Model Prediksi Infrastructure Resilience Sebagai Sistem Pendukung Pengambilan Keputusan Berbasis Bayesian Network

Agustin, Andini Dwi (2021) Model Prediksi Infrastructure Resilience Sebagai Sistem Pendukung Pengambilan Keputusan Berbasis Bayesian Network. Masters thesis, Institut Teknologi Sepuluh Nopember.

This is the latest version of this item.

[img] Text
view_usp=sharing - Accepted Version
Restricted to Repository staff only until 1 October 2023.

Download (75kB) | Request a copy
[img] Text
view_usp=sharing - Accepted Version
Restricted to Repository staff only until 1 October 2023.

Download (67kB) | Request a copy
Official URL: https://drive.google.com/file/d/191mkY14bM88ePUxZL...

Abstract

Kondisi geologi Indonesia yang dikelilingi oleh wilayah yang berada pada pertemuan lempeng dan deret gunung api, umumnya banyak terdapat patahan aktif, yang sering menyebabkan terjadinya gempa bumi. Gempa bumi merupakan bencana alam yang sangat mempengaruhi kondisi fisik infrastruktur, sehingga menyebabkan menurunnya atau tidak berfungsinya suatu kota. Untuk mengurangi dampak kerusakan akibat bencana dibutuhkan penerapan konsep resilience agar kota-kota mulai bergerak dalam konteks perubahan menuju ketahanan infrastruktur terhadap bencana dengan identifikasi langkahlangkah yang dapat meningkatkan kinerja infrastruktur untuk memenuhi resilience goal. Kejadian bencana yang tidak terduga menyebabkan model prediksi infrastructure resilience seharusnya berbasis ketidakpastian (uncertainty). Penelitian ini bertujuan untuk membuat model prediksi resilience infrastructure berbasis uncertainty. Bayesian Network digunakan sebagai tools untuk menginferensi model, dengan variabel-variabel penelitian didapatkan dari studi literatur secara mendalam. Conditional probability tabel didapat dengan mengkombinasikan data historis dan Fuzzy logic. Penelitian ini menggunakan studi kasus kota/kabupaten di Sumatera Barat yang terdampak gempa. Tiga skenario lokasi berdasarkan klasifikasi dari Badan Nasional Penanggulangan Bencana (BNPB) dibuat untuk merepresentasikan kondisi sesungguhnya dari kasus gempa yaitu skenario pada lokasi kota 60KM, 100 KM dan 140 KM dari pusat gempa. Untuk validasi model digunakan metode óne test only validation. Hasil penelitian menunjukkan bahwa model ini mampu memprediksi kondisi resilience kota/kabupaten yang terdampak, dengan akurasi 73,68%. Keluaran dari model ini, selain bisa digunakan untuk memprediksi kondisi resilience infrastruktur kota, juga dapat digunakan sebagai Early warning system untuk pengambilan keputusan rekonstruksi maupun rehabilitasi infrastruktur ====================================================================================================== The geological condition of Indonesia, which is surrounded by areas located at the confluence of plates and volcanic series, generally has many active faults, which often cause earthquakes. Earthquakes are natural disasters that greatly affect the physical condition of infrastructure, causing the decline or non-functioning of a city. To reduce the impact of disaster damage, it is necessary to apply the concept of resilience so that cities begin to move in the context of changes towards infrastructure resilience against disasters by identifying steps that can improve infrastructure performance to meet the resilience goal. Unexpected disaster events cause infrastructure resilience prediction models to be based on uncertainty. This study aims to create a predictive model for resilience infrastructure. Bayesian Network-based research method for model inference with research variables obtained from an in-depth study of the literature. Conditional Probability Table is obtained by combining historical data and fuzzy logic. This study uses case studies of cities/districts in West Sumatra that were affected by the earthquake. Three scenarios were made to present the actual conditions of the earthquake case, namely scenarios in the city location of 60KM, 100KM and 140KM from the epicenter. For model validation, the one test only validation method is used. The results show that this model able to predict the resilience conditions of the affected cities/districts quite well, with an accuracy of 73.68%. The output of this model, apart from being able to predict the condition of urban infrastructure resilience, can also be used as an early warning system for decision making on infrastructure reconstruction and rehabilitation.

Item Type: Thesis (Masters)
Uncontrolled Keywords: Bayesian-Network, conditional-probability-table, earthquakes, resilience-infrastructure
Subjects: T Technology > TA Engineering (General). Civil engineering (General)
T Technology > TA Engineering (General). Civil engineering (General) > TA169 Reliability (Engineering)
Divisions: Faculty of Civil, Planning, and Geo Engineering (CIVPLAN) > Civil Engineering > 22101-(S2) Master Thesis
Depositing User: Andini Dwi Agustin
Date Deposited: 14 Aug 2021 06:00
Last Modified: 14 Aug 2021 06:00
URI: https://repository.its.ac.id/id/eprint/86230

Available Versions of this Item

Actions (login required)

View Item View Item