Applying a supervisory control system based on a petri net model for fault tolerant handling of AUVs in water ingress accidents

The issue of air aspiration is a major challenge for AUVs’ propulsion systems working based on combustion engines. Some AUVs have snorkels to take fresh air from sea surface, so keeping heights of these AUVs is important. A reliable and safe bilge sub-system survives the AUV in emergency cases. In this paper, a three layers supervisory control system has been employed to handle a bilge subsystem dealing with faults (e.g., pump failures) and allows to keep the inlet of the AUV’s snorkel upon sea water. The highest layer of the supervisory control system as a mission layer is modeled by a decision tree and the decision tree is built on petri net models. In addition, our results represented the bilge might not be sufficient to deal with the effects of water ingress. Using an accumulator in the supervisory control system to keep the height can improve fault tolerant of AUV after water ingress.