Prediksi Kendali pada Furnace dengan Menggunakan Neural Network Model Predictive Control (NNMPC) di PPSDM Migas Cepu

Danandyo, Dimas (2022) Prediksi Kendali pada Furnace dengan Menggunakan Neural Network Model Predictive Control (NNMPC) di PPSDM Migas Cepu. Undergraduate thesis, Institut Teknologi Sepuluh Nopember.

[img] Text
02311740000087-Undergraduate_Thesis.pdf - Accepted Version
Restricted to Repository staff only until 1 April 2024.

Download (2MB) | Request a copy

Abstract

Crude oil atau minyak mentah merupakan bahan dasar untuk olahan produk minyak bumi yang diproses melalui proses pemisahan fraksi distilasi di kolom distilasi pada kilang sehingga didapatkan minyak bumi olahan sesuai dengan spesifikasi yang dibutuhkan, di mana proses pengolahannya sendiri terbagi menjadi beberapa bagian antara lain pemanasan, pemisahan, penguapan, pengembunan, dan pendinginan. Crude oil dipanaskan di dalam furnace dengan temperatur sebagai variabel yang paling penting yang mana akan sangat menentukan kualitas dari olahan hasil produksi. Jaringan syaraf tiruan sendiri digunakan untuk merancang neural network model predictive control (NNMPC) pada penelitian tugas akhir ini untuk memprediksi kinerja dan performansi sistem kendali yang ideal karena kelebihannya dalam melakukan pembelajaran dari data input dan output suatu proses. Pada penelitian yang dilakukan, NNMPC menunjukkan performa yang lebih baik ditinjau dari karena rise time, settling time, peak time yang lebih cepat; serta error steady state (ESS) yang lebih rendah dibandingkan dengan pengendali PI yang digunakan sebagai pembanding langsung. ================================================================================================ Crude oil is the raw material for the processing of petroleum products which is processed through a distillation separation process at the refinery so that refined petroleum is obtained according to the required specifications, where the processing process itself is divided into several parts including heating, separation, evaporation, condensation, and cooling. Crude oil in the furnace run the process with temperature as the most important variable which will greatly determine the quality of the processed products. The artificial neural network itself is used to design a neural network model predictive control (NNMPC) in this undergraduate thesis research to predict the performance of an ideal control system because of its advantages in learning from input and output data of a process. In the research conducted, NNMPC showed better performance quickly in terms of rising time, settling time, and peak time; also, lower steady state error (ESS) than the PI controller which is used as a direct comparison.

Item Type: Thesis (Undergraduate)
Uncontrolled Keywords: Furnace, Artificial Neural Network, NNMPC
Subjects: Q Science > QA Mathematics > QA76.87 Neural networks (Computer Science)
T Technology > TJ Mechanical engineering and machinery > TJ217.6 Predictive Control
Divisions: Faculty of Industrial Technology and Systems Engineering (INDSYS) > Physics Engineering > 30201-(S1) Undergraduate Thesis
Depositing User: Dimas Danandyo
Date Deposited: 22 Feb 2022 01:28
Last Modified: 22 Feb 2022 01:28
URI: https://repository.its.ac.id/id/eprint/94694

Actions (login required)

View Item View Item