Sistem Klasifikasi Kontaminan Pewarna untuk Proteksi Saluran Drainase Menggunakan Artificial Neural Network

Aditama, Febrian Raffles (2026) Sistem Klasifikasi Kontaminan Pewarna untuk Proteksi Saluran Drainase Menggunakan Artificial Neural Network. Other thesis, Institut Teknologi Sepuluh Nopember.

Text 2040221146-Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only Download (3MB) | Request a copy

Abstract

Sistem pembuangan air hujan di PT X saat ini hanya bergantung pada sensor ketinggian air tanpa deteksi kualitas air, sehingga berisiko meloloskan air yang terkontaminasi polutan pewarna industri ke saluran drainase eksternal. Kelemahan inspeksi manual yang lambat ini membutuhkan solusi pemantauan cerdas yang responsif. Penelitian ini bertujuan untuk merancang purwarupa simulasi Sistem Klasifikasi Kontaminan Pewarna menggunakan sensor TCS34725 dan algoritma Artificial Neural Network (ANN). Sistem ini bekerja dengan mengakuisisi intensitas spektrum warna (RGBC) dari sampel air di dalam tangki sampling terisolasi yang disirkulasikan secara otomatis. Data tersebut kemudian diproses menggunakan metode edge inferencing pada mikrokontroler ESP32 untuk mengklasifikasikan kondisi air ke dalam 10 kelas tingkat keparahan dari pewarna Malachite Green, Methyl Violet, dan Carmoisine Red pada rentang konsentrasi 6,25 ppm hingga 500 ppm. Hasil pengujian menunjukkan bahwa arsitektur model ANN (4-16-4) sukses mengklasifikasikan seluruh sampel uji dengan tingkat akurasi mencapai 100%. Saat terdeteksi kontaminasi yang melewati batas aman, ESP32 secara andal mengirimkan instruksi lockdown ke instrumen PLC Omron via protokol Modbus RS-485 dengan rata-rata latensi sangat rendah sebesar 42,75 ms. Prototipe ini terbukti efektif dalam memutus daya pembuangan secara aktual, menjadikannya sistem proteksi ganda (fail-safe) yang efisien untuk mencegah pencemaran drainase.
====================================================================================================================================
The rainwater drainage system at PT X currently relies solely on water level sensors without water quality detection, thus risking the leakage of water contaminated with industrial dye pollutants into external drainage channels. This slow manual inspection weakness requires a responsive intelligent monitoring solution. This study aims to design a simulation prototype of a Dye Contaminant Classification System using a TCS34725 sensor and an Artificial Neural Network (ANN) algorithm. This system works by acquiring the color spectrum intensity (RGBC) of water samples in an isolated sampling tank that is automatically circulated. The data is then processed using the edge inferencing method on an ESP32 microcontroller to classify water conditions into 10 severity classes of Malachite Green, Methyl Violet, and Carmoisine Red dyes at a concentration range of 6.25 ppm to 500 ppm. The test results show that the ANN model architecture (4-16-4) successfully classifies all test samples with an accuracy level reaching 100%. When contamination exceeding safe limits is detected, the ESP32 reliably sends a lockdown instruction to the Omron PLC instrument via the Modbus RS-485 protocol with an extremely low average latency of 42.75 ms. This prototype has proven effective in actually cutting off the drain power, making it an efficient fail-safe system for preventing drain contamination.

Item Type:	Thesis (Other)
Uncontrolled Keywords:	Kolorimetri, Artificial Neural Network, TCS34725, Simulasi Interlock, Drainase Industri, Colorimetry, TCS34725, Interlock Simulation, Industrial Drainage
Subjects:	Q Science > QA Mathematics > QA76.87 Neural networks (Computer Science) T Technology > T Technology (General) > T55 Industrial Safety T Technology > TA Engineering (General). Civil engineering (General) > TA1573 Detectors. Sensors T Technology > TD Environmental technology. Sanitary engineering > TD420 Water pollution
Divisions:	Faculty of Vocational > 36304-Automation Electronic Engineering
Depositing User:	Febrian Raffles Aditama
Date Deposited:	26 Aug 2026 00:55
Last Modified:	26 Aug 2026 00:55
URI:	http://repository.its.ac.id/id/eprint/144329

Actions (login required)

View Item