Sistem Kendali Underwater Remotely Operated Vehicles Pada Tingkat Kedalaman Air Menggunakan Metode Kendali Adaptif

Abidin, Ali Zainal (2018) Sistem Kendali Underwater Remotely Operated Vehicles Pada Tingkat Kedalaman Air Menggunakan Metode Kendali Adaptif. Masters thesis, Institut Teknologi Sepuluh Nopember.

[img] Text
07111450040010-Master_Thesis.pdf - Published Version
Restricted to Repository staff only

Download (5MB) | Request a copy

Abstract

Pada tingkat kedalaman air tertentu, kecepatan serta arah arus air dapat berubah tidak teratur dan memiliki tekanan yang berbeda. Faktor tersebut merupakan parameter yang mempengaruhi stabilisasi underwater ROV (remotely operated vehicle) pada saat di dalam air. Underwater ROV harus dapat bertahan (hold position) pada tingkat kedalaman air dan terhadap gangguan (perubahan arus air) yang ada di dalam air. Dalam menghadapi masalah tersebut, digunakan sistem kendali adaptif karena memiliki parameter kendali yang dapat beradaptasi terhadap gangguan (disturbance). Pada penelitian ini, sistem kendali adaptif dirancang menggunakan neural network dan sistem kendali PID (Propotional Integral Derivative). Sistem kendali PID digunakan sebagai kontrol gerak underwater ROV, sedangkan neural network difungsikan sebagai identifikasi terhadap gangguan yang sudah dikenali (learning) terlebih dahulu, hasil identifikasi digunakan untuk menentukan nilai parameter dari sistem kendali PID. Nueral network terdiri dari 3 layer (input, hidden, dan output), input layer memiliki 3 neuron (error, integral error dan derivative error), hidden layer memiliki 8 neuron dan output layer memiliki 3 neuron (Kp, Ki dan Kd). Komputansi sistem kendali NN berbasis mikrokontroler arduino Nano 328. Pengujian dilakukan dengan memberi arus air buatan yang berasal dari Bilge Pump 1100GPH, menghasilkan gaya dorong arus atas dan bawah sebagai gangguan kepada underwater ROV dengan besar nilai yang bervariasi. Pada penelitian dilakukan pengujian terhadap sistem kendali adaptif NN-PID. Pada gangguan atas memiliki nilai overshoot 20%, rise time 0,8 detik, settling time 2,4 detik dan error steady state (ESS) 6,50%. Pada gangguan bawah, overshoot 20%, rise time 2,0 detik, settling time 4,6 detik dan ESS 6,45%. Hasil pengujian underwater ROV yang dilakukan terhadap gangguan yang berubah-ubah menunjukkan sistem kendali NN-PID dapat beradaptasi dengan baik. ============ At certain water depth levels, the velocity and direction of the water current may change irregularly and have different pressure. These factors are parameters that affect the navigation and stabilization of underwater ROV (remotely operated vehicle) in the water. Underwater ROV must be able to maintain a position (hold position) at the level of water depth and the disturbance (changes in water flow) in the water. In the face of the problem, an adaptive control system is used because it has a control parameter that can adapt to disturbance. In this research, built NN-PID adaptive control system of two control system, that is between PID control system (Propotional Integral Derivative) and Neural Network control system. PID control functions as control over underwater motion of ROV, while Neural Network control is used as identification of previously recognized disturbances and the output value of the system is used as a tunning parameter of the PID control system. The Nueral Network system consists of 3 layers (input, hidden, and output), the input layer has 3 neurons (error, integral error and derivative error), the hidden layer has 8 neurons and the output layer has 3 neurons (Kp, Ki and Kd). Computing NN control system based on arduino Nano 328 microcontroller. The test is carried out by giving artificial water flow from the Bilge Pump 1100GPH, resulting in an upper and lower current thrust as a disturbance to the underwater ROV with varying values. In the experiment, we tested the NN-PID adaptive control system. At the top disturbance has a 20% overshoot value, 0.8 second rise time, 2,4 seconds settling time and error steady state (ESS) of 6,50%. On the downside, overshoot 20%, rise time 2,0 seconds, settling time 4,6 seconds and ESS 6,45%. The results of underwater ROV testing performed against the mutable disturbances indicate the NN-PID control system can adapt well.

Item Type: Thesis (Masters)
Uncontrolled Keywords: hold position, NN-PID, neural network, underwater ROV (remotely operated vechile).
Subjects: T Technology > TJ Mechanical engineering and machinery > TJ211 Robotics.
T Technology > TJ Mechanical engineering and machinery > TJ213 Automatic control.
T Technology > TJ Mechanical engineering and machinery > TJ223 PID controllers
Divisions: Faculty of Electrical Technology > Electrical Engineering > 20101-(S2) Master Thesis
Depositing User: Ali zainal abidin
Date Deposited: 06 Aug 2018 08:30
Last Modified: 06 Aug 2018 08:30
URI: http://repository.its.ac.id/id/eprint/55459

Actions (login required)

View Item View Item