Functional and behavioural modelling for dependability in automated production systems

Within the exploitation of the automated production systems (APS) framework, supervision and monitoring take the dependability into account. In order to reach these objectives, a method is being developed at the Laboratoire d'Automatique et d'Informatique Industrielle de Lille in a supervision and monitoring framework. The main functions are detection, diagnosis, modes management and faults recovery. This paper focuses on two kinds of modelling: functional modelling and behavioural modelling. On the one hand, these models are used to implement the diagnosis and the modes management functions. On the other hand, the complementarity of these models enables them to be used within the alarm processing system framework. Indeed, the design of intelligent alarm processing systems (IAPS) is a critical problem. One common problem is the number of alarms that operators have to manage. This paper focuses on the problem of the management of avalanches of alarms after the occurrence of failures in a complex process. To deal with this problem, different types of alarm are defined from the functional and behavioural modelling of the process to be surveyed. Filtering rules, based on the concepts of validation and inhibition, are also proposed.