Vennligst spre budskapet om denne stillingen til gode søkere.
-Hilsen Sigurd
---------------------------------------------
Dr.ing. stipendiat
Modellbasert regulering/optimalisering
---------------------------------------------
I samarbeide med Statoil og flere europiske universiteter og bedrifter startes det opp et forskningsprosjekt
innen modellbasert regulering og optimalisering (INCOOP). Med utgangspunkt i eksistrende teknologi for
modellbasert prediktiv regulering og sanntids optimalisering, er siktemålet er å legge grunnlaget for neste
generasjons systemer innen dette området.
Ved oppstarten består programmet av minst fem doktorstipendiater finansiert av industribedriftene, hvorav
en stipendiat skal være ved NTNU. Det vil være et nært samarbeide mellom universitetsmiljøene og det tas
sikte på at et utenlandsopphold inngår som en del av studiet. Selv om doktorgradsarbeidene vil ta
utgangspunkt i konkrete industriprosesser, så er målsettingen å utvikle nye metoder av generell karakter.
Vi tilbyr et sterkt faglig miljø innenfor områdene regulering, dynamikk, optimalsiering og design av
prosessanlegg organisert gjennom styrkeområdet PROST.
Passende bakgrunn er sivilingeniør eller tilsvarende med studieretning kjemi, maskin eller regulering
(kybernetikk) og med interesse for prosessfag og matematiske fag.
Nærmere opplysinger fås ved henvendelse til professor Sigurd Skogestad, NTNU (tel. 7359 4154) eller
rådgiver Stig Strand, Statoil F&U (tel. 7358 4759). Se også http://www.kjemi.unit.no/~skoge/stipendiat.html .
Søknad sendes innen 15. august 1997 til:
Institutt for kjemiteknikk, NTNU, 7034 Trondheim
>
>
Litt om INCOOP:
---------------------------------------------------
Proposal INCOOP Consortium june 1997 - june 2001
---------------------------------------------------
Global Work Package, june 1997 - june 2001 including detailed Milestones
and Deliverables, june 1997 - june 1998
0. Partners involved
The work to be described below will be executed by the University
Partners in the INCOOP consortium
- Norwegian Univ. of Science and Technology (Skogestad)
- RWTH Aachen Univ. of Technology (Marquardt)
- Eindhoven Univ. of Technology (Backx, van de Bosch)
- Delft Univ. of Technology (Bosgra, Grievink)
by five doctural students, under the guidance of the university groups. Support
will be provided by the Supplier Partners in the consortium
- Aspentec, NL/USA
- MDC Co., UK
in a generic sense (availability of tools, technical support), and
in certain detailed parts of the work package as indicated. Support will
also be provided by the Process Industry Partners
- Shell
- Dow
- Bayer
- Statoil
regarding making available plant data and experimental
facilities. As agreed, the work package will be directed towards a
detailed description of the first 5 x 18 months including formulation of
deliverables and milestones. A global plan will be given for the
subsequent 5 x 30 months.
1. Principal goals
Two principal goals of the INCOOP consortium during its first phase
1997-2001 are:
a. To realize a first step in an improved technology for the integration
of advanced process control and real-time process optimization as
compared to the industrially applied state-of-the-art
b. To show the benefits of this first step of technology improvement by
implementing prototype software realizing the improved technology on two
benchmarking case studies
note: State-of-the-art refers to the technology which is currently
provided by suppliers or by operating company technology groups and used
in industrial applications. Improved technology on the other hand refers
to:
1. Known results in literature not yet applied in industrial routine
2. New methods and technology to be developed in INCOOP
These technical objectives have to result in:
* Further reduction of standard deviations of critical process and
product variables with 30 % compared with currently applied MPC
* Reduction of average transition time between operating points with at
least a factor 2 by improved transition trajectory optimization
* Closer operation to true economic optima of plants resulting in
realization of at least better than 70 % of the potential achievable
economic benefits by the fully integrated use of more accurate models,
better model consistencies, better performing model based control
systems and improved data processing techniques.
2. General Approach
The development strategy will pursue a smooth migration path from
present to future technology. The approach will start on specific plant
applications, and proceed in the direction of generalization. Focus will
be on four main issues:
(1) operational integration of optimization and control layers
(2) take dynamics of operation into full consideration
(3) investigate how the work process can be shortened
(4) integrate knowledge over application areas
The main steps to be performed in the approach are:
Part A.
Formulation of two bench-marking case studies, one aiming at the
refinery industry, one aiming at chemical process industry. The two
selected processes for the case study are:
(1) fluidized catalytic cracker (FCC)
(2) HCl electrolysis process, the so-called Bayer-Hoechst-Uhde process
(BHU)
Part B.
Formulating rigorous dynamic simulation models for each of the
processes, allowing each simulation to be considered as the 'process'
which will be available in real time to study and implement operational
control and real-time optimization technology improvements.
Part C.
Development, software implementation and (simulated) experimentation
with state-of-the-art solutions of MPC and RTO for both processes.
Part D.
Development of improved technology for integrated optimization and
control. Attention to be aimed towards multi-time-scale data
interpretation, multi-time-scale, economic optimization-directed
control, and model-based plant/model/data interpretation for process
supervision. Demonstration of technology improvement and benefits by
implementation and experimentation on both 'processes'. Bringing the
results of both case studies together in a generalization step, leading
to the establishment of obtained improvements and results. In addition,
a thorough evaluation of all results must lead to the diagnosis of
possible improvements and to the formulation of necessary further
developments.
***************************************************************************
Sigurd Skogestad, Professor Fax: +47-7359-4080
Chemical Engineering Phone: +47-7359-4154
Norwegian Univ. of Sci. and Techn. (NTNU) Home: +47-7393-6855
N-7034 Trondheim email: skoge@kjemi.unit.no
Norway http://www.kjemi.unit.no/~skoge
***************************************************************************