Research On Maintenance Strategies For Civil Aviation Fuel Systems Based On Multilingual Contextual Systems

The fuel system of civil aircraft plays an important role in the flight process,mainly for storing,distributing,and supplying fuel.In recent years,frequent civil aviation accidents have prompted researchers to conduct numerous studies around fault diagnosis and maintenance,gradually forming the theory of preventive maintenance,from past maintenance after the event,regular maintenance to the current state-based maintenance.Reliability-centered maintenance(RCM)is a fault maintenance method based on equipment failures,aiming to improve equipment safety and reliability while reducing maintenance costs.This method has been widely used in practice,but the current reliability system has subjectivity issues,and most fault diagnosis and maintenance systems may produce misjudgments and misdiagnoses due to the aging of faults and the operating conditions and environment of equipment,indicating the need for further improvement in system maintenance.This paper is proposed a multi-context reliability fault maintenance system,which establishes a fault knowledge base and uses Answer Set Programming(ASP)to summarize reliability into functional attributes,fault attributes,and maintenance attributes.Compared with traditional maintenance systems implemented using databases,maintenance systems implemented using Answer Set Programming have the advantages of interpretability,quick insertion of knowledge,and loose conflict resolution.At the same time,the system can express the maintenance process knowledge more comprehensively and in detail.The prototype system implemented using Answer Set Programming demonstrates that it can intuitively and efficiently represent maintenance system knowledge and logical libraries and has some practical value.To address the subjectivity of reliability systems,this paper is proposed a decision tree method based on causal measurement,which adds causal classification criteria and generates causal gain ratio to provide a basis for node splitting based on information gain ratio,thereby classifying the three reliability attributes in the fault knowledge base and achieving decision-making according to fault types.This paper is confirmed the parameter selection principle of the method through 10-fold cross-evaluation experiments on multiple public datasets and verifies its effectiveness,which can provide assistance for classification decision-making.This paper is established a multi-context system to build a fault maintenance system based on civil aviation fuel systems.Specifically,the fault knowledge base,fault model classification,fault risk and consequence assessment,and logical decision-making are set as corresponding contexts,and bridge rules are used to facilitate communication between the different contexts,making full use of the advantages of each context to optimize the maintenance strategy.Finally,the adaptability and simplicity of the system were verified in the experiment of fuel supply failures in civil aviation fuel systems,and suitable maintenance strategies were successfully provided for faults.This research is provided a new solution for the field of civil aviation fuel system fault maintenance and has important theoretical and practical value.