Online monitoring of a distributed building automation system to verify large sequences of bus messages by causal Petri net models

Distributed systems for building automation have already exhibited a high degree of functional complexity. The integration of AAL (Ambient Assisted Living), Smart Grid and energy saving features result in more functional relationships. To handle this complexity, the software architecture of distributed automation systems is increasingly based on hierarchy concepts. The depending software components in this hierarchy have to communicate via a bus system, which results in large sequences of messages for a communication process. The recognition of these causal sequences is necessary for a systematical identification of faults in the system. Due to the high amount of parallel processes, it is more difficult to assign non- deterministic events to a causal sequence of events. This contribution presents a case study of online monitoring of a distributed automation system for building automation to verify the causal relationships between bus messages. That assumes an establishment of a chronologic total order of permitted sequences, which are specified by causal Petri nets. A universal interface allows a direct link between this sequence model and the real automation system to capture specific communication sequences online. Based on the online recognition of message sequences during the operation phase, further analysis of fault or failure causes is supported by this kind of online monitoring system.