Work Area: Uncertainty Management
Keywords fuzzy control, fuzzy sets, genetic algorithms, mass assignment, neural nets
Start Date: 1 October 92 / Duration: 36 months / Status: running
[ participants / contact ]
Abstract FALCON will discuss and develop scientific knowledge and available technologies in the area of fuzzy control in order to provide a basis for fuzzy controllers that are superior to existing ones. This includes learning and adaptation of fuzzy controllers, fuzzy inference, and design analysis and performance of fuzzy controllers.
The aim of this working group is to meet the Japanese and American challenge in the area of fuzzy control by pooling European expertise in the areas of artificial intelligence, control engineering and operations research. The theory of fuzzy control will be advanced to the point where better fuzzy controllers than those now available can be built efficiently. In particular, work will focus on three subareas of special importance: (i) learning and adaptation of fuzzy controllers, (ii) fuzzy inference, and (iii) design analysis and performance of fuzzy controllers.
The group will further its objectives through its own internal contact network which, depending upon the topic in question, will involve meetings of two or more partners. To improve the efficiency of communication, the nine partners have been structured into three clusters according to the three topics involved. One or two meetings per year are planned for each cluster. In addition, each year there will be one meeting of the coordinators of the three clusters. In 1993 and 1995 the group as a whole will also organise a meeting of the entire working group, this will be open to the public. The first and second European conferences on fuzzy technology will be organised. The group also intends to serve as one of the nodes of a Network of Excellence.
The working group contains participants involved in control engineering who started work in the field of fuzzy control in the early 1970s. They are mainly located in the control engineering laboratories of universities (Delft). Tools are already available for experiments in real-time environments and three well-known artificial intelligence groups are participating (Toulouse, Bristol and Brussels). Operations research is primarily represented by the chair of operations research in Aachen, where intensive empirical, axiomatic, and mathematical research on fuzzy set theory has been performed since 1972.
Since FALCON was started progress was made in several different areas. According to the three clusters of this working group, the research activities focused on the respective investigations. Improvements of genetic algorithms and their application to adapt fuzzy controllers will be presented at international conferences (see below). Here the integration of neural networks for learning tasks is one very promising field of research which will be considered in the future. A tool for data analysis is implemented which contains inference mechanisms for advanced fuzzy reasoning schemes. This tool supports the construction and application of fuzzy controllers. Prototypes where mobile robots are controlled have pointed out the need and potentials of fuzzy control for this kind of application. All these results will be presented at international conferences (see below).
This working group aims to enhance the potential for future technological breakthroughs by identifying novel techniques for the formal verification of the complex VLSI systems that will be required by the European IT industry. The progress of this working group will be regularly presented at international conferences and workshops and in publications.
The activities taken by the members of FALCON and by the working group coordinator include:
ELITE - D
European Laboratory for Intelligent Techniques Engineering
D - 52076 AACHEN
Université Libre de Bruxelles-IRIDIA - B
Universidad de Malaga - E
Université Paul Sabatier-IRIT - F
Siemens Automotive SA - F
CNRS-LAAS - F
Université de Savoie-LAMII - F
Technische Universiteit Delft - NL
University of Bristol - UK
Queen Mary and Westfield College - UK
Lund Institute of Technology - S
Prof. Dr. H.J. Zimmerman
FALCON - 6017, August 1994
please address enquiries to the ESPRIT Information Desk
html version of synopsis by Nick Cook