Setiawan, Erik Tri Yudha (2019) Desain Soft Sensor Konduktivitas Elektrik Terhadap pH Pada Tangki Nutrisi Hidroponik. Masters thesis, Institut Teknologi Sepuluh Nopember.
Text
02311445000037-Undergraduate_Theses.pdf - Accepted Version Restricted to Repository staff only until 1 October 2022. Download (2MB) | Request a copy |
Abstract
Tangki nutrisi hidroponik merupakan sebuah tangki yang berfungsi sebagai tempat terjadinya reaksi larutan nutrisi untuk tanaman pada sistem hidroponik. Salah satu larutan nutrisi yang digunakan adalah larutan nutrisi AB mix. Beberapa besaran yang berpengaruh terhadap kualitas larutan nutrisi AB mix diantaranya konduktivitas elektrik (EC) dan pH. Soft sensor didesain untuk memudahkan pembacaan nilai konduktivitas elektrik sehingga mendapatkan nilai yang lebih cepat dan efisien dengan menggunakan metode jaringan syaraf tiruan. Soft sensor ini memiliki satu variabel masukan berupa pH dan satu variabel keluaran berupa konduktivitas elektrik. Variasi yang diberikan berupa jumlah node sebesar 5, 10, dan 15 serta variasi algoritma pembelajaran berupa gradient descent with momentum and adaptive learning rate (traingdx) dan algoritma Levenberg Marquardt (trainlm). Hasil menunjukkan bahwa perubahan jumlah node pada jaringan syaraf tiruan akan mempengaruhi nilai bobot dan bias yang dihasilkan. Algoritma pembelajaran trainlm mempunyai iterasi yang lebih sedikit dibandingkan algoritma traingdx.
Kata kunci : konduktivitas elektrik, jaringan syaraf tiruan, pH, tangki nutrisi hidroponik
===============================================================================================
The hydroponic nutrient tank is a tank that functions as a reaction place for nutrient solutions for plants in hydroponic systems. One nutrient solution used is the AB mix nutrient solution. Some quantities that affect the quality of the AB mix nutrient solution include electrical conductivity (EC) and pH. Soft sensors are designed to facilitate reading of electrical conductivity values so that they get a faster and more efficient value using the artificial neural network method. This soft sensor has one input variable in the form of pH and one output variable in the form of electrical conductivity. The variations given in the form of the number of nodes of 5, 10, and 15 and variations of the learning algorithm in the form of gradient descent with momentum and adaptive learning rate (traingdx) and Levenberg Marquardt algorithm (trainlm). The results show that changes in the number of nodes in artificial neural networks will affect the value of the weight and bias produced. The trainlm learning algorithm has fewer iterations than the traingdx algorithm.
Keyword: artificial neural network, electrical conductivity, pH, hydroponic nutrient tank
Item Type: | Thesis (Masters) |
---|---|
Additional Information: | RSF 629.89 Set d-1 2019 |
Uncontrolled Keywords: | konduktivitas elektrik, jaringan syaraf tiruan, pH, tangki nutrisi hidroponik |
Subjects: | T Technology > T Technology (General) T Technology > TK Electrical engineering. Electronics Nuclear engineering |
Divisions: | Faculty of Industrial Technology and Systems Engineering (INDSYS) > Physics Engineering > 30201-(S1) Undergraduate Thesis |
Depositing User: | Erik Tri Yudha Setiawan |
Date Deposited: | 12 Jan 2022 02:46 |
Last Modified: | 12 Jan 2022 02:49 |
URI: | http://repository.its.ac.id/id/eprint/61948 |
Actions (login required)
View Item |