Description and diagnosis of malfunctions in autonomous uninhabited underwater vehicles based on ontologies

The paper proposes new ontologies for describing and diagnosing malfunctions of subsystems of autonomous underwater vehicles (AUVs) as part of the development of a previously created approach to the intelligent diagnosis of these vehicles using knowledge bases. To implement this approach, it is proposed to use two knowledge bases, one of which describes malfunctions, signs for their identification and actions necessary to adapt to the negative consequences caused by the appearance of these malfunctions. Another database contains information about the main subsystems of the AUVs, devices and replaceable modules, as well as the states in which they may be. Knowledge bases are used in the operation of intelligent control and emergency systems (ICES) on the AUV boards, which allow for comprehensive functional diagnostics of all subsystems of the device, including payload modules, as well as adaptation to the consequences of emerging malfunctions, taking into account the actual cause-and-effect relationships between them. The proposed ontologies for the indicated knowledge bases are compact, which makes it possible to use ICES on the AUV boards. To maintain up-to-date information about possible malfunctions, it is proposed to carry out continuous remote maintenance of ICES using the software Internet platform for the development and support of intelligent cloud services IACPaaS. To test the operability of ICES, its prototype was developed, which implements the functions of processing knowledge bases and developing solutions for adapting to faults. The simulation results of the prototype operation showed that ICES fulfills the assigned tasks and can be implemented on the on-board computers of existing AUVs using known programming technologies.