Challenging in Instrument and Process Control on Oil and Gas Industry
SISTEM KONTROL PREDIKTIF (MODEL PREDICTIVE CONTROL) SEBAGAI SALAH SATU ALTERNATIF SISTEM KONTROL PADA INDUSTRI MINYAK - GAS
Bambang L. Widjiantoro
JURUSAN TEKNIK FISIKA – FAK. TEKNOLOGI INDUSTRI INSTITUT TEKNOLOGI SEPULUH NOPEMBER (ITS) 2008 1
Karakteristik Proses Umumnya proses di iindustri U d t i memiliki iliki karakteristik k kt i tik sebagai berikut : • Kompleks • Nonlinier N li i • Memiliki waktu tunda • Unstable U t bl open lloop • Nonminimum phase
2
Sistem Kontrol Berbasis Model Nonlinear model based control is an integrated approach for process analysis, l i control t l and d optimization ti i ti where h th the same steady t d state, t t nonlinear, process model is used at each stage. The use of this type of predict the control action required q to meet the control model to p objectives can be expected to provide improved performance over simple, linear models as the model gives more precise information about the effects of manipulated variables on the controlled variables variables.
*) Anshari R.M, Nonlinear Model Based Process Control – Applications in Petroleum Refining, Springer 2000
3
Konsep Sistem Kontrol Prediktif Trayektori acuan Set point point, r Output model, y ˆ (k)
Output proses, proses y(k)
Sinyal y kontrol,, u(k) ( )
t k + N1
t k + Nu
tk + N2 4
Keuntungan Sistem Kontrol Prediktif Predictive Control/Model Predictive Control is the only advanced control technique – that is, more advanced than standard PID control To have had a significant and widespread impact on industrial process control. t l Th The penetration t ti off predictive di ti control t l into i t industrial i d t i l practice ti has also been helped by the facts that y g idea is easy y to understand • Its underlying • Its basic formulation extends to multivariable plants with almost no modification. • It is more powerful than PID control control, even for single loops without constraints, even on difficult loops such as those containing long time delays.
*) Maciejowski, Predictive Control with Constraints, Prentice Hall 2002 5
Diagram Blok Sistem Kontrol Prediktif
Kontroler
r(t)
Optimasi
+
u(t)
y(t) Proses
Model
yˆ (t + N 1 ) K yˆ (t + N 2 )
+
6
Surveyy Sistem Kontrol Prediktif
Zhu Y (1998), Multivariable process identification for MPC : the asymptotic method and its application, Journal of Process Control.
Qin S.J, Badgwell T.A (2003), A Survey of Industrial Model Predictive Control Technology, Control Engineering Practice.
Terdapat lebih dari 2200 aplikasi sistem kontrol prediktif linier di industri dengan g 70% p penerapannya p y p pada industri refining g dan petrokimia
Hussain M.A (1999), Review of the application of neural networks in chemical process control – simulatio and online implementation, Artificial Intelligent in Engineering
MPC ttechnology h l h has b been widely id l and d successfully f ll applied li d iin th the refinery and petrochemical industry.
Terdapat banyak applikasi MPC pada proses kimia termasuk kolom distilasi distilasi, CSTR CSTR, dll
Chandra PVS, Model Predictive Control of Ethyl Acetate Reactive Distillation Column and Engell S (2000), Neural networks for modelling and control of reactive distillation column
Aplikasi MPC pada kolom distilasi
7
Survey Kontrol Prediktif
Zanin AC (2002), Integrating real time optimization into the model predictive controller of the FCC system, system Control Engineering Practice
Algoritma optimasi pada MPC untuk FCC sistem
Wang (2004), W (2004) A new robust b t model d l predictive di ti control t l method, th d Journal of Process Control
Algoritma g baru pada MPC
8
Prediktif Kontrol di Teknik Fisika FTI – ITS (1)
Short research di Department p of Chemical Engineering, g g Univ. Dortmund Germany Sponsored by DAAD Design of Neural Networks Based Nonlinear Multivariable MPC on Reactive Distillation Column
9
Prediktif Kontrol di Teknik Fisika FTI – ITS (2) condenser
reflux
separation zone distillate tank
feed
reaction zone
reboiler
heat supply
10
Prediktif Kontrol di Teknik Fisika FTI – ITS (3) Sistem Kontrol Prediktif Nonlinier dengan model dan kontroler JST Fraksi mol MeAc (Output #1) 0.92
(x 10 00%)
0.9 0.88 0.86 0 84 0.84 0
200
400
600
800
1000
1200
1400
1600
Laju aliran produk 0.094
(mol/s)
0.092 0.09 0.088
putus-putus : set point kontinu : output proses
0.086 0 084 0.084 0
200
400
600
800 Cacah
1000
1200
1400
1600
11
Prediktif Kontrol di Teknik Fisika FTI – ITS (4)
Pengembangan multivariabel MPC serta implementasi online li Proses tangki bertingkat (quadruple tanks) Didanai Hibah Bersaing Ditjen Dikti Depdiknas
12
Kontrol Prediktif di Teknik Fisika FTI – ITS (6)
13
Kontrol Prediktif di Teknik Fisika FTI – ITS (7) Performansi Sistem Kontrol Prediktif Nonlinier
Level ta angki 1 (cm)
65
60
55
50
0
1000
2000
3000 Cacah
4000
5000
6000
0
1000
2000
3000 Cacah
4000
5000
6000
Level ttangki 2 (cm)
45
40
35
30
25
14
Kesimpulan Sistem kontrol prediktif memiliki beberapa keunggulan yang dapat dimanfaatkan sebagai salah satu alternatif sistem kontrol di industri migas. Penelitian yang intensif memang perlu terus dilakukan dengan melibatkan banyak kalangan baik akademisi, industri, maupun produksen instrumen
15
