system (SAMS) data using a neural network approach.
300 $a 203 p.
500 $a Source: Dissertation Abstracts International, Volume: 64-01, Section: B, page: 0390.
500 $a Adviser: Alok Sinha.
502 $a Thesis (Ph.D.)--The Pennsylvania State University, 2002.
520 $a In order to support the individual microgravity science experiments and the microgravity missions, the Space Acceleration Measurement System (SAMS) (Baugher et al., 1993) has been developed by the NASA Lewis Research Center. The amount of data collected by a SAMS unit during a shuttle mission is in the several gigabytes (GB) range. In order to assess experimental results and to develop future experiments and facilities, a knowledge base must be created for carrier acceleration environments and effects of various disturbances such as crew activities, thruster firings, glove box fans, compressors, etc. Using the large amount of existing data, a cause and effect relationship between the type of disturbance and the acceleration environment for the shuttle and, eventually, the space station should be developed. In this thesis, neural networks have been used to achieve this objective.
520 $a Three neural networks have been used to classify SAMS data, a supervised learning multilayered neural network (MNN) and two unsupervised learning neural networks, self organizing map (SOM) and adaptive resonance theory 2-A (ART2-A). The unsupervised neural networks are needed due to an insufficient amount of a priori information to train the MNN for every event. The SOM has the advantage of global ordering, where similar weight vectors exist nearby on a neural array. ART2-A is able to add clusters as new information is presented. This is important since not all combinations of events are encountered in previous mission data. Input patterns for all three neural networks are formed from the power spectral density (PSD) of SAMS data. The PSDs are primarily calculated using Welch's method. Autoregressive coefficients and wavelet transform coefficients are used to form input patterns for the ART2-A neural network.
520 $a Since events are beginning and ending at irregular intervals during a mission, a segmentation algorithm has been adopted to better separate the data into stationary sections. This increases the likelihood that each section of data has only one set of events occurring.
520 $a The system is used to classify SAMS data is real-time for the STS-087 mission carrying the fourth United States Microgravity Payload (USMP-4). The total system is implemented including input pattern generation, clustering, and classification. An initial knowledge base is generated from STS-075 data, which carried USMP-3.
590 $a School code: 0176.
650 4 $a Engineering, Mechanical. $3 783786
650 4 $a Artificial Intelligence. $3 769149
650 4 $a Physics, Astronomy and Astrophysics. $3 1019521
690 $a 0548
690 $a 0800
690 $a 0606
710 2 0 $a The Pennsylvania State University. $3 699896
773 0 $t Dissertation Abstracts International $g 64-01B.
790 1 0 $a Sinha, Alok, $e advisor
790 $a 0176
791 $a Ph.D.
792 $a 2002
856 4 0 $u http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=3077001