Det blir tirsdag 6/12 kl.13.15 paa NTH.
Tittel paa avhandling er: Topology-based diagnosis for chemicalprocess plants.
------------------------
Proeveforelesning kl.10.15: Modelling of hybrid control system
Hybrid control systems include sequencing, logic control and
continuous control. Examples of this is the transition between control
structures based on process operating conditions, and sequence-driven
control of batch-reactors.
1) Give a comparative study of model representations for hybrid systems.
2) Assess the possibilities of using representation in 1) for designing
control systems.
3) Give examples of the use of these representations in
industry.
------------------
Vi arrangerer en kjemi/systemteknisk gjesteforelesning onsdag 7/12 kl.10.15
paa B-343.
Intelligent Systems In The Process Industries
Prof. James F. Davis
Department of Chemical Engineering and Laboratory for AI Research
Ohio State University
Columbus, OH
Many problems in process operations are ill-structured and share
certain generic
characteristics that make traditional approaches to automation difficult
if not
impossible. Intelligent System technologies offer a set of powerful
concepts and
techniques that are indispensable in addressing the challenges
faced in process
systems engineering. In particular, intelligent systems are poised to play a
central role in the emerging paradigm of computer integrated manufacturing
(CIM)
applied to chemical process industries. Partial CIM implementations are quite
prevalent and have resulted in the anticipated benefits of improved
efficiency,
better product quality and lower operating costs.
The value-added benefits of intelligent systems technologies can be
described in
terms of qualitative modeling, cognition, deliberation to achieve
generalization,
adaptation and autonomy. These incorporate major aspects of "intelligent"
functionality observed in the different techniques and applications.
As each is
"built" into an intelligent system, the improved performance brought by the
system to the particular application increases. Knowledge-based systems and
neural networks offer important techniques for producing these intelligent
functionalities. We explore integrated intelligent systems through several
industrial implementations that address monitoring, detection, data
interpretation, diagnosis and HAZOP.
-------------------
Kan du 'spre' dette hos dere.
Velkommen (jeg tror det blir interessant)!
Bjarne