IDENTIFIKASI PROSES DALAM LUP TERTUTUP KONTROL TEMPERATUR PRODUK ATAS PADA PLANT AMMONIA STRIPPER DENGAN MENGGUNAKAN ALGORITMA LUUS JAAKOLA

• Main Author: ANGGANA, GIAN
• Format: Final Project
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/10494
• Institution: Institut Teknologi Bandung

In industrial application, accurate mathematical model is needed to design a good controller. System identification is a method to construct mathematical model from a dynamic process using measured data. In this research, a non-linear closed loop system has been identified using two approaches, direct approach and joint input-output approach. Closed loop identification needs different methods from open loop identification. In this research, Luus Jaakola algorithm is used for closed loop identification. Luus Jaakola algorithm is usually used for optimization problems, but in this research it is used for system identification. In this research, ARX, Hammerstein and neural network are the models that are used in system identification. Ammonia Stripper which is located in PT. XYZ is a closed loop system that is used in this research. This plant is a part of ammonia plant that is used to separate ammonia and water. The results of closed loop identification using Luus Jaakola algorithm show that this algorithm is a good method for system identification. RMSE from identification and validation phase show a small value, 0.011 for ARX model, 0.015 for Hammerstein model and 0.012 for neural network model.