Sistem Deteksi Anomali Data Arus Motor Dan Beban Mekanik Lift Kargo Menggunakan Metode Isolation Forest

Pamungkas, Okky Surya (2026) Sistem Deteksi Anomali Data Arus Motor Dan Beban Mekanik Lift Kargo Menggunakan Metode Isolation Forest. Other thesis, Institut Teknologi Sepuluh Nopember.

Text 2040221112-Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only Download (4MB) | Request a copy

Abstract

Lift kargo memerlukan pemantauan berkelanjutan karena penyimpangan hubungan arus motor dan beban mekanik dapat mengindikasikan kondisi operasi tidak normal. Penelitian ini merancang sistem monitoring dan deteksi anomali lift kargo sebagai peringatan dini menggunakan Isolation Forest. Arus keluaran inverter Schneider ATV320 yang merepresentasikan arus agregat dua motor serta beban dari load cell indicator LM8-RR4D dibaca Raspberry Pi 4 melalui Modbus RTU, disimpan pada InfluxDB, dan ditampilkan pada dashboard. Model forward dan reverse dibangun secara terpisah menggunakan 200 isolation tree dan ukuran subsampel 256. Pelatihan menggunakan 1.855 data normal yang terdiri atas 1.490 data forward dan 365 data reverse. Pengujian menggunakan 564 data, yaitu 464 data normal dan 100 data anomali berbasis tiga skenario gangguan. Threshold persentil ke-95 sebesar 0,5981 untuk forward dan 0,5832 untuk reverse menghasilkan 100 true positive, 441 true negative, 23 false positive, dan tanpa false negative. Secara gabungan, model memperoleh accuracy 95,92%, precision kelas anomali 81,30%, recall 100%, dan F1-score 89,69%. Sistem mampu mendeteksi seluruh anomali skenario dan menampilkan status WARNING sebagai indikasi awal. Namun, hasil deteksi tetap memerlukan pemeriksaan lebih lanjut oleh operator atau teknisi karena data anomali yang digunakan dalam pengujian belum berasal dari gangguan aktual.
===================================================================================================================================
Cargo lifts require continuous monitoring because deviations in the relationship between motor current and mechanical load may indicate abnormal operating conditions. This study develops a cargo lift monitoring and anomaly detection system as an early warning system using Isolation Forest. The output current of the Schneider ATV320 inverter, representing the aggregate current of two motors, and load data from an LM8-RR4D load cell indicator were acquired by a Raspberry Pi 4 via Modbus RTU, stored in InfluxDB, and displayed on a dashboard. Separate forward and reverse models were developed using 200 isolation trees and a subsample size of 256. Training used 1,855 normal observations, comprising 1,490 forward and 365 reverse observations. Testing used 564 observations, consisting of 464 normal observations and 100 scenario-based anomalies representing three fault conditions. The 95th-percentile thresholds of 0.5981 for the forward direction and 0.5832 for the reverse direction produced 100 true positives, 441 true negatives, 23 false positives, and no false negatives. The combined evaluation yielded an accuracy of 95.92%, anomaly-class precision of 81.30%, recall of 100%, and an F1-score of 89.69%. The system detected all scenario-based anomalies and displayed a WARNING status as an early indication. However, further inspection by an operator or technician remains necessary because the anomaly data used in the evaluation were not obtained from actual faults.

Item Type:	Thesis (Other)
Uncontrolled Keywords:	Isolation Forest, deteksi anomali, arus motor, beban mekanik, lift kargo, anomaly detection, motor current, mechanical load, cargo lift
Subjects:	T Technology > TK Electrical engineering. Electronics Nuclear engineering > TK2785 Electric motors, Induction. T Technology > TK Electrical engineering. Electronics Nuclear engineering > TK7870.23 Reliability. Failures T Technology > TK Electrical engineering. Electronics Nuclear engineering > TK7871.674 Detectors. Sensors T Technology > TK Electrical engineering. Electronics Nuclear engineering > TK7882.P3 Pattern recognition systems
Divisions:	Faculty of Vocational > 36304-Automation Electronic Engineering
Depositing User:	Okky Surya Pamungkas
Date Deposited:	05 Aug 2026 09:13
Last Modified:	05 Aug 2026 09:13
URI:	http://repository.its.ac.id/id/eprint/144085

Actions (login required)

View Item