Salsabila, Berliana Khansa (2023) Optimasi Proses Dosing Koagulan pada Instalasi Pengolahan Air Minum Menggunakan Metode Jaringan Saraf Tiruan Berbasis Propagasi Balik. Masters thesis, Institut Teknologi Sepuluh Nopember.
![[thumbnail of 6014221051-Master_Thesis.pdf]](http://repository.its.ac.id/style/images/fileicons/text.png) |
Text
6014221051-Master_Thesis.pdf - Accepted Version Restricted to Repository staff only until 1 April 2025. Download (4MB) | Request a copy |
Abstract
Kebutuhan air minum semakin lama semakin meningkat. Instalasi Pengolahan Air Minum (IPAM) dituntut untuk dapat menyediakan air minum dengan mengolah air baku dari berbagai sumber, salah satunya air sungai, menjadi air bersih yang memiliki kualitas yang baik dan aman untuk dikonsumsi masyarakat. Pengolahan air minum konvensional yang melibatkan unit koagulasi sangat banyak diterapkan di Indonesia. Pada unit koagulasi, terdapat proses pembubuhan koagulan dan pengadukan cepat. Pembubuhan koagulan pada IPAM konvensional mengandalkan proses jartest untuk menentukan dosis koagulan yang mana memakan waktu lama sehingga untuk kondisi air dengan variabilitas tinggi, menyebabkan ketidakakuratan dalam penyesuaian dosis di IPAM. Penelitian ini bertujuan untuk mengoptimasi proses pembubuhan dosis koagulan yang sesuai dengan kualitas air baku, dilihat dari lima parameter utama yang sangat mempengaruhi proses koagulasi, yaitu kekeruhan, pH, suhu, warna, dan konduktivitas listrik. Penelitian dilakukan dengan mengumpulkan data primer dan data sekunder penentuan dosis koagulan berdasarkan kondisi air yang diolah. Data diolah dengan metode jaringan saraf tiruan multi-layer dengan algoritma adaptasi-backpropagation menggunakan aplikasi MATLAB. Penelitian ini juga menguji statistik tingkat signifikansi kelima parameter kualitas air. Hasil pengujian menunjukkan signifikansi paling besar secara berurutan yakni parameter kekeruhan, warna, konduktivitas listrik, pH, dan suhu. Dilakukan dua permodelan dengan salah satunya menghilangkan parameter suhu dan pH. Untuk model dengan 5 input parameter, arsitektur jaringan yang paling cocok dengan data terdiri dari 5 input layer, 6 hidden layer, dan 1 output layer. Sedangkan arsitektur jaringan dengan 3 input parameter terdiri dari 3 input layer, 4 hidden layer, dan 1 output layer. Input layer berisi nilai parameter kualitas air. Output layer berisi dosis optimum koagulan yang diperoleh dari hasil jartest. Jenis koagulan yang digunakan adalah alumunium sulfat dengan konsentrasi 1%. Hasil penelitian ini berupa pemodelan empiris dosis optimum koagulan. Jaringan saraf tiruan yang melibatkan 5 parameter kualitas air menghasilkan model sederhana dosis koagulan = (kekeruhan x 0.298) - (suhu x 0.138) + (pH x 0.134) + (warna x 0.165) + (konduktivitas listrik x 0.163). Jaringan saraf tiruan dengan 3 parameter kualitas air menghasilkan model dosis koagulan = (kekeruhan x 0.306) + (warna x 0.159) +(konduktivitas listrik x 0.151)
==================================================================================================================================
Fulfillment of drinking water needs is a basic human right. Drinking Water Treatment Plants (DWTPs) are required to be able to provide drinking water by treating raw water from various sources, one of which is river water, into clean water that has good quality and is safe for public consumption. Conventional drinking water treatment involving coagulation units is widely applied in Indonesia. In the coagulation process, coagulant addition and rapid mixing are carried out. Coagulant application in conventional water treatment plants relies on the jartest process to determine the coagulant dosage which is time consuming, so that if water conditions have high variability, it will cause inaccuracies in dose adjustments at DWTPs. This research aim to optimize the process of coagulant dosing according to the quality of incoming raw water, seen from five main parameters that greatly affect the coagulation process, i.e turbidity, pH, temperature, color, and electrical conductivity. The research was conducted by collecting primary and secondary data to determine coagulant dosage based on the condition of the treated water. The data is processed using a multi-layer artificial neural network with an adaptationbackpropagation algorithm using the MATLAB software. This study also tested the level of significance of the five water quality parameters. The test results showed the greatest significance sequentially are turbidity, color, electrical conductivity, pH, and temperature. Two models were carried out with one of them excluding temperature and pH parameters. For a model with 5 input parameters, the network architecture that best fits the data consists of 5 input layers, 6 hidden layers, and 1 output layer. While the network architecture with 3 input parameters consists of 3 input layers, 4 hidden layers, and 1 output layer. The input layer contains water quality parameter values. The output layer contains the optimum coagulant dose obtained from the results of the jartest. The type of coagulant used was aluminum sulfate with a concentration of 1%. The results of this study are empirical modeling of the optimum coagulant dosage. An artificial neural network involving 5 water quality parameters produces a simple model of coagulant dosage = (turbidity x 0.298) - (temperature x 0.138) + (pH x 0.134) + (color x 0.165) + (electrical conductivity x 0.163). Artificial neural network with 3 water quality parameters produces a model of coagulant dosage = (turbidity x 0.306) + (color x 0.159) +(electrical conductivity x 0.151)
| Item Type: | Thesis (Masters) |
|---|---|
| Uncontrolled Keywords: | artificial neural network, coagulant dosage, coagulant dose optimization, coagulation process, drinking water treatment, dosis koagulan, jaringan saraf tiruan, koagulasi, optimasi dosis koagulan, pengolahan air minum |
| Subjects: | Q Science > QA Mathematics > QA76.87 Neural networks (Computer Science) T Technology > TD Environmental technology. Sanitary engineering > TD259.2 Drinking water. Water quality T Technology > TD Environmental technology. Sanitary engineering > TD433 Water treatment plants T Technology > TD Environmental technology. Sanitary engineering > TD455 Chemical precipitation. Coagulation. Flocculation. Water--Purification--Flocculation. |
| Divisions: | Faculty of Civil, Planning, and Geo Engineering (CIVPLAN) > Environmental Engineering > 25101-(S2) Master Thesis |
| Depositing User: | Berliana Khansa Salsabila Salsabila |
| Date Deposited: | 27 Jan 2023 06:35 |
| Last Modified: | 27 Jan 2023 06:35 |
| URI: | http://repository.its.ac.id/id/eprint/95519 |
Actions (login required)
 |
View Item |