Simulasi pemodelan inversi data magnetotelurik 1-D menggunakan metode algoritma genetik

Romdhani , Nur Arifah (2003) Simulasi pemodelan inversi data magnetotelurik 1-D menggunakan metode algoritma genetik. Undergraduate thesis, FMIPA Undip.

[img]PDF
Restricted to Repository staff only

3236Kb
[img]
Preview
PDF
18Kb
[img]
Preview
PDF
359Kb
[img]
Preview
PDF
581Kb
[img]
Preview
PDF
399Kb
[img]
Preview
PDF
591Kb
[img]
Preview
PDF
397Kb
[img]
Preview
PDF
331Kb
[img]PDF
Restricted to Repository staff only

773Kb
[img]
Preview
PDF
334Kb
[img]
Preview
PDF
1817Kb

Abstract

Inversi dalam geofisika. dapat digunakan untuk memperkirakan struktur bawah permukaan dan data geofisika secara objektif. Masalah inversi non-linear dapat diselesaikan dengan menggunakan Algoritma Genetik (AG) yang bertujuan sebagai solusi respon model anatara data sintetik dengan data inversi dami data MT-1D. Algoritma ini menganalogikan optimasi fungsi objektif sebagai proses kehidupan mahl-uk hid-up. Pada tahap awal dibangkitkan suatu populasi yang terdiri atas beberapa rangkaian bilangan biner random dan disebut sebuah populasi kromosom. Setiap kromosom terdapat representasi model parameter yang dinamakan gen. Selanjutnya dengan mengevaluasi milt sebagai fungsi objektif dilakukan proses seleksi untuk menentukan kromosom. Dalam kasus inversi data magnetotelurik 1-dirnensi, gen dalam sate kromosom merepresentasikan harga ketebalan dan resitivitas tiap lapisan Proses ini berlangsung secara iteratif hingga generasi tertentu dimana telah diperoleh milt yang minimal. Hash inversi data sintetik dengan algoritma menghasilkan nilai rtes misfit yang relatif kecil dan secara geometris yang dapat menyelesaikan model sintetik adalah pada maksimum generasi 290; pc = 0,005; pm = 0,4; dan jumlah parameter 9. Inverse method in geophysics can he use to predict under suface structure from geophysic data in a objective manner. Non-linear inverse problem can he solved by using Genetic Algorithm (GA) which aim to as solution of respon model between synthetic data with data of inversi of data of magnetotelurik 1 dimension. This algorithm optimizes an objective function in a manner by which biological organisms multiplicate themselves. At the first stage a population which consist of some array of binary random digits generated, this called a population of chromosomes. In each chromosome contained representation of model parameters, which called the genes. Furthermore, by evaluating misfit as objective function a selection was done to choose the fittest chromosome. In case of magnetotelluric 1-dimensional inversion, the genes in a chromosome represents the value of thickness and resistivities in each layer respectively This process is an iterative one, until a specific generation with minimum misfit was attained. The results of the inversion with this algorithm give relatively small misfit and geometrically can resolve the synthetic model on maximum generation 290, probability of crossover 0.005, probability mutation 0,4, and 9 parameters. xiv

Item Type:Thesis (Undergraduate)
Subjects:Q Science > QD Chemistry
Q Science > QC Physics
Divisions:Faculty of Science and Mathematics > Department of Physics
ID Code:30301
Deposited By:Mr UPT Perpus 1
Deposited On:26 Oct 2011 09:32
Last Modified:26 Oct 2011 09:32

Repository Staff Only: item control page