Aviation equipment fault analysis method and system based on ontology semantic retrieval

Abstract

The invention provides an aeronautical equipment fault analysis method and system based on ontology semantic retrieval, and relates to the technical field of aeronautical equipment maintenance safeguard.The method comprises the steps that S1, an aeronautical equipment fault ontology model is established, retrieval conditions are set, and fault ontology characteristic parameters are determined; s2, performing semantic data matching of the aviation equipment fault ontology concept feature parameters; s3, aviation equipment fault ontology data characteristic parameter matching is carried out; s4, calculating the weighted semantic similarity of the aeronautical equipment fault ontology concept feature parameters; and S5, judging the fault type of the aviation equipment, and performing auxiliary processing on the aviation equipment fault. According to the method, actual fault data retrieval is converted into matching of concepts in the fault ontology library and semantics of the concepts, and retrieval is performed by using the standard concepts, so that the recall ratio and precision ratio of fault data retrieval are effectively improved; and a concept matching similarity method is provided by combining depth and density factors of the ontology model, aeronautical equipment fault ontology semantic retrieval is realized, and accurate fault auxiliary diagnosis is provided.

Description

TECHNICAL FIELD

[0001] The application relates to the technical field of aviation equipment fault evaluation, and particularly relates to an aviation equipment fault analysis method based on ontology semantic retrieval and a system thereof. BACKGROUND

[0002] Aviation equipment belongs to a large, complex, high-technology and high-quality product system. Due to the large system structure, complex functions and the crisscross of sub-systems and sub-sub-systems, the probability of faults is relatively large, and local faults can sometimes cause major accidents and huge losses or even casualties. Therefore, it is necessary to reduce the probability of faults by improving reliability and preventing and eliminating faults by maintenance. Faults are inherent characteristics of products, and products will have faults in the use process, which cannot complete the task or function as required, and even affect safety. As a complex equipment system, aviation equipment faults have the characteristics of complexity, hierarchy, correlation, time delay and uncertainty. The most direct method to solve the fault problem is to reuse similar fault data. The main way of fault data application is knowledge retrieval. Traditional information retrieval technology generally applies simple word matching rules, and only the search term can be retrieved in the search object. The representation ability of related information is limited, and the semantic information cannot be described and reflected. However, the input search term and the knowledge background, search ability and search experience are related, and it can be only a number of synonyms, near synonyms or related terms of a certain concept. Therefore, similar concepts related to the user's search request cannot be retrieved due to different words, the retrieval efficiency is low, and the retrieval result lacks accuracy and comprehensiveness.

[0003] Semantics is a concept that can be recognized by a computer, which is the relationship between symbols and expressions constructed on a certain grammar and the objects they describe. Semantic retrieval is a retrieval mechanism that provides information that can be read and understood by a machine through semantic information description and carries out retrieval and reasoning according to certain logical rules. Semantic retrieval applies a set of standard concepts to map the search term to its synonyms, near synonyms and related semantic related terms, and uses a set of standard concepts for retrieval, which can effectively improve the recall rate and precision rate of information retrieval. The key to realizing semantic retrieval lies in the expression of knowledge and knowledge-based reasoning. Knowledge retrieval is the reverse process of knowledge organization, and the realization of knowledge retrieval based on concept semantics must rely on the support of the knowledge organization system.

[0004] Ontology is the explicit formal specification of a shared conceptualization, which contains modeling primitives such as concepts, relations, functions, axioms, instances, etc. as a way of organizing knowledge, can explicitly describe the setting of domain concepts, reflect the semantic information of concepts through the relations between concepts, give the simple terms explicit background knowledge, and thus make the implicit relations clear and guarantee the consistency of semantics. The concept of Ontology originally comes from the field of philosophy, which is used to study the nature of the objective world. In the field of science and technology, the recognized setting is that Ontology is the explicit formal specification of a shared conceptualization. This setting contains four meanings: share refers to the idea of capturing common sense knowledge, reflects the recognized concept set in the relevant field, and is aimed at the group rather than the individual; conceptualization refers to the model obtained by abstracting the relevant concepts of some phenomena in the objective world, which represents the meaning independent of the specific environmental state; explicit refers to clearly setting the types of all concepts and the relation constraints between concepts; formal refers to the fact that Ontology should be computer understandable, i.e. computer processable. The goal of Ontology is to capture the fault data of the relevant field, provide a common understanding of the fault data of the field, determine the commonly recognized vocabulary in the field, and give the explicit setting of the vocabulary and the mutual relations between the vocabulary from different levels of formalization. The language used by Ontology can be divided into non-formal, semi-formal and formal Ontology language according to the formalization degree of representation and description. The higher the formalization degree of Ontology, the more conducive to automatic processing by computer. Ontology can explicitly describe the setting of domain concepts, reflect the semantic information of concepts through the relations between concepts, and give the simple terms explicit background fault data, thus making the implicit relations clear and guaranteeing the consistency of semantics. Since Ontology can make the communication between people and computers or between computers based on the consensus of the field to be communicated, it is suitable for the representation of aviation equipment fault data. The Ontology-based aviation equipment fault data representation method represents the real world itself through modeling, which is independent of the task. The Ontology-based topological representation of various types of aviation equipment fault data establishes the explicit association between them, thus realizing their unification, and also provides the explicit semantics of aviation equipment fault data.

[0005] Fault data is all the valuable experience and information related to product faults, involving fault products, fault modes, fault causes, fault effects, fault handling, fault cases, fault diagnosis, etc. The adoption of Ontology technology for aviation equipment fault data organization can provide domain concepts and their relations that can be understood by computers, and also provides support for semantic retrieval. SUMMARY

[0006] In order to solve the above-mentioned prior art, the purpose of the present application is to provide an aviation equipment fault analysis method and system based on ontology semantic retrieval, by normalizing the retrieval condition, converting it into a concept in the ontology rule constraint, converting the actual fault data retrieval into the matching of the concept and its semantics in the fault ontology library, searching and reasoning on the semantic layer of the fault data, effectively improving the recall rate and precision of the fault data retrieval by using a set of standardized concepts for retrieval; a semantic retrieval model based on ontology is established from the aspects of semantic expansion and semantic matching, a concept matching similarity method based on semantic distance is proposed in combination with the depth and density factors of the ontology model, and a semantic retrieval method and its implementation process based thereon, the fault data semantic retrieval application support for users is provided through the function of the fault analysis system, the comprehensive and accurate retrieval of the fault data is realized, and strong support is provided for solving the fault problem.

[0007] Specifically, in one aspect, the present application provides an aviation equipment fault analysis method based on ontology semantic retrieval, comprising the following steps: S1: establishing an aviation equipment fault ontology model, setting a retrieval condition, and determining fault ontology characteristic parameters; if the retrieval item of the aviation equipment fault data corresponds to the concept characteristic parameters of the fault ontology, step S2 is executed; if the retrieval item of the aviation equipment fault data corresponds to the data characteristic parameters of the fault ontology, step S3 is executed; S2: expanding the semantic data of the aviation equipment fault ontology concept characteristic parameters, calculating the depth factor and the density factor of the aviation equipment fault ontology concept characteristic parameters, and determining the semantic distance between the fault ontology concept characteristic parameters; calculating the semantic similarity between the aviation equipment fault ontology concept characteristic parameters, and performing semantic matching of the aviation equipment fault ontology concept characteristic parameters; S3: performing aviation equipment fault ontology data characteristic parameter matching, specifically including: corresponding matching of data characteristic parameters, fuzzy matching of data characteristic parameters, and text data characteristic parameter matching; S4: performing steps S2 and S3 on the aviation equipment fault ontology concept characteristic parameters and data characteristic parameters respectively, adding the obtained characteristic parameter similarities, performing normalization processing on the weight of all retrieval characteristic parameters, calculating the weighted semantic similarity of the aviation equipment fault ontology concept characteristic parameters, and specifically: ; Wherein, is the weighted semantic similarity between the instance and the instance in the aviation equipment fault ontology; is the weight of the i-th characteristic parameter; is the weight of the i-th characteristic parameter; is the weight of the i-th characteristic parameter; a semantic similarity function of the feature parameter values; for the first aviation equipment fault ontology instance; for the second aviation equipment fault ontology instance; for retrieving the feature parameter number; for retrieving the total number of feature parameters; S5: According to the semantic retrieval result of the aviation equipment fault in step S4, the aviation equipment fault type is determined.

[0008] Preferably, step S2 is specifically: S21: Expanding the semantic data of the aviation equipment fault ontology concept feature parameter, specifically including: data synonym expansion, feature parameter expansion, hierarchical expansion, axiom expansion and rule expansion; S22: Calculating the depth factor and density factor of the aviation equipment fault ontology concept feature parameter, and determining the semantic distance between the fault ontology concept feature parameters; S23: Calculate the semantic similarity between aviation equipment fault ontology concept feature parameters, and perform semantic matching of aviation equipment fault ontology concept feature parameters.

[0009] Preferably, step S22 is specifically: S221: Determine the depth of the aviation equipment fault ontology concept feature parameter is ; Set the depth factor of the concept hierarchical depth factor , calculate the depth factor of the aviation equipment fault ontology concept feature parameter; S222: Determine the density of the aviation equipment fault ontology concept feature parameter is ; Set the density factor of the concept density factor , calculate the density factor of the aviation equipment fault ontology concept feature parameter; S223: In the fault concept tree of the aviation equipment fault ontology concept feature parameter, calculate the semantic distance from the concept feature parameter node to the concept feature parameter node, get the comprehensive semantic distance between two concept feature parameters in the aviation equipment fault ontology .

[0010] Preferably, the depth factor of the aviation equipment fault ontology concept feature parameter in step S221 specifically includes: the upper semantic relationship depth factor and the lower semantic relationship depth factor; The upper semantic relationship depth factor is determined by the root node of the aviation equipment fault concept tree, and the lower semantic relationship depth factor is determined by the bottom leaf node of the aviation equipment fault concept tree, specifically: ; ; wherein, is a superordinate semantic relation depth factor of the concept feature parameter ; is an aeronautical equipment fault ontology concept feature parameter; is a superordinate semantic relation depth factor of the concept feature parameter ; is a subordinate semantic relation depth factor of the concept feature parameter ; is a depth of the aeronautical equipment fault concept tree ; is the aeronautical equipment fault concept tree.

[0011] Preferably, the density factor of the aeronautical equipment fault ontology concept feature parameter in step S222 comprises a superordinate semantic relation density factor and a subordinate semantic relation density factor; the superordinate semantic relation density factor is determined by a root node of the aeronautical equipment fault concept tree, and the subordinate semantic relation density factor is determined by a bottom layer leaf node of the aeronautical equipment fault concept tree, and specifically: ; ; wherein, is a superordinate semantic relation density factor of the concept feature parameter C; is a superordinate node of the concept feature parameter C; is an out-degree of the superordinate node of the concept feature parameter C; is a subordinate semantic relation density factor of the concept feature parameter ; is an out-degree of the concept feature parameter .

[0012] Preferably, step S223 multiplies the reference value of the superordinate and subordinate relation semantic distance by a comprehensive adjustment coefficient combined with the depth factor and the density factor, to obtain an actual value of the semantic distance of the superordinate and subordinate semantic relation ; the comprehensive adjustment coefficient combined with the depth factor and the density factor is a depth factor of the superordinate and subordinate semantic relation of the concept feature parameter plus a density factor multiplied by an adjustment coefficient , and specifically: ; wherein, is the actual value of the semantic distance of the superordinate and subordinate semantic relation; is the reference value of the superordinate and subordinate relation semantic distance; is a depth factor of the superordinate and subordinate semantic relation of the concept feature parameter ; Conceptual feature parameters Density factor of hierarchical semantic relationship; This is the adjustment coefficient.

[0013] Preferably, step S23 specifically includes: Calculate the comprehensive semantic distance between concept feature parameters in the fault ontology. The comprehensive semantic distance is the sum of the semantic distances of all directed edges of the shortest path between concept feature parameters. Calculate the semantic similarity between concept feature parameters in the fault ontology. The specific steps are as follows: ; ; ; in, Conceptual feature parameters in the fault ontology arrive The comprehensive semantic distance; Conceptual feature parameters arrive semantic distance; Conceptual feature parameters arrive The shortest path is a directed edge; Number the directed edges of the shortest path; This represents the total number of directed edges in the shortest path. These are the first concept feature parameters; These are the characteristic parameters of the second concept; Conceptual feature parameters To concept feature parameters The maximum semantic distance; This is the root node of the aviation equipment fault concept tree; Conceptual feature parameters in the fault ontology to the root node semantic distance; The root node in the faulty entity To concept feature parameters semantic distance; Conceptual feature parameters in the fault ontology With concept feature parameters Semantic similarity between them; As a regulating factor; Conceptual feature parameters in the fault ontology To concept feature parameters Semantic distance.

[0014] Preferably, step S3 specifically includes: S31: The corresponding matching of the feature parameters of the aviation equipment fault ontology data is data corresponding matching of the data feature parameters of the Boolean type, the enumeration type, the numerical type and the date time type, so as to obtain a similarity function of the data feature parameter corresponding matching; S32: The fuzzy matching of the feature parameters of the aviation equipment fault ontology data needs to divide the limited interval of the set domain, and through a similarity function of the data feature parameter fuzzy matching, the similarity of the two data feature parameters is obtained. S33: The text type data feature parameter is quantified and similarity calculation is performed in a way of vocabulary matching; the real word matching of the text data feature parameter is performed; and the similarity of the text data feature parameter is obtained.

[0015] Preferably, the step S5 is specifically: setting the weight of each concept feature parameter and the semantic similarity threshold of the retrieval result, for screening the retrieval result to improve the semantic retrieval accuracy; and eliminating the result instance with the semantic similarity less than the similarity threshold in the step S4, so as to obtain a final semantic retrieval result instance set, sort and output the aviation equipment fault ontology semantic retrieval result according to the similarity from large to small, and judge the aviation equipment fault type according to the aviation equipment fault semantic retrieval result.

[0016] On the other hand, the application provides a fault analysis system of an aviation equipment fault analysis method based on ontology semantic retrieval, which comprises: a fault ontology module, an attribute weight and similarity threshold setting module, a fault data semantic retrieval module and an aviation equipment fault judgment output module. The fault ontology module is used for operating and viewing the aviation equipment fault ontology data, and comprises the structure construction of the concept, the relationship, the attribute and the instance of the aviation equipment fault ontology. The attribute weight and similarity threshold setting module is used for setting the weight of the aviation equipment fault data retrieval item corresponding to the aviation equipment fault ontology concept attribute and the semantic similarity threshold of the retrieval result, so as to perform the semantic similarity calculation of the aviation equipment fault data and the retrieval condition and the screening of the retrieval result. The fault data semantic retrieval module is an interactive module of the fault analysis system and the semantic retrieval, realizes the concept feature parameter matching similarity method based on the semantic distance through the combination of the aviation equipment fault ontology model and the semantic retrieval, inputs the retrieval condition, calls the fault ontology module and the attribute weight and similarity threshold setting module to perform the semantic similarity calculation, and outputs the aviation equipment fault ontology semantic retrieval result. The aviation equipment fault judgment output module compares the output aviation equipment fault ontology semantic retrieval result of the fault data semantic retrieval module with the aviation equipment fault condition, and judges the fault type of the aviation equipment.

[0017] Compared with the prior art, the application has the following beneficial effects: (1) This invention can solve the problem of comprehensive and accurate retrieval of aviation equipment fault data. In view of the problems of low retrieval efficiency and lack of accuracy and comprehensiveness of retrieval results in traditional information retrieval technology, this invention introduces a set of standardized concepts of fault data ontology to standardize the retrieval conditions and transform them into concepts in ontology rule constraints. This transforms the actual fault data retrieval into the matching of concepts and their semantics in the fault ontology library. It is a search and reasoning process at the semantic layer of fault data. By using a set of standardized concepts for retrieval, the recall and precision of fault data retrieval can be effectively improved.

[0018] (2) This invention establishes an ontology-based semantic retrieval model from the aspects of semantic expansion and semantic matching. It proposes a concept matching similarity method based on semantic distance by combining the depth and density factors of the ontology model, as well as a semantic retrieval method and its implementation process based on this. Through the fault analysis system function, it provides user-oriented fault data semantic retrieval application support, realizes comprehensive and accurate retrieval of fault data, and provides strong support for solving fault problems. Attached Figure Description

[0019] Figure 1 This is a flowchart illustrating the aviation equipment fault analysis method based on ontology semantic retrieval of the present invention. Figure 2 This is a structural diagram of the aircraft equipment fault model of the present invention; Figure 3 This is a schematic diagram of the basic semantic relationships of the aviation equipment fault entity in this invention; Figure 4 This is a schematic diagram illustrating the domain-specific semantic relationships of the aviation equipment fault ontology in this invention; Figure 5 This is a schematic diagram illustrating the actual semantic distance of the hierarchical relationship involving depth and density factors in this invention. Detailed Implementation

[0020] Hereinafter, embodiments of the present invention will be described with reference to the accompanying drawings.

[0021] This invention proposes a method for analyzing aviation equipment faults based on ontology semantic retrieval, such as... Figure 1 As shown, an aviation equipment fault ontology model is established, search conditions are set, and fault ontology feature parameters are determined; semantic data matching of aviation equipment fault ontology conceptual feature parameters is performed; aviation equipment fault ontology data feature parameter matching is performed; weighted semantic similarity of aviation equipment fault ontology conceptual feature parameters is calculated; aviation equipment fault type is determined, and aviation equipment fault processing is assisted; the process includes the following steps: Step S1: Establish an aircraft equipment fault ontology model, set search conditions, and determine fault ontology characteristic parameters.

[0022] The knowledge source of the aviation equipment fault ontology is analyzed based on ontology modeling semantics, and the ontology is composed of five elements of concept, relation, function, axiom and instance, wherein the concept forms a classification hierarchy, and the association and constraint between the concepts are expressed by the relation, function and axiom; according to the constituting elements of the ontology, the sources of the aviation equipment fault knowledge include: fault field related standard specifications, comprehensive and professional thesauri, fault related engineering design and information management systems, fault field expert experience knowledge, fault field related books, papers, reports and other literature materials and network information resources and the like; the fault related engineering design and information management systems such as PDM, CAX, ERP, MES and the like.

[0023] As shown in Figure 2 The aviation equipment fault ontology model is a model for expressing and organizing the aviation equipment fault knowledge by the ontology method, and the aviation equipment fault ontology model constructed by the application is composed of a product domain ontology, a case domain ontology, a diagnosis domain ontology and a fault core ontology reflecting the essential features of the fault. In the aviation equipment fault ontology model, the fault core sub-ontology reflects the essential features of the fault, and is composed of five core concepts of fault product, fault mode, fault cause, fault influence and fault treatment; the product domain sub-ontology describes the concepts and relations of the fault related aviation products and equipment structures and the like, mainly including the core concepts of aviation products, equipment structures and the like and the relations thereof; the case domain sub-ontology describes the related concepts and relations of the actual fault occurrence, mainly including the core concepts of fault situation, support resources, fault business and the like and the relations thereof; the diagnosis domain sub-ontology describes the concepts and relations of the aircraft fault diagnosis model, mainly including the core concepts of system model, fault detection model, fault propagation model and the like and the relations thereof; there are cross concepts and association relations between the sub-ontologies, and the complete and consistent aviation equipment fault field knowledge is expressed.

[0024] The retrieval item of the aviation equipment fault data is converted into a feature parameter item of the aviation equipment fault ontology model; and the retrieval value of the aviation equipment fault data is converted into a feature parameter value of the fault ontology. In the aviation equipment fault ontology model, the feature parameters of the aviation equipment fault data are divided into concept feature parameters and data feature parameters. If the retrieval item of the aviation equipment fault data corresponds to the concept feature parameter of the fault ontology, step S2 is executed. If the retrieval item of the aviation equipment fault data corresponds to the data feature parameter of the fault ontology, step S3 is executed.

[0025] Step S2: expanding the semantic data of the aviation equipment fault ontology concept feature parameter and matching, as shown in Figure 3 The implementation steps of the semantic data expansion and matching of the aviation equipment fault ontology concept feature parameter in the embodiment of the application are shown in the figure, and specifically include the following sub-steps: Step S21: expanding semantic data of the concept feature parameters of the aviation equipment fault ontology, specifically including data synonym expansion, feature parameter expansion, hierarchy expansion, axiom expansion, and rule expansion, etc., to obtain a fault concept tree or a fault network graph in the fault data space of the aviation equipment field. The complex semantic data relationship between the various aviation equipment fault ontology concept feature parameters is composed of a semantic data network graph, and the concept feature parameters are nodes in the graph, and the connection between the nodes represents the relationship between the concept feature parameters. Semantic data retrieval realizes retrieval from a point of the retrieval condition, and the semantic data relationship network realizes retrieval of an aviation equipment fault semantic feature surface through semantic data expansion.

[0026] Step S22: calculating the depth factor and the density factor of the aviation equipment fault ontology concept feature parameters, and determining the semantic distance between the fault ontology concept feature parameters. The semantic distance between the aviation equipment fault ontology concept feature parameters refers to the path length connecting two concept feature parameters in a directed and weighted concept feature parameter network graph, that is, the sum of the weights of all edges on the path connecting the two concept feature parameters, which is used to quantitatively describe the strength of the association between the concept feature parameters. The semantic distance is related to the number, type, direction of the edges in the path connecting the concept feature parameter nodes, and the depth and density of the concept feature parameter nodes. In the aviation equipment fault ontology model, different semantic relationships correspond to different semantic distances, including synonymous, superordinate, subordinate, related, and other basic relationships, as well as fault cause and effect relationship, product structure relationship, problem solving relationship, product replacement relationship, case reference relationship, and other field-specific semantic relationships. The complex semantic relationship between the various concept feature parameters of the aviation equipment fault ontology is composed of a semantic data network graph, and the concept feature parameters are nodes in the graph, and the directed edges between the nodes represent the above-mentioned relationships, and the weight on the edge is the semantic distance. The superordinate and subordinate relationships are first set with a semantic distance reference value.

[0027] The embodiment of the present application determines the retrieval fault data object as aircraft fault data, as shown in Figure 4 The basic semantic relationship example graph of the fault ontology model is shown in the figure, and the aircraft fault data is composed of a product domain ontology, a case domain ontology, a diagnosis domain ontology, and a fault core ontology reflecting the fault nature characteristics. There is an overlap between the sub-field ontologies, which expresses a complete and consistent aircraft fault field. As shown in Figure 3 The field-specific semantic relationship of the fault ontology model is shown in the figure, and the semantic distance between the concepts in the aircraft fault ontology is set. The basic relationships of the aircraft fault ontology include synonym, superordinate, subordinate, and related, and the semantic distance is set as shown in Figure 4The schematic diagram illustrates the special semantic relationships in the aircraft fault ontology, including fault causal relationships, product structure relationships, problem-solving relationships, product replacement relationships, and case reference relationships. The semantic distance values ​​of various conceptual feature parameter relationships are set as shown in Table 1. The semantic distance of the superior and subordinate relationships in the table is the baseline value. When considering depth factors and density factors, depth factors and density factors need to be added for comprehensive calculation to obtain the actual value.

[0028] Table 1. Semantic distance table of various conceptual relationships in the aircraft fault ontology. Type Source concept class Target concept class Semantic distance Synonym 0 Hypernym Hyponym Superclass 3* Subclass Hyponym Superclass 5* Subclass Related Class 7 Feature parameter class Case reference Fault case 1 Fault case Product structure Aerospace product 2 Aircraft structure Product replacement Aerospace product 1 Aerospace product Fault cause Fault cause 6 Fault mode Problem solving Fault mode 6 Step S221: Calculate the depth factor of the conceptual feature parameters of the aviation equipment fault ontology. In the hierarchical fault concept tree of the aviation equipment fault ontology conceptual feature parameters, the deeper the nodes of two conceptual feature parameters are in the classification hierarchy, the closer their relationship and the smaller their semantic distance; conversely, the deeper the nodes, the larger their semantic distance. The node depth of the conceptual feature parameters of the aviation equipment fault ontology refers to the level of the node of the conceptual feature parameter in the hierarchical fault concept tree. The depth of the root node is 1, the depth of its child nodes is 2, and so on. Conceptual feature parameters The depth is root node The depth is ,and =1. Fault concept tree depth of the aircraft equipment fault entity. This refers to the maximum number of levels of nodes in the fault concept tree, i.e. As the node depth of concept feature parameters increases, the semantic distance between concept feature parameters decreases; a depth factor reflecting the depth of concept hierarchy is set. The depth factor for calculating the conceptual characteristic parameters of the aircraft equipment fault ontology is as follows: For concept feature parameters The depth of the superordinate semantic relation concept features -1 is taken as the square root and then the reciprocal is taken to obtain the concept feature parameters. Hyper-semantic relation depth factor Since the root node of the aviation equipment fault concept tree has a depth of 1, and the root node does not have a superior relationship, the conceptual feature parameters... It cannot be the root node, that is... >1; The higher-level semantic relation depth factor of the concept feature parameter C is obtained as: ; in, Conceptual feature parameters The depth factor of the higher-level semantic relation; For the conceptual characteristic parameters of the fault body of aviation equipment; Depth of the concept characteristic parameter .

[0029] Depth of the concept characteristic parameter Take the reciprocal of the square root to obtain the lower semantic relationship depth factor of the concept characteristic parameter ; Since the depth of the bottom leaf node of the aviation equipment failure concept tree is , and it does not have a lower relationship, the concept characteristic parameter cannot be the bottom leaf node, i.e. ; The lower semantic relationship depth factor of the concept characteristic parameter is: ; wherein, is the lower semantic relationship depth factor of the concept characteristic parameter ; is the depth of the aviation equipment failure concept tree ; is the depth of the aviation equipment failure concept tree .

[0030] In the aviation equipment failure concept tree, the root node does not have a higher semantic relationship, and the bottom node does not have a lower semantic relationship.

[0031] Step S222: Calculate the density factor of the aviation equipment failure ontology concept characteristic parameter; In the failure concept tree of the aviation equipment failure ontology concept characteristic parameter, the area density is generally different, i.e. the degree of detail of the concept characteristic parameter description is not uniform; The greater the density of a certain concept characteristic parameter node, the higher the degree of refinement of the concept characteristic parameter at that place, the greater the similarity between the nodes it subdivides, and the smaller the semantic distance.

[0032] The density of the aviation equipment failure ontology concept characteristic parameter node refers to the out-degree of the upper node of the concept characteristic parameter node, i.e. the number of nodes containing the concept characteristic parameter node. The density of the concept characteristic parameter is , the upper node is , and the out-degree of the upper node is , then the density of the concept characteristic parameter is , and cannot be the root node. Correspondingly, the density of the failure concept tree is the out-degree of the node containing the most child nodes, i.e. ​​​​As the node density of concept feature parameters increases, the semantic distance between concept feature parameters decreases. A density factor β, reflecting the concept density factor, is set, and the density factor of the concept feature parameters of the aviation equipment fault ontology is calculated, specifically as follows: For concept feature parameters The parent node out of degree Take the reciprocal to obtain the concept feature parameters. Hyperordinate semantic relation density factor Since the root node of the aviation equipment fault concept tree has a depth of 1 and no hierarchical relationship, the conceptual characteristic parameters... It cannot be the root node, that is... >1; Obtain conceptual feature parameters The hyperordinate semantic relation density factor is: ; in, The hyperordinate semantic relation density factor of the concept feature parameter C; This is the parent node of the concept feature parameter C; Let C be the out-degree of the parent node of the concept feature parameter C.

[0033] For concept feature parameters out of degree Taking the reciprocal, we obtain the lower semantic relation density factor of the concept feature parameter C. Due to the concept tree of aircraft equipment failure The depth of the bottom leaf nodes is Furthermore, they do not have a subordinate relationship, therefore the conceptual characteristic parameters It cannot be a bottom-level leaf node, i.e. < The lower semantic relation density factor of the concept feature parameter C is: ; in, Conceptual feature parameters The hyponym semantic relation density factor; The depth of the concept tree for aircraft equipment failures; Conceptual feature parameters The degree of departure.

[0034] Step S223: Calculate the semantic distance between the conceptual feature parameters of the aviation equipment fault ontology; in the fault concept tree of the aviation equipment fault ontology conceptual feature parameters, calculate the conceptual feature parameters. Node-to-concept feature parameters The semantic distance of a node is as follows: The benchmark value of semantic distance between hierarchical relationships The semantic distance actual value of the hypernym and hyponym semantic relation is obtained by multiplying the comprehensive adjustment coefficient of the depth factor and the density factor The comprehensive adjustment coefficient of the depth factor and the density factor of the hypernym and hyponym semantic relation is the depth factor of the hypernym and hyponym semantic relation of the concept characteristic parameter The depth factor of the hypernym and hyponym semantic relation of the concept characteristic parameter The density factor The adjustment coefficient The adjustment coefficient The adjustment coefficient is used to adjust the importance of the density factor relative to the depth factor, and the value range is [0, 1]; according to the benchmark value, the depth factor and the density factor of the hypernym and hyponym semantic relation, the semantic distance actual value of the hypernym and hyponym semantic relation of each level concept is calculated as follows: ; Wherein, The semantic distance actual value of the hypernym and hyponym semantic relation; The benchmark value of the hypernym semantic distance should be coordinated with the semantic distance fixed value of other relations; The depth factor of the hypernym and hyponym semantic relation of the concept characteristic parameter The depth factor of the hypernym and hyponym semantic relation of the concept characteristic parameter ; The density factor of the hypernym and hyponym semantic relation of the concept characteristic parameter The density factor of the hypernym and hyponym semantic relation of the concept characteristic parameter ; The density factor of the hypernym and hyponym semantic relation of the concept characteristic parameter ; The adjustment coefficient adjusts the importance of the density factor relative to the depth factor, and the value range of the adjustment coefficient is [0, 1].

[0035] As shown in Fault handling , the method is applied to calculate the concept depth factor and the concept density factor between each concept in the aircraft fault ontology, and then the actual semantic distance of the hypernym and hyponym relation is obtained. The comprehensive semantic distance between the concept characteristic parameters of the aviation equipment fault ontology is calculated; the semantic distance between the concept characteristic parameter and the concept characteristic parameter is the shortest path from the concept characteristic parameter node to the concept characteristic parameter node, and the Floyd method is used to obtain the shortest path from the concept characteristic parameter node to the concept characteristic parameter node in the ontology concept characteristic parameter graph, and the weight sum of the shortest path is the shortest path from the concept characteristic parameter to the concept characteristic parameter ​The semantic distance. The Floyd method, also known as the interpolation method, is a method that uses dynamic programming to find the shortest path between multiple source points in a given weighted graph. Conceptual feature parameters in the fault ontology. arrive Comprehensive semantic distance for arrive The shortest path is the sum of the semantic distances of all directed edges. After determining the weights of all directed edges in the aviation equipment fault ontology network graph, the semantic distance between two concept feature parameter nodes is the shortest weighted path length; the comprehensive semantic distance between two concept feature parameters in the aviation equipment fault ontology is obtained as follows: ; in, Conceptual feature parameters in the fault ontology arrive The comprehensive semantic distance; Conceptual feature parameters Node to The sum of weighted distances of all edges in the shortest path of a node; Conceptual feature parameters arrive semantic distance; Conceptual feature parameters arrive The shortest path is a directed edge; Number the directed edges of the shortest path; This represents the total number of directed edges in the shortest path. These are the first concept feature parameters; These are the characteristic parameters of the second concept.

[0036] like Figure 4 The figure shows the calculation of the comprehensive semantic distance between two concepts using the comprehensive semantic distance calculation function proposed in this invention. In the example, the shortest path value of the semantic distance from "aviation product" to "support resource" is 7; the semantic distance from "avionics product" to "mechanical product" is 2.42+4.04=6.46, and the semantic distance from "support resource" to "failure cause" is 3.13+5.21=8.34. It can be seen that the semantic distance between co-occurring concepts with deeper levels is small; the semantic distance between the synonyms "engine" and "engine engine" is 0.

[0037] Step S23: Calculate the semantic data similarity between the conceptual feature parameters of the aviation equipment fault ontology, and perform semantic data matching of the conceptual feature parameters of the aviation equipment fault ontology. The core of semantic data matching lies in correctly and objectively setting and quantifying the degree of semantic data similarity between conceptual feature parameters. Semantic data similarity is the substitutability and semantic conformity of the vocabulary of conceptual feature parameters. It reflects the similarity of two conceptual feature parameters in semantic data. The higher the semantic similarity, the higher the matching degree. The semantic similarity between two conceptual feature parameters can be measured by the semantic distance between them. That is, the smaller the semantic distance between two conceptual feature parameters, the greater their semantic similarity. The purpose of calculating the semantic distance between the conceptual feature parameters of the aviation equipment fault ontology is to calculate the semantic similarity between them. The semantic similarity function is used to convert the semantic distance between the conceptual feature parameters of the aviation equipment fault ontology into the semantic similarity between them.

[0038] Calculate the maximum semantic distance between conceptual feature parameters in the aviation equipment fault ontology. In the aviation equipment fault ontology, all conceptual feature parameters share a common root node. When the semantic distance between two conceptual feature parameters is equal to the sum of their semantic distances to the root node, it indicates that the two conceptual feature parameters have little similarity other than being conceptual feature parameters. This situation is defined as having a similarity of 0, i.e., the maximum semantic distance. The maximum semantic distance between conceptual feature parameters in the aviation equipment fault ontology is the sum of the semantic distance from one conceptual feature parameter node to the root node through its superior relation and the semantic distance from the root node to another conceptual feature parameter node through its subordinate relation. To concept feature parameters maximum semantic distance These are conceptual feature parameters in the fault ontology. to the root node semantic distance With the root node To concept feature parameters semantic distance The summation of the features. Conceptual characteristic parameters. Node-to-concept feature parameters The maximum semantic distance between nodes is specifically: ; in, Conceptual feature parameters To concept feature parameters The maximum semantic distance; This is the root node of the aviation equipment fault concept tree; Conceptual feature parameters in the fault ontology to the root node semantic distance; The root node in the faulty entity To concept feature parameters Semantic distance.

[0039] Since a subclass in the aircraft equipment body can inherit multiple superclasses, when calculating the maximum semantic distance, it is necessary to first determine the fault concept tree to which the two concept feature parameters belong, and take the fault concept tree with the smallest node depth of the concept feature parameters as the standard.

[0040] Calculate the semantic data similarity of conceptual feature parameters of the fault ontology of aviation equipment. The semantic similarity calculation function must satisfy several characteristics: the function output range is [0, 1]; the semantic similarity is 1 when the semantic distance is 0; the semantic similarity is 0 when the semantic distance is the maximum semantic distance Dmax; semantic similarity decreases with semantic distance, that is, semantic similarity decreases as the semantic distance increases. Conceptual feature parameters in the fault ontology. With concept feature parameters semantic similarity between It is 1 minus the conceptual feature parameters in the fault entity. To concept feature parameters Comprehensive semantic distance and arrive The opening of the maximum ratio of semantic distance Power; Regulatory factor This is used to adjust the rate at which semantic similarity decreases as semantic distance increases, and its value range is positive integers; the semantic similarity calculation function proposed in this invention is: ; in, Conceptual feature parameters in the fault ontology With concept feature parameters Semantic similarity between them; It is a moderating factor used to adjust the rate at which semantic similarity decreases as semantic distance increases; The root node in the faulty entity To concept feature parameters Semantic distance.

[0041] By applying the computational function proposed in this invention, the maximum semantic distance between concept feature parameters is calculated, thereby obtaining the semantic similarity between concept feature parameters. (Refer to...) Figure 5 If we take the parameter θ=1 in the semantic similarity calculation function, the following is an example of similarity calculation between some conceptual feature parameter classes of the aircraft fault ontology.

[0042] The semantic distance of "aviation product" to "guarantee resource" concept characteristic parameter is 7, the maximum semantic distance is 3.13+5.21=8.34, and the semantic similarity of "aviation product" to "guarantee resource" concept characteristic parameter is 1-(7 / 8.34)=0.16.

[0043] The semantic distance of "aviation product" to "guarantee resource" concept characteristic parameter is 7, the maximum semantic distance is 3.13+5.21=8.34, and the semantic similarity of "aviation product" to "guarantee resource" concept characteristic parameter is 1-(7 / 8.34)=0.16.

[0044] The semantic distance of "aviation product" to "guarantee resource" concept characteristic parameter is 7, the maximum semantic distance is 3.13+5.21=8.34, and the semantic similarity of "aviation product" to "guarantee resource" concept characteristic parameter is 1-(7 / 8.34)=0.16.

[0045] The semantic distance of "aviation product" to "guarantee resource" concept characteristic parameter is 7, the maximum semantic distance is 3.13+5.21=8.34, and the semantic similarity of "aviation product" to "guarantee resource" concept characteristic parameter is 1-(7 / 8.34)=0.16.

[0046] Step S3: Perform aviation equipment fault ontology data characteristic parameter matching; as Figure 5 Figure 4 shown is the matching implementation step of the fault ontology data characteristic parameter, the value range of the data characteristic parameter is various data types, and the data types in the aviation equipment fault ontology include Boolean type, enumeration type, numerical type, date and time type, and string type. The semantic similarity matching method of the data characteristic parameter includes corresponding matching, fuzzy matching and text matching, and a suitable matching method is adopted according to the actual application situation; specifically, the following sub-steps are included: Step S31: Perform corresponding matching of aviation equipment fault ontology data characteristic parameters. Since the quantifiable data characteristic parameters are themselves an accurate description of the state of things, they can be directly calculated, and in principle, each data value is not similar to other data values that are not equal. Therefore, for quantifiable data characteristic parameters, a corresponding matching method is used to calculate the similarity. In the fault ontology, the data characteristic parameters of the data types of Boolean type, enumeration type, numerical type and date and time type are matched, and the similarity function of the corresponding matching of the data characteristic parameters is specifically: ; wherein, is the corresponding matching semantic similarity between the data characteristic parameter and the data characteristic parameter . The first aviation equipment failure data characteristic parameter is; The second aviation equipment failure data characteristic parameter is.

[0047] Step S32: fuzzy matching of the aviation equipment failure ontology data characteristic parameter. The value range of the Boolean type and enumeration type data characteristic parameter type is discrete, while the value range of the numerical type and date time type data characteristic parameter is discrete and continuous. For the quantifiable continuous data value range, interval division is performed, and the data characteristic parameters in one interval have certain similarity, and the data characteristic parameters not in the same interval are not similar, that is, fuzzy matching of the data characteristic parameter. Fuzzy matching avoids the loss of semantic similarity in a certain range of corresponding matching, and can achieve a better semantic matching result. In the aviation equipment failure field, a large number of uncertain factors exist, and statistical data are often described by fuzzy interval values or fuzzy semantics, for example, the mean time between failures (MTBF) of a product is described as “about 400 to 500 hours”, and the fuzzy description of “low failure rate” is also used. The limited interval of the divided domain is calculated, and the similarity of two data characteristic parameters is 0. The similarity function of the data characteristic parameter fuzzy matching is specific as follows: ; Wherein, is the fuzzy matching semantic similarity between the data characteristic parameter and the data characteristic parameter ; is the fuzzy limited interval; is the fuzzy limited interval length.

[0048] The data fuzzy matching example of the embodiment of the application for matching the data characteristic parameter in the aircraft failure ontology is as follows. The characteristic parameter value of the fluid pressure characteristic parameter of a certain duct is divided into three intervals, that is, [0, 20), [20, 60), and [60, +∞), and the unit is Mpa. Then, if the data characteristic parameter = 30, the data characteristic parameter = 50, since , are in the interval [20, 60), the similarity is 1- |30-50| / |60-20|=1-20 / 40=0.5. If the data characteristic parameter = 30, the data characteristic parameter = 80, since , are not in the same set interval, the similarity is 0.

[0049] Step S33: Perform aerial equipment failure ontology text data feature parameter matching. The string text type data feature parameter is difficult to match, so the text type data feature parameter is quantified and similarity is calculated by using the word matching method. The text data feature parameter is first processed by word segmentation and replacement, and then the real words contained in the two text data feature parameters are compared to calculate the matching degree of the two text data, specifically: The text data feature parameter is decomposed by word, and the virtual word is removed; the word in the synonym dictionary is replaced after the word segmentation processing, so that the words with the same meaning have the same expression; the synonym dictionary is obtained from the comprehensive or professional thesaurus, such as NASA thesaurus, defense science and technology thesaurus, aviation science and technology data thesaurus, and military logistics thesaurus. The real word matching of the text data feature parameter is performed; the similarity of the text data feature parameter is calculated, and the number of real words that are the same or synonymous in the two text data feature parameters is divided by the average number of real words in the two text data feature parameters. The similarity of the text data feature parameter is calculated, specifically: ; Wherein, is the text data semantic similarity between the text data feature parameter and the text data feature parameter . is the first aerial equipment failure text data feature parameter; is the first aerial equipment failure text data feature parameter; is the number of real words in the text data feature parameter . is the number of real words in the text data feature parameter . is the number of real words that are substantially the same in the text data feature parameter and the text data feature parameter . In the embodiment, substantially the same means the same or synonymous.

[0050] The aerial equipment failure text data feature parameter matching example in the embodiment of the application is: as described in the failure of "2 number engine's thrust cable wear and tear fracture" and "right thrust cable fracture", the semantics are completely the same, but the expression of the two text data is not completely the same, and direct text data matching processing cannot conclude that they are the same, therefore, the text data is first processed by word segmentation and replacement, and then the matching degree of the two text data is calculated by comparing the real words contained in the two text data. The word segmentation processing, i.e. the text data is decomposed by word, and the virtual word is removed, for example, = "2 number engine's thrust cable wear and tear fracture", after word segmentation, it is "2nd engine ~ thrust cable ~ wear ~ break"; replacement processing, replacing the word processed after the word segmentation with the word in the synonym dictionary, such as replacing "2nd engine" with "right engine" and replacing "thrust cable" with "thrust cable", the purpose is to make the words with the same semantics have the same expression; real word matching, that is, calculating the similarity of text data feature parameters, applying the text matching similarity function proposed in the application, after word segmentation and replacement processing, "right engine ~ thrust cable ~ wear ~ break", "right engine ~ thrust cable ~ break", Count( )=4, Count( )=3, Count(S)=3, the matching degree of the text data and is 0.86.

[0051] Step S4: After the processing of step S2 and step S3 of the ontology concept feature parameters and data feature parameters corresponding to all search conditions are performed, all feature parameter similarities of each result instance are added, and then the normalization processing of all search feature parameter weights is performed, the semantic similarity of the search result instance corresponding to all search conditions is obtained, and the weighted semantic similarity of the aviation equipment fault ontology concept feature parameter is calculated.

[0052] The semantic search of the aviation equipment fault ontology is the instance search of the aviation equipment fault field concept feature parameter corresponding to the search condition, that is, the similarity of different instances of the same concept feature parameter is calculated, which is the comprehensive calculation result of the feature parameter similarity between instances; the weighted semantic similarity of the aviation equipment fault ontology concept feature parameter calculated is: ; wherein, is the similarity between the instance and the instance in the aviation equipment fault ontology; is the weight of the th feature parameter; is the semantic similarity function of the th feature parameter value, corresponding to the concept similarity of the concept feature parameter or the semantic similarity of the data feature parameter; is the first aviation equipment fault ontology instance; is the second aviation equipment fault ontology instance; is the search feature parameter number; is the total number of search feature parameters.

[0053] The embodiment of the present application calculates the weighted semantic similarity of the concept feature parameters of the aviation equipment fault ontology after calculating the concept feature parameter similarity and the data feature parameter similarity of all the search conditions, and then outputs the semantic search result instance set sorted by the similarity. In the aircraft fault data search, the fault part is "landing gear lifting cylinder hydraulic conduit", the corresponding ontology concept depth is 5, and the density is 6; the fault mode is "leakage", the corresponding ontology concept depth is 6, and the density is 3; the occurrence time is "static maintenance" enumerated data feature parameter, and the corresponding matching is used as the search condition. The semantic search is performed on the real aviation equipment fault case library handled by the enterprise in a certain time period, 6 semantic similar fault cases are obtained, and the similarity is 1.0, 0.87, 0.76, 0.76, 0.69 and 0.63. The feature parameter values of the result fault case with the similarity of 1.0 are completely the same as the search condition, and the feature parameter values of the other result cases are different from the search condition. In the process of handling the fault problem, the actual situation of the aviation equipment fault problem is combined, and the appropriate case is selected from the similar case set obtained by the semantic search as the troubleshooting reference.

[0054] Step S5: According to the semantic search result of the aviation equipment fault, the aviation equipment fault type is judged. The result instances with the semantic similarity less than the similarity threshold value in step S4 are removed, and then the final semantic search result instance set is obtained, and the aviation equipment fault ontology semantic search result is outputted in the order of the similarity from large to small.

[0055] Before the aviation equipment fault ontology instance matching semantic search is performed, the search item, that is, the concept feature parameter, is determined, the weight of each concept feature parameter and the semantic similarity threshold value of the search result are set. The concept feature parameter is the property performance of a certain aspect of the concept, the concept feature parameter weight reflects the importance of each feature of the concept, and the semantic similarity threshold value is used for screening the search result to improve the semantic search accuracy.

[0056] The second aspect of the embodiment of the present application provides a fault analysis system of the aviation equipment fault analysis method based on the ontology semantic search, which comprises a fault ontology module, an attribute weight and similarity threshold value setting module, a fault data semantic search module and an aviation equipment fault judgment output module.

[0057] The fault ontology module is used for operating and viewing the aviation equipment fault ontology data, including the construction, graphical display, consistency inspection and reasoning rule setting of the structure and content of the concepts, relationships, attributes, instances and the like of the aviation equipment fault ontology, and provides a basis for the semantic search of the aviation equipment fault.

[0058] The attribute weight and similarity threshold setting module is configured to set the weight of the aviation equipment fault data retrieval item corresponding to the aviation equipment fault ontology concept attribute and the semantic similarity threshold of the retrieval result, support the semantic similarity calculation of the aviation equipment fault data and the retrieval condition, and the screening of the retrieval result, so as to improve the semantic retrieval accuracy.

[0059] The fault data semantic retrieval module is an interactive module of the fault analysis system and the semantic retrieval. By combining the aviation equipment fault ontology model and the semantic retrieval, the concept feature parameter matching similarity method based on the semantic distance is realized. The retrieval condition is input in the module, the semantic similarity calculation is performed by calling the fault ontology module and the attribute weight and similarity threshold setting module, and the aviation equipment fault ontology semantic retrieval result is output.

[0060] The aviation equipment fault judgment output module compares the aviation equipment fault ontology semantic retrieval result output by the fault data semantic retrieval module with the aviation equipment fault condition, and judges the specific fault type of the aviation equipment.

[0061] The embodiment of the present application has the following advantages: the present application introduces the fault data ontology specification concept set to solve the problems of low retrieval efficiency, lack of accuracy and comprehensiveness of the retrieval result of the traditional information retrieval technology, standardizes the retrieval condition, converts it into the concept in the ontology rule constraint, and changes the actual fault data retrieval into the matching of the concept and the semantics in the fault ontology library. It is a search and reasoning process on the semantic layer of the fault data. Through the description of the embodiment, it is proved that the method can solve the problems of comprehensive and accurate retrieval of the aviation equipment fault data, and effectively improve the recall rate and precision rate of the fault data retrieval. The present application establishes the ontology-based semantic retrieval model from the aspects of semantic expansion and semantic matching, proposes the concept matching similarity method based on the semantic distance combined with the depth and density factors of the ontology model, and the semantic retrieval method and the implementation process based on the method. The fault data semantic retrieval application support is provided for the user through the function of the fault analysis system, the comprehensive and accurate retrieval of the fault data is realized, strong support is provided for solving the fault problem, and it is proved through the implementation process of the embodiment that the method can better realize the aviation equipment fault retrieval and meet the actual engineering requirements.

[0062] The above-described embodiments are only used to describe the preferred embodiments of the present application, and do not limit the scope of the present application. Without departing from the design spirit of the present application, various modifications and improvements of the technical solutions of the present application made by those skilled in the art shall fall within the protection scope of the present application determined by the claims.

Claims

1. An aviation equipment fault analysis method based on ontology semantic retrieval, characterized in that: It comprises the following steps: S1: Establishing an aviation equipment fault ontology model, setting retrieval conditions, and determining fault ontology characteristic parameters; If the retrieval item of the aviation equipment fault data corresponds to the concept characteristic parameters of the fault ontology, step S2 is executed; if the retrieval item of the aviation equipment fault data corresponds to the data characteristic parameters of the fault ontology, step S3 is executed; S2: Expanding the semantic data of the aviation equipment fault ontology concept characteristic parameters, calculating the depth factor and the density factor of the aviation equipment fault ontology concept characteristic parameters, and determining the semantic distance between the fault ontology concept characteristic parameters; calculating the semantic similarity between the aviation equipment fault ontology concept characteristic parameters, and performing semantic matching of the aviation equipment fault ontology concept characteristic parameters; S3: Performing aviation equipment fault ontology data characteristic parameter matching, specifically including: corresponding matching of data characteristic parameters, fuzzy matching of data characteristic parameters, and text data characteristic parameter matching; S4: Executing steps S2 and S3 for the aviation equipment fault ontology concept characteristic parameters and data characteristic parameters respectively, adding the obtained characteristic parameter similarities, performing normalization processing on the weight of all retrieval characteristic parameters, calculating the weighted semantic similarity of the aviation equipment fault ontology concept characteristic parameters, specifically: ； in, Examples of faults in aviation equipment and examples Weighted semantic similarity between them; For the first The weights of each feature parameter; For the first A semantic similarity function for each feature parameter value; This is an example of a fault in the first type of aviation equipment. This is an example of a faulty component in the second type of aviation equipment. For the retrieval feature parameter number; To retrieve the total number of feature parameters; S5: According to the semantic retrieval result of the aviation equipment fault in step S4, the type of the aviation equipment fault is determined.

2. The ontology-based semantic retrieval based aviation equipment failure analysis method according to claim 1, characterized in that: Step S2 specifically comprises: S21: Expanding the semantic data of the aviation equipment fault ontology concept characteristic parameters, specifically including: data synonym expansion, characteristic parameter expansion, hierarchical expansion, axiom expansion, and rule expansion; S22: Calculating the depth factor and the density factor of the aviation equipment fault ontology concept characteristic parameters, and determining the semantic distance between the fault ontology concept characteristic parameters; S23: Calculating the semantic similarity between the aviation equipment fault ontology concept characteristic parameters, and performing semantic matching of the aviation equipment fault ontology concept characteristic parameters.

3. The ontology-based semantic retrieval based aviation equipment failure analysis method according to claim 2, characterized in that: Step S22 specifically comprises: S221: determining the depth of the aviation equipment fault ontology concept characteristic parameter The depth of the aviation equipment fault ontology concept characteristic parameter is ; set the depth factor of the concept hierarchical depth factor , calculate the depth factor of the aviation equipment fault ontology concept characteristic parameter; S222: determining the density of the aviation equipment fault ontology concept characteristic parameter The density of the aviation equipment fault ontology concept characteristic parameter is ; a density factor β of the concept density factor is set, and a density factor of the aviation equipment fault ontology concept characteristic parameter is calculated; S223: In the fault concept tree of the fault ontology concept feature parameters of the aviation equipment, the fault ontology concept feature parameters of the aviation equipment are calculated Node to concept feature parameter Semantic distance of the node, to obtain the comprehensive semantic distance between two concept feature parameters in the aviation equipment fault ontology .

4. The ontology-based semantic retrieval based aviation equipment failure analysis method according to claim 3, characterized in that: The depth factor of the aviation equipment fault ontology concept characteristic parameters in step S221 specifically includes: the upper semantic relationship depth factor and the lower semantic relationship depth factor; the upper semantic relationship depth factor is determined through the root node of the aviation equipment fault concept tree, and the lower semantic relationship depth factor is determined through the bottom layer leaf node of the aviation equipment fault concept tree, specifically: ； ； wherein, is a superordinate semantic relation depth factor for the concept feature parameter is an aeronautical equipment failure ontology concept feature parameter; is a superordinate semantic relation depth factor for the concept feature parameter is a superordinate semantic relation depth factor for the concept feature parameter is a depth of an aeronautical equipment failure concept tree is an aeronautical equipment failure concept tree.​​​​ 5. The ontology-based semantic retrieval based aviation equipment failure analysis method according to claim 3, characterized in that: The density factor of the aviation equipment fault ontology concept characteristic parameters in step S222 includes: the upper semantic relationship density factor and the lower semantic relationship density factor; the upper semantic relationship density factor is determined through the root node of the aviation equipment fault concept tree, and the lower semantic relationship density factor is determined through the bottom layer leaf node of the aviation equipment fault concept tree, specifically: ； ； wherein, is a superordinate semantic relation density factor for the concept feature parameter C; is a superordinate node for the concept feature parameter C; is an out-degree of the superordinate node for the concept feature parameter C; is a subordinate semantic relation density factor for the concept feature parameter is an out-degree of the subordinate node for the concept feature parameter .​ 6. The ontology-based semantic retrieval based aviation equipment failure analysis method according to claim 3, characterized in that: Step S223 multiplies the reference value of the semantic distance of the hypernym-hyponym relationship by the adjustment coefficient of the depth factor and the density factor to obtain the actual value of the semantic distance of the hypernym-hyponym relationship ; the adjustment coefficient of the depth factor and the density factor is the concept feature parameter ; the depth factor of the hypernym-hyponym relationship of the concept feature parameter ; the density factor is added ; and the adjustment coefficient is multiplied ; and the actual value of the semantic distance of the hypernym-hyponym relationship is obtained , specifically as follows: ； wherein, is the semantic distance actual value for the hyponym-hypernym semantic relationship; is the semantic distance reference value for the hyponym-hypernym relationship; is the depth factor for the hyponym-hypernym semantic relationship of the concept feature parameter is the density factor for the hyponym-hypernym semantic relationship of the concept feature parameter is the density factor for the hyponym-hypernym semantic relationship of the concept feature parameter is the density factor for the hyponym-hypernym semantic relationship of the concept feature parameter is the adjustment coefficient.

7. The ontology-based semantic retrieval based aviation equipment failure analysis method according to claim 2, characterized in that: Step S23 specifically comprises: The comprehensive semantic distance between the concept characteristic parameters in the fault ontology is calculated, the comprehensive semantic distance being the sum of the semantic distances of all directed edges of the shortest path between the concept characteristic parameters, the semantic similarity between the concept characteristic parameters in the fault ontology is calculated, and the specific steps are as follows: ； ； ； wherein, is the semantic distance of the concept feature parameter in the fault ontology to the root node ; is the semantic distance of the concept feature parameter to ; is the shortest path directed edge of the concept feature parameter to ; is the shortest path directed edge number; is the total number of shortest path directed edges; is the first concept feature parameter; is the second concept feature parameter; is the semantic distance of the concept feature parameter to the concept feature parameter ; is the root node of the aviation equipment fault concept tree; is the semantic distance of the concept feature parameter in the fault ontology to the root node ; is the semantic distance of the root node in the fault ontology to the concept feature parameter ; is the semantic similarity between the concept feature parameter in the fault ontology and the concept feature parameter ; is the adjustment factor; is the semantic distance of the concept feature parameter in the fault ontology to the concept feature parameter .​​​​​​​​ 8. The ontology-based semantic retrieval based aviation equipment failure analysis method according to claim 1, characterized in that: Step S3 specifically comprises: S31: The corresponding matching of the aviation equipment fault ontology data feature parameters is the data corresponding matching of the Boolean type, enumeration type, numerical type and date time type data feature parameters, to obtain the similarity function of the data feature parameter corresponding matching; S32: The fuzzy matching of the aviation equipment fault ontology data feature parameters needs to divide the limited interval of the set domain, and the similarity of the two data feature parameters is obtained through the similarity function of the data feature parameter fuzzy matching; S33: The text type data feature parameter is quantified and similarity calculation is performed in the way of vocabulary matching; the real word matching of the text data feature parameter is performed; and the text data feature parameter similarity is obtained.

9. The ontology-based semantic retrieval based aviation equipment failure analysis method according to claim 1, characterized in that: Step S5 is specifically: setting the weight of each concept feature parameter and the semantic similarity threshold of the retrieval result, which is used for filtering the retrieval result to improve the semantic retrieval accuracy; eliminating the result instance with a semantic similarity less than the similarity threshold in step S4, thereby obtaining the final semantic retrieval result instance set, sorting and outputting the aviation equipment fault ontology semantic retrieval result according to the similarity from large to small; and judging the aviation equipment fault type according to the aviation equipment fault semantic retrieval result.

10. A failure analysis system for the ontology-based semantic retrieval based failure analysis method of one of claims 1 to 9, characterized in that It comprises: a fault ontology module, an attribute weight and similarity threshold setting module, a fault data semantic retrieval module and an aviation equipment fault judgment output module; The fault ontology module is used for operating and viewing the aviation equipment fault ontology data, including the structure construction of the concepts, relationships, attributes and instances of the aviation equipment fault ontology; The attribute weight and similarity threshold setting module is used for setting the weight of the aviation equipment fault data retrieval item corresponding to the aviation equipment fault ontology concept attribute and the semantic similarity threshold of the retrieval result, so as to perform the semantic similarity calculation of the aviation equipment fault data and the retrieval condition and the filtering of the retrieval result; The fault data semantic retrieval module is the interactive module of the fault analysis system and the semantic retrieval, which realizes the concept feature parameter matching similarity method based on the semantic distance by combining the aviation equipment fault ontology model and the semantic retrieval, inputs the retrieval condition, calls the fault ontology module and the attribute weight and similarity threshold setting module to perform the semantic similarity calculation, and outputs the aviation equipment fault ontology semantic retrieval result; The aviation equipment fault judgment output module compares the output aviation equipment fault ontology semantic retrieval result of the fault data semantic retrieval module with the aviation equipment fault condition, and judges the fault type of the aviation equipment.

Images