Introduction to the department

Address

Phone: +32 (9) 264.59.32 (secretary)
Fax: +32 (9) 264.62.20
e-mail: secretaries@biomath.UGent.be

How to get here

Research

Research Goals

• Disciplines include water quality, bioprocess technology, food technology, econometrics, software process management, environmental technology, medical science, process control, production planning, traffic control, agricultural technology, etc.
• The basis of these studies is a rigorous modelling and simulation methodology which prescribes how to deploy numerical analysis, computer science, statistics, biometrics, control theory, experimental design, systems analysis, artificial intelligence, etc. in the analysis, design, optimisation and control of complex systems. Currently, the most important aspect of complexity is not only the number of system components, but also the diversity and heterogeneity of these components (e.g., hardware and software building blocks, continuous and discrete components, etc.).

Research Approach

1. Deductive modelling and simulation (a priori knowledge to model):

   A modelling methodology and supporting environment are developed in which system models are constructed from general principles and a priori knowledge. Subsequently, a simulation environment is generated to investigate the complex system under study (virtual experimentation). A distributed (WWW) implementation of this environment is achieved using intelligent agent technology.

2. Inductive modelling (data to model):

   Inductive modelling starts from data, from which models for the process are built. The application domain consists mainly of ill-defined systems, with the assumption that little or no deterministic system knowledge is available. The main objective is to develop conglomerate information extracting techniques (inductive (traditional: statistical; new: entropy-based) + deductive). Optimal Experimental Design techniques are deployed for efficient generation and collection of experimental data.

3. Applications of modelling and simulation (iterating with the model):

   The optimisation of bioprocesses is pursued using an optimal mix of inductive and deductive techniques. This activity acts as a validator of the above theoretical results and as a generator of new ideas looking for an efficient solution. The systems under study are related to environmental biotechnology (integrated wastewater management, sustainable development, etc.), fermentation technology (production of antibiotics, etc.), and food and feed biotechnology.

Perspectives of BIOMATH Research and Development

1. Deductive modelling and simulation:

   Further development and validation of the methodology. Realisation and use of software environments for computer-aided modelling and simulation. Introduction of symbolic manipulation to increase the efficiency of the simulation environment. Integration of the environment within a real-time process control context.

2. Inductive modelling:

   Further refining of model selection using data-driven feature extraction techniques, and implementing this in a real-time process environment. The integration of statistics and system theory. Use of genetic algorithms and AI techniques to improve pattern recognition capabilities. Development of on-line optimal experimental design techniques for model selection and parameter estimation of non-linear models. Design of methods for the building and identfication of spatio-temporal models, including aspects of extreme value statistics.

3. Applications of modelling and simulation:

   Dealing with modelling and optimisation tasks of systems characterised by a higher level of integration, and larger spatial and temporal scale, (e.g. integrated urban wastewater management). Development of the Hardware-In-the-Loop concept for real-time model selection and parameter estimation using optimally designed experiments. Uncertainty analysis to assess the reliability of simulation results and to support decision making.

A selection of recent BIOMATH publications

The most recent publications of our department can be found here.

Keywords

Mathematical Modelling, Bioprocess Technology, Simulation, Optimisation, Process Control, Statistics, Systems Analysis, Biotechnology, Environment, Identification, Sensors, Sustainable Development, Food Technology, Feed Technology, Agriculture, Experimental Design, Computer Science, Biometrics, Extreme Value Statistics, Uncertainty Analysis

University-wide description

Click here for a university-wide description of the department.