Fault predictability detection method for discrete event system and fault predictor

Abstract

The invention discloses a fault predictability detection method for a discrete event system and a fault predictor. A class 1 mark state and a class 2 mark state are used, and the system can be quicklyjudged to be unpredictable when the class 2 mark state exists in the system, so that the efficiency is higher. Model simplification is further proposed to finally obtain a simplified verifier automaton VG, so that the spatial complexity is reduced. Required observables are fewer, i.e. fewer sensing devices are needed when judging the system fault predictability and fault prediction, so that the stability of the fault predictability detection method and the fault prediction is higher. The verifier automaton VG finally obtained by the fault predictability detection method can also serve as a fault predictor to perform online prediction on the system fault.

Description

technical field

[0001] The invention relates to the technical field of fault prediction of discrete event systems, in particular to a fault predictability detection method and a fault predictor of discrete event systems. Background technique

[0002] With the widespread use of embedded computing devices and communication networks, the system integration of engineering systems is increasing and the functions are becoming more and more complex. Among them is a class of complex systems in which events start and stop at discrete moments with a certain degree of randomness. For example, the berthing of ships in the port, the start and stop of machine tools on the production line, the connection and disconnection of telephones, the progress and exit of a certain operation in the computer system, etc., such systems are called discrete event systems. Discrete event systems are a class of dynamic systems in which discrete events interact according to certain operating rules and lead...

Images