MOVES-Seminar, 26.10.2007, 13:30
Tomas Krilavičius
Vytautas Magnus University, Kaunas, Lithuania
Hybrid Systems in Process Algebraic Context
Abstract:
Computer controlled systems are almost omnipresent nowadays. We expect
them to function properly at any time we need them. The malfunctioning
of home electronics just irritates us, but glitches in a power plant
may threaten life, and faults in nuclear missile control facility may
bring the end to civilisation. It puts very high reliability
requirements on such systems. Hybrid systems combine continuous
real-time behaviour and discrete events. Research in hybrid systems
aims at providing means for reliable design and production of such
systems. Process algebra is a theoretical framework for the modelling
and analysis of the behaviour of concurrent discrete event systems
that has been developed within computer science in past quarter
century. It has generated a deeper understanding of the nature of
concepts such as observable behaviour in the presence of
nondeterminism, system composition by interconnection of concurrent
component systems, and notions of behavioural equivalence of such
systems. It has contributed fundamental concepts such as bisimulation,
and has been successfully used in a wide range of problems and
practical applications in concurrent systems. We believe that the
basic tenets of process algebra are highly compatible with the
behavioural approach to dynamical systems. In our contribution we
present Behavioural Hybrid Process Calculus an extension of classical
process algebra that is suitable for the modelling and analysis of
continuous and hybrid dynamical systems. It provides a natural
framework for the concurrent composition of such systems, and can deal
with nondeterministic behaviour that may arise from the occurrence of
internal switching events. Standard process algebraic techniques lead
to the characterisation of the observable behaviour of such systems as
equivalence classes under some suitably adapted notion of
bisimulation. Moreover, we discuss future developments of BHPC as
well as general directions in hybrid systems research.
Computer controlled systems are almost omnipresent nowadays. We expect
them to function properly at any time we need them. The malfunctioning
of home electronics just irritates us, but glitches in a power plant
may threaten life, and faults in nuclear missile control facility may
bring the end to civilisation. It puts very high reliability
requirements on such systems. Hybrid systems combine continuous
real-time behaviour and discrete events. Research in hybrid systems
aims at providing means for reliable design and production of such
systems. Process algebra is a theoretical framework for the modelling
and analysis of the behaviour of concurrent discrete event systems
that has been developed within computer science in past quarter
century. It has generated a deeper understanding of the nature of
concepts such as observable behaviour in the presence of
nondeterminism, system composition by interconnection of concurrent
component systems, and notions of behavioural equivalence of such
systems. It has contributed fundamental concepts such as bisimulation,
and has been successfully used in a wide range of problems and
practical applications in concurrent systems. We believe that the
basic tenets of process algebra are highly compatible with the
behavioural approach to dynamical systems. In our contribution we
present Behavioural Hybrid Process Calculus an extension of classical
process algebra that is suitable for the modelling and analysis of
continuous and hybrid dynamical systems. It provides a natural
framework for the concurrent composition of such systems, and can deal
with nondeterministic behaviour that may arise from the occurrence of
internal switching events. Standard process algebraic techniques lead
to the characterisation of the observable behaviour of such systems as
equivalence classes under some suitably adapted notion of
bisimulation. Moreover, we discuss future developments of BHPC as
well as general directions in hybrid systems research.
