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

Inverse gravity modeling for depth v...

King, Thomas Steven.

FindBook

Google Book

Amazon

博客來

Inverse gravity modeling for depth varying density structures through genetic algorithm, triangulated facet representation, and switching routines.

紀錄類型:	書目-語言資料,印刷品 : Monograph/item
正題名/作者:	Inverse gravity modeling for depth varying density structures through genetic algorithm, triangulated facet representation, and switching routines./
作者:	King, Thomas Steven.
面頁冊數:	461 p.
附註:	Adviser: Tien-Chang Lee.
Contained By:	Dissertation Abstracts International67-03B.
標題:	Geophysics. -
電子資源:	http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=3210413
ISBN:	9780542599576

Inverse gravity modeling for depth varying density structures through genetic algorithm, triangulated facet representation, and switching routines.
King, Thomas Steven.

Inverse gravity modeling for depth varying density structures through genetic algorithm, triangulated facet representation, and switching routines. - 461 p.

Adviser: Tien-Chang Lee.

Thesis (Ph.D.)--University of California, Riverside, 2006.

A hybrid gravity modeling method is developed to investigate the structure of sedimentary mass bodies. The method incorporates as constraints surficial basement/sediment contacts and topography of a mass target with a quadratically varying density distribution. The inverse modeling utilizes a genetic algorithm (GA) to scan a wide range of the solution space to determine initial models and the Marquardt-Levenberg (ML) nonlinear inversion to determine final models that meet pre-assigned misfit criteria, thus providing an estimate of model variability and uncertainty.

ISBN: 9780542599576Subjects--Topical Terms:

535228
Geophysics.

Inverse gravity modeling for depth varying density structures through genetic algorithm, triangulated facet representation, and switching routines.
LDR:03389nam 2200301 a 45 001 965829
005 20110908
008 110908s2006 eng d
020 $a 9780542599576
035 $a (UnM)AAI3210413
035 $a AAI3210413
040 $a UnM $c UnM
100 1 $a King, Thomas Steven. $3 1288580
245 1 0 $a Inverse gravity modeling for depth varying density structures through genetic algorithm, triangulated facet representation, and switching routines.
300 $a 461 p.
500 $a Adviser: Tien-Chang Lee.
500 $a Source: Dissertation Abstracts International, Volume: 67-03, Section: B, page: 1338.
502 $a Thesis (Ph.D.)--University of California, Riverside, 2006.
520 $a A hybrid gravity modeling method is developed to investigate the structure of sedimentary mass bodies. The method incorporates as constraints surficial basement/sediment contacts and topography of a mass target with a quadratically varying density distribution. The inverse modeling utilizes a genetic algorithm (GA) to scan a wide range of the solution space to determine initial models and the Marquardt-Levenberg (ML) nonlinear inversion to determine final models that meet pre-assigned misfit criteria, thus providing an estimate of model variability and uncertainty.
520 $a The surface modeling technique modifies Delaunay triangulation by allowing individual facets to be manually constructed and non-convex boundaries to be incorporated into the triangulation scheme. The sedimentary body is represented by a set of uneven prisms and edge elements, comprised of tetrahedrons, capped by polyhedrons. Each underlying prism and edge element's top surface is located by determining its point of tangency with the overlying terrain. The remaining overlying mass is gravitationally evaluated and subtracted from the observation points. Inversion then proceeds in the usual sense, but on an irregular tiered surface with each element's density defined relative to their top surface.
520 $a Efficiency is particularly important due to the large number of facets evaluated for surface representations and the many repeated element evaluations of the stochastic GA. The gravitation of prisms, triangular faceted polygons, and tetrahedrons can be formulated in different ways, either mathematically or by physical approximations, each having distinct characteristics, such as evaluation time, accuracy over various spatial ranges, and computational singularities. A decision tree or switching routine is constructed for each element by combining these characteristics into a single cohesive package that optimizes the computation for accuracy and speed while avoiding singularities.
520 $a The GA incorporates a subspace technique and parameter dependency to maintain model smoothness during development, thus minimizing creating nonphysical models. The stochastic GA explores the solution space, producing a broad range of unbiased initial models, while the ML inversion is deterministic and thus quickly converges to the final model. The combination allows many solution models to be determined from the same observed data.
590 $a School code: 0032.
650 4 $a Geophysics. $3 535228
690 $a 0373
710 2 0 $a University of California, Riverside. $3 791122
773 0 $t Dissertation Abstracts International $g 67-03B.
790 $a 0032
790 1 0 $a Lee, Tien-Chang, $e advisor
791 $a Ph.D.
792 $a 2006
856 4 0 $u http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=3210413