PROCESS
MODELLING, CONTROL
AND OPTIMISATION GROUP
The Process Modelling, Control and Optimisation
research group in the
Staff members:
|
|
A/Prof. Tuan
Pham
Optimisation, Numerical Methods |
|
Dr. Jie Bao
Robust process control, Decentralised control,
Control of membrane processes |
|
|
A/Prof.
Soji Adesina
Process modelling |
|
Dr. Dianne
Wiley
Modelling, optimisation and control of membrane
processes |
|
|
Dr. David
Clements
Modelling, optimisation and control of nonlinear
processes, Control and optimisation of membrane processes |
|
|
Research Area/Projects:
Process Modelling
Activities are carried out on
the modelling of various processes from both the fundamental and empirical
aspects. A common focus of the modelling group is the application of
optimisation methods, to maximise profits and minimise environmental impacts to
the various process models. Some specific areas the research group
are involved with are:
§
Modelling
and Optimisation of Chemical Reactors – dynamic and steady state
modelling/optimisation of fixed bed, fluidised bed, trickle bed,
and slurry reactors.
§
Modelling
and Optimisation of Combined Reaction and Separation Systems – modelling
and optimisation of catalytic membrane reactors (CMR). Specific research is being carried out
on CMR’s for the production of hydrogen and sulphur from hydrogen
sulphide gas.
§
Development
of Models for the Food and Bioprocessing Industries –Development of
refrigeration software for use in the food industries to lower costs and
improve product quality. Modelling of biofiltration processes for removal of
contaminants by microorganisms.
§
Probabilistic
Modelling of Complex Systems – development of modelling methods for
systems with inherent uncertainties with application to the process industries
as well as chemical, environmental, and biological systems in various scales.
§
Dynamic
Modelling of Membrane Processes - developing dynamic process models that can be
used for improved control of pressure driven membrane systems. The major
dynamic phenomenon associated with membrane systems, flux decline and fouling
are being studied and mathematical models are being developed, using a 'hybrid
modelling' method, which combines both physical and empirical modelling
approaches.
Process Control
Process control research is
being conducted for both theoretical studies and applications.
§
Passivity-based
robust process control
– This is a new approach to robust process control using
the Passivity Theorem. Instead of using the norm bound of the uncertainty, as
in mainstreamed robust control (such as H¥ control), uncertainty
is characterised in terms of passivity. Several new robust stability conditions
have been derived and robust control synthesis methods have been developed
based on the proposed stability conditions. Current work focuses on developing
practical methods to estimate passivity indices of nonlinear unstructured uncertainties
and extension to nonlinear process control based on the Passivity Theorem.
§
Decentralised control studies
– This work focuses on interaction analysis for Multi-Input Multi-Output
(MIMO) processes, the stability conditions for decentralised control and multi-loop controller
design for MIMO
processes. A new passivity-based method to analyse Decentralised Integral Controllability
has been derived. This result is being extended to stability conditions of
general decentralised control systems. Performance limit implied by the
stability condition is being studied. Recent work includes studies on
decentralised failure-tolerant control, which concerns decentralised unconditional stability, and the synthesis of controllers
that achieves closed-loop stability while one or more control loops are
switched off due to failure of sensors and/or actuators.
§
Advanced Control of Membrane
Process - This
work aims to develop a dynamic process model and advanced control schemes for pressure driven membrane systems. Current methodologies, while functional,
are conservative, narrow and slow and do not take advantage of process
improvements achievable with tight active control. The expected outcomes
include a validated model, control strategies that maximize productivity and
minimize fouling during normal operation as well as during start-up and
shut-down.
Optimisation
§
One of the
most common uses of process models is to use them to determine the optimal
manner to design and operate the process, to maximise profits and minimise
environmental impact. Activities
related to the development and application of optimisation methods for process
design and operation are being conducted.
Some specific areas include:
§
Optimal
Process Design and Operation of Membrane Plants – membranes are being
increasingly used for separation in many industries. Research is being
conducted to determine the best configuration for the membrane module network,
as well as to determine the optimal operation strategy (flowrates, cleaning
cycles, membrane replacement).
Methods being developed include Mix Integer Non-Linear Programming, and
a novel application of fuzzy logic for preliminary designs.
§
Hybrid
Optimisation Methods – development of hybrid optimisation methods, which
combine the advantages of conventional hill climbing methods with the
flexibility of new stochastic method such as parallel simulated annealing and
competitive genetic algorithms. Applications to process and plant optimisation.
Numerical
Methods
§
Research
is also being carried out to develop appropriate numerical solution techniques
for moving boundary front problems such as systems undergoing chemical reaction
or phase change (freezing, thawing).
Research into special numerical methods to handle these types of
situations is being conducted.
These include: adaptive finite difference, orthogonal collocation on
finite elements, quasi-linearisation and invariant imbedding methods. The novel
combination of these techniques for computationally-efficient algorithms is
also being pursued.
