The research activity in this area has concerned both the development of knowledge-based systems devoted to significant problems of industrial diagnosis (fault diagnosis in power transmission networks and preventive diagnosis of power transformers) and theoretical contributions concerning automated diagnosis of a class of discrete event systems, called active systems.
Power transmission networks are large distributed systems, endowed with
automated reaction capabilities with respects to several kinds of faults and
anomalous events.
When a fault occurs, the networks reacts in a possibly complex manner:
identifying the anomaly which caused a reaction is then a crucial task for
restoring normal network operation.
An original approach, based on the notion of history reconstruction, has been
developed for this task and implemented in a knowledge-based system prototype
called SHORT.
Relevant publications:
P. Baroni, U. Canzi, G. Guida, SHORT: a knowledge-based system for fault diagnosis in power transmission networks, Proc. ISAI 95, 8th Int.Symposium on Artificial Intelligence, Monterrey, MX, 1995, 199-208
P. Baroni, U. Canzi, G. Guida, Fault diagnosis through history reconstruction: an application to power transmission networks, Expert Systems with Applications, 12(1), 37-52, 1997
P.Baroni, G. Lamperti, P. Pogliano, G. Tornielli, M. Zanella, A diagnostic engine for power transmission networks, Proc. of ICI&C 97 International Conference on Informatics and Control, San Pietroburgo, Russia, Giugno 1997, 1076-1086
P.Baroni, G. Lamperti, P. Pogliano, G. Tornielli, M. Zanella, Automata-based reasoning for the diagnosis of short circuits in power transmission networks, Proc. AIENG 1997, 12th International Conference on the Applications of Artificial Intelligence in Engineering, Capri, I, Luglio 1997, 415-440
P.Baroni, G. Lamperti, P. Pogliano, G. Tornielli, M. Zanella, A multi-interpretation approach to fault diagnosis in power transmission networks, Proc. SAFEPROCESS'97, IFAC Symposium on Fault Detection, Supervision and Safety for Technical Processes, Hull, UK, Agosto 1997, 961-966
Power transformers are very expensive and critical components in electrical
power systems; their outage is very costly and very dangerous catastrophic
events, such as explosion or fire, may occur if their preventive diagnosis is
not carried out appropriately.
Preventive diagnosis of power transformers is a complex task typically carried
out by experts since it requires the integration of several types of
domain knowledge for the interpretation of data of various nature. Moreover both
data and knowledge are typically affected by incompleteness and
uncertainty.
A knowledge-based system for state assessment and preventive diagnosis of
power transformers, called ASTRA, has been developed and successfully tested.
Relevant publications:
P. Baroni, F. Cremonesi, G. Guida, S. Mussi, S. Yakov, ASTRA: a knowledge based system for state assessment, preventive diagnosis, and intervention planning of power transformers, Proc. ISAP 94 International Conference on Intelligent System Application to Power Systems, Montpellier, F, 1994, 463-470
P. Baroni, G. Guida, and S. Mussi, Distributed and uncertain reasoning in a knowledge-based system for preventive diagnosis of power transformers, Proc. 25th IEEE International Conference on Systems, Man, and Cybernetics, Vancouver, B.C., 1995, 1801-1806
P. Baroni, G. Guida, S. Mussi, A case study of technology transfer in the field of knowledge-based technology: the 9-ingredients approach, Proc. 25th IEEE International Conference on Systems, Man, and Cybernetics, Vancouver, B.C., 1995, 3210-3215
P. Baroni, G. Guida, G. Lamperti, M. Zanella, Model-based situation assessment of dynamic physical systems, Proc. AIENG 1997, 12th International Conference on the Applications of Artificial Intelligence in Engineering, Capri, I, Luglio 1997, 592-612
P. Baroni, G. Guida, S. Mussi, State assessment and preventive diagnosis of power transformers: a knowledge-based approach, International Journal of Engineering Intelligent Systems, 5(2), 1997, 91-105
P. Baroni, G. Guida, G. Lamperti, S. Mussi, Developing knowledge-based applications in engineering: quality, productivity, and technology transfer issues, In L.C. Jain (Ed.), Soft Computing Techniques in Knowledge-Based Intelligent Engineering Systems - Approaches and Applications, Springer Verlag, 1997, 115-140.
Building on the experience in the development of the SHORT prototype for diagnosis in power transmission networks, a significant class of discrete event systems, called active system, has been identified and formally characterized. The class of active systems covers a large family of practical systems to which previous automated diagnosis approaches were not satisfactorily applicable. Theoretical contributions and diagnostic algorithms have been provided for the class of active systems, with particular attention to the use of modular and incremental techniques in order to ensure their applicability to large scale systems.
Relevant publications:
P. Baroni, G. Lamperti, P. Pogliano, M. Zanella, Diagnosis of active systems, Proc. of ECAI '98, European Conf. on Artificial Intelligence, Brighton, UK, 1998, 274-278
P. Baroni, G. Lamperti, P. Pogliano, M. Zanella, Diagnosis of large active systems, Artificial Intelligence, 110 (1), 1999, 135-183
P. Baroni, G. Lamperti, P. Pogliano, M. Zanella, Diagnosis of a class of distributed discrete-event systems, IEEE Trans. on Systems, Man, and Cybernetics-Part A, 30 (6), 2000, 731-752