INTERPRETASI STRUKTUR BAWAH PERMUKAAN DATA GAYABERAT MENGGUNAKAN ALGORITMA JARINGAN SARAF TIRUAN STUDI KASUS DAERAH PANAS BUMI UNGARAN, JAWA TENGAH

*Ratih Rundri Utami  -  Departemen Fisika, Fakultas Sains dan Matematika, Universitas Diponegoro, Semarang, Indonesia
Agus Setyawan  -  Departemen Fisika, Fakultas Sains dan Matematika, Universitas Diponegoro, Semarang, Indonesia
Rahmat Gernowo  -  Departemen Fisika, Fakultas Sains dan Matematika, Universitas Diponegoro, Semarang, Indonesia
Published: 1 Oct 2016.
View
Open Access
Citation Format:
Abstract

Artificial neural networks have been used in an application of geophysical such as seismic, electromagnetic, restivity, and gravity. In this study, artificial neural network system used is the method of propagation of gravity to produce anomalies corresponding to the desired anomalies on the geothermal area of Mount Ungaran, Central Java. In the training process to produce the best weight with 4 hidden layer with a correlation coefficient of 0.99 and the testing process using the results of the best training with a correlation coefficient of 0.97 and a yield value that resembles Bouguer anomaly in the research area., so it can be seen under the surface of the structure with the results of the best network where there is a high density value of 2.70 to 2.80 g/cm3 in lava basalt as geothermal systems Mount Ungaran. Density 2.40 to 2.80 g/cm3 Low contained in the surface area of Mount Ungaran with the majority of sedimentary rocks of andesitic pyroclastic products of Mount Ungaran Young.

Keywords: Artificial Neural Network, Inversion, Mount of Ungaran, Basalt

Article Metrics: