Hasil Pencarian  ::  Simpan CSV :: Kembali

"Permasalahan terbesar dalam pengendalian reaktor alir tangki berpengaduk adalah sistem yang sangat tidak linear dan multivariabel.Sistem pengendalian konvensional tidak dapat mengontrol sistem semacam ini dengan optimal, sehingga kemurnian produk yang dihasilkan rendah.Multiple Model Predictive Control (MMPC)digunakan untuk mengatasi masalah pengendalian proses yang nonlinear dan melibatkan banyak variabel. Beberapa MPC lokal digunakan pada MMPC diperoleh dengan metode yang baru dikembangkan, Representative Model Predictive Control (RMPC). Penelitian ini menggunakan model reaktor alir tangki berpengaduk yang disimulasikan dengan perangkat lunak MATLAB. Variabel yang dimanipulasi adalah suhu inlet pendingin dan konsentrasi umpan sedangkan variabel yang dikontrol adalah komposisi produk. Untuk perubahan set point konsentrasi produk dari 8,5 sampai 8,6; disarankan menggunakan MMPC 4,1,2.

---

The biggest problem in controlling Continuous Stirred Tank Reactor (CSTR) is nonlinearity in the system. Conventional control system can not optimally control this system, therefore decrease the purity of product. Multiple Model Predictive Control (MMPC), that can be used to control nonlinear and multivariable system, tried to be used on this system. Some local MPC used for MMPC based on new developed method, Representative Model Predictive Control (RMPC).

This thesis using CSTR model which is simulated by MATLAB software. The manipulated variable are cooler inlet temperature and feed concentration, and controlled variable is residual concentration. For the change of residual concentration set point from 8.5 to 8.6 change, the MMPC 4,1,2. is recommended."

"ABSTRAKAdanya ketidaklinearan pada reaktor alir tangki berpengaduk mampu menyebabkan gangguan ketika proses sedang berjalan. Gangguan tersebut menyebabkan turunnya kualitas produk, sehingga diperlukan penanganan terhadap gangguan. Skripsi ini membahas penggunaan Representative Model Predictive Control (RMPC) dalam memilih beberapa model predictive control (MPC) lokal yang kemudian dikombinasikan untuk membuat Multi Model Predictive Control (MMPC), dan digunakan untuk menangani gangguan pada proses. Penelitian ini menggunakan model reaktor Bequette dan disimulasikan menggunakan perangkat lunak MATLAB. Variabel bebasnya adalah konsentrasi feed sedangkan variabel kontrolnya adalah konsentrasi produk dan suhu reaktor. Hasil dari penelitian menunjukkan IAE MMPC lebih kecil dari IAE PI.

---

ABSTRACTExisting nonlinearity in continuous stirred tank reactor can cause disturbances when the process is running. Those disturbances cause a decline in product quality, so that disturbances rejection control is needed. The use of Representative Model Predictive Control (RMPC) in selecting some of the local Model Predictive Control (MPC) and then combined to make Multi Model Predictive Control (MMPC) are discussed and explained. MMPC, a Bequette reactor model, and MATLAB software were used and applied to handle disturbances and simulate. Manipulated variable is feed concentration while the controlled variables are product concentration and reactor temperature. The results of this study show that IAE value of MMPC is smaller than IAE value of PI."

"Untuk menangani gangguan pada proses operasi nonlinear diperlukan suatu bentuk pengendalian. Representative Model Predictive Control (RMPC) adalah salah satu cara untuk memperoleh sekumpulan MPC lokal yang dapat merepresentasikan keseluruhan rentang operasi. MPC lokal ini nantinya digunakanpada Multiple Model Predictive Control (MMPC) untuk mensimulasikanproses operasi nonlinear multi variabel.Skripsi ini membahas penggunaan RMPC untuk memilih beberapa MPC lokal yang kemudian digunakan sebagai model pada MMPC untuk menangani gangguan. Penelitian ini menggunakan model kolom distilasi biner ?Kolom A? yang disimulasikan dengan perangkat lunak MATLAB. Variabel yang dimanipulasi adalah laju refluks dan laju boil up sedangkan variabel yang dikontrol adalah komposisi produk distilat dan komposisi produk bawah. Hasil IAE MMPC dibandingkan dengan IAE kontroler PI konvensional. Untuk gangguan single step; MPC terbaik dengan IAE 0,2564, lebih baik dari IAE kontroler PI 0,7494.Sedangkan untuk gangguan multi step; MMPC terbaik dengan IAE 0,7730, lebih baik dari IAE kontroler PI 0,9808.

---

In order to handle disturbances in the nonlinear operation some form of control is required. Representative Model Predictive Control (RMPC) is one way to obtain a set of local MPC which able to represent the entire operating range. The local MPC is later used in the Multiple Model Predictive Control (MMPC) to simulate the operation of nonlinear multi-variable process. This thesis discusses the use of RMPC to select some local MPC which is then used as a model for dealing with disturbances in the MMPC. This study uses a model of a binary distillation column "Column A" which is simulated with MATLAB software. The manipulated variable is the rate of reflux and boil-up rate, while the controlled variable is the product composition of the distillate and bottom product composition. MMPC IAE results compared with conventional PI controller IAE. For single step disturbance; the best MPC with IAE 0.2564, is better than PI controller IAE 0.7494. As for the multi-step disturbance; the best MMPC with IAE 0.7730, is better than PI controller IAE 0.9808."

"Optimalisasi pabrik regasifikasi liqufied natural gas LNG penting dilakukan untuk meminimilasi biaya, khususnya biaya operasional. Oleh karena itu penting untuk memilih desain pabrik regasifikasi LNG dan mendapatkan kondisi operasi yang optimum serta mempertahankan kondisi operasi yang optimum tersebut melalui implementasi model predictive control MPC. Kriteria optimalnya adalah minimumnya jumlah energi yang digunakan dan atau integral of square error ISE. Hasilnya, disain yang optimum adalah menggunakan skema 2 dengan penghematan energi sebesar 40. Sedangkan kondisi operasi yang optimum terjadi jika suhu keluaran vaporizer sebesar 6oC. Untuk mempertahankan kondisi optimum tersebut diperlukan MPC dengan setelan parameter P prediction horizon , M control horizon dan T sampling time sebagai berikut: pengendali tekanan tangki penyimpanan: 90, 2, 1; tekanan produk: 95, 2, 1; suhu vaporizer: 65, 2, 2; dan suhu heater: 35, 6, 5, dengan nilai ISE pada set point tracking masing-masing 0,99, 1792,78, 34,89 dan 7,54, atau peningkatan kinerja pengendalian masing-masing sebesar 4,6 , 63,5 , 3,1 dan 58,2 dibandingkan kinerja pengendali PI. Penghematan energi yang dapat dilakukan pengendali MPC saat terjadi gangguan pada kenaikan suhu air laut 1oC adalah 0,02 MW dan pengendali MPC juga mengurangi error terhadap kualitas produk sebesar 34,25 dibandingkan dengan menggunakan pengendali PI.

---

Optimization of liquified natural gas LNG regasification plant is important to minimize costs, especially operational costs. Therefore, it is important to select the LNG regasification plant design and obtain optimum operating conditions while maintaining the optimum operating conditions through the implementation of model predictive control MPC. The optimal criterion is the minimum amount of energy used and or the integral of square error ISE.

As a result, the optimum design is to use scheme 2 with an energy savings of 40 . While the optimum operating conditions occur if the vaporizer output temperature is 6oC. In order to maintain the optimum conditions, MPC is required with parameter setting P prediction horizon, M control horizon and T sampling time as follows tank storage pressure controller 90, 2, 1 product pressure 95, 2, 1 temperature vaporizer 65, 2, 2 and temperature heater 35, 6, 5, with ISE value at set point tracking respectively 0.99, 1792.78, 34.89 and 7.54, or improvement of control performance respectively 4.6, 63.5 , 3.1 and 58.2 compared to PI controller performance.

The energy savings that MPC controllers can make when there is a disturbance in sea temperature rise of 1oC is 0.02 MW and MPC controller also reduces error to product quality by 34.25 compared to the PI controller."

"Recent developments in model-predictive control promise remarkable opportunities for designing multi-input, multi-output control systems and improving the control of single-input, single-output systems. This volume provides a definitive survey of the latest model-predictive control methods available to engineers and scientists today.The initial set of chapters present various methods for managing uncertainty in systems, including stochastic model-predictive control. With the advent of affordable and fast computation, control engineers now need to think about using “computationally intensive controls,” so the second part of this book addresses the solution of optimization problems in “real” time for model-predictive control. The theory and applications of control theory often influence each other, so the last section of Handbook of Model Predictive Control rounds out the book with representative applications to automobiles, healthcare, robotics, and finance.The chapters in this volume will be useful to working engineers, scientists, and mathematicians, as well as students and faculty interested in the progression of control theory. Future developments in MPC will no doubt build from concepts demonstrated in this book and anyone with an interest in MPC will find fruitful information and suggestions for additional reading."

"Model Predictive Control is an important technique used in the process control industries. It has developed considerably in the last few years, because it is the most general way of posing the process control problem in the time domain. The Model Predictive Control formulation integrates optimal control, stochastic control, control of processes with dead time, multivariable control and future references. The finite control horizon makes it possible to handle constraints and non linear processes in general which are frequently found in industry. Focusing on implementation issues for Model Predictive Controllers in industry, it fills the gap between the empirical way practitioners use control algorithms and the sometimes abstractly formulated techniques developed by researchers. The text is firmly based on material from lectures given to senior undergraduate and graduate students and articles written by the authors"

"Model Predictive Control is an important technique used in the process control industries. It has developed considerably in the last few years, because it is the most general way of posing the process control problem in the time domain. The Model Predictive Control formulation integrates optimal control, stochastic control, control of processes with dead time, multivariable control and future references. The finite control horizon makes it possible to handle constraints and non linear processes in general which are frequently found in industry. Focusing on implementation issues for Model Predictive Controllers in industry, it fills the gap between the empirical way practitioners use control algorithms and the sometimes abstractly formulated techniques developed by researchers. The text is firmly based on material from lectures given to senior undergraduate and graduate students and articles written by the authors"

"Model Predictive Control (MPC) merupakan sistem pengendalian yang menggunakan model berdasarkan data hasil pengukuran keluaran (output) saat ini atau masa sebelumnya untuk memprediksi nilai dari variabel proses (input) pada masa yang akan datang. Pada penelitian ini, sistem pengendalian MPC digunakan untuk menangani pengendalian proses variabel jamak dalam unit operasi Continous Stirred Tank Reactor (CSTR) dengan reaksi pembuatan propylene glycol. Model dinamik sesuai dengan kondisi operasi yang dapat mewakili interaksi antara variabel jamak dibuat untuk diterapkan pada sistem pengendali. Sistem pengendalian proses disimulasikan dengan menggunakan perangkat lunak Unisim R390.1. Simulasi pengendalian proses dilakukan untuk menghasilkan performa pengendalian yang optimum dan untuk mengendalikan variable jamak yang saling berinteraksi dalam sistem pada CSTR. Optimasi pada sistem pengendalian dilakukan dengan cara tuning terhadap parameter-parameter MPC seperti model horizon (N), waktu sampel (T), prediction horizon (P), dan control horizon (M).Hasil dari simulasi menunjukkan Model F sebagai model dinamik terbaik pada pengendali MPC multivariable mampu menangani jangkauan perubahan setpoint dalam rentang perubahan yang kecil dari 0,33 ke 0,331 dengan IAE sebesar 0,10602. Secara keseluruhan, pengendali MPC belum dapat mengendalikan sistem CSTR secara optimum berdasarkan nilai IAE, namun pengendali MPC lebih mampu menjaga kestabilan sistem dibandingkan dengan pengendali PI.Model Predictive Control (MPC) are control system which use model based on value output variable at present or past to predict value of future process variable. In this research, MPC control system use to handle multivariable process control in unit operation Continous Stirred Tank Reactor (CSTR) with propylene glycol reaction system. Dynamics model based on operating condition which representative interaction between multivariable are made to implement in control system. Process control system simulating in Unisim R390.1 software. The simulation of process control aims to achieve optimum performance of controller and to control interaction between multivariable in CSTR system. Optimasion will be doing in system control with MPC parameters tuning such as model horizon (N), time sampling (T), prediction horizon (P), and control horizon (M).

The Results show that Model F as the best model in MPC multivariable can control the change of setpoint in short length from 0,33 to 0,331 with 0,10602 IAE. Overall, MPC controller can?t controlled CSTR system with optimum result based on IEA value, but MPC can make system more stabile than PI controller."

"Penilitian ini meninjau kinerja pengendalian MPC dengan model empirik Auto-Regressive Exogenous pada proses produksi formaldehid di PT. X. MPC digunakan untuk mengendalikan laju alir umpan steam, tekanan evaporator, temperatur udara, dan ketinggian cairan evaporator. Hasil dari penelitian ini dibandingkan dengan penelitian Wahid dan Salman (2020) yang menggunakan pengendali MPC dengan model FOPDT. Kinerja pengendalian diukur menggunakan parameter IAE dan ISE dengan dua jenis pengujian, yaitu set-point tracking dan disturbance rejection. Hasil menunjukkan bahwa model yang diidentifikasi memiliki nilai fit to estimation lebih dari 95% dan mampu merepresentasikan data aktual dengan nilai kesalahan (RMSE) lebih kecil daripada model FOPDT. Parameter pengendali yang optimum secara berurutan (T, P, dan M) adalah (10,10,2) untuk FIC-102, (10,30,2) untuk PIC-101, (10,10,2) untuk TIC-101, dan (10,60,2) untuk LIC-101. Terdapat perbaikan kinerja pengendalian berdasarkan parameter IAE dan ISE, pada uji SP tracking sebesar 86,63% dan 85,56% untuk FIC-102, 74,36% dan 87,28% untuk PIC-101, 56,27% dan 20,45% untuk TIC-101, serta 87,35% dan 84,65% untuk LIC-101. Sedangkan untuk disturbance rejection perbaikannya sebesar 95,85% dan 96,75% untuk FIC-102, 85,95% dan 96,81% untuk PIC-101, 43,06% dan -30,0% pada TIC-101, serta -85,06% dan -539,13% pada LIC-101. Berdasarkan hasil penelitian, pengendali dengan model ARX memberikan kinerja pengendalian yang lebih baik karena dapat merepresentasikan proses aktual secara lebih akurat.

---

This study reviews the performance of MPC control with the Auto-Regressive Exogenous empirical model in the formaldehyde production process at PT. X. MPC is used to control the steam feed flow rate, evaporator pressure, air temperature, and evaporator liquid level. The results of this study were compared with the research of Wahid and Salman (2020) which used MPC controllers with the FOPDT model. Control performance is measured using IAE and ISE parameters with two types of tests, namely set-point tracking and disturbance rejection. The results show that the identified model has a fit to estimation value of more than 95% and is able to represent actual data with an error value (RMSE) smaller than the FOPDT model. The optimum control parameters sequentially (T, P, and M) are (10,10,2) for FIC-102, (10,30,2) for PIC-101, (10,10,2) for TIC-101 , and (10,60,2) for LIC-101. There is an improvement in control performance based on IAE and ISE parameters, in the SP tracking test of 86.63% and 85.56% for FIC-102, 74.36% and 87.28% for PIC-101, 56.27% and 20,45% on TIC-101, and 87.35% and 84.65% on LIC-101. Meanwhile, for disturbance rejection, the improvements are 95.85% and 96.75% for FIC-102, 85.95% and 96.81% for PIC-101, 43.06% and -30.0% on TIC-101, and -85.06% and -539.13% on LIC-101. Based on the research results, the controller with the ARX model provides better control performance because it can represent the actual process more accurately."