Auxiliary troubleshooting method of air traffic control equipment and operation and maintenance terminal

Abstract

The invention discloses an air traffic control equipment auxiliary troubleshooting method and an operation and maintenance terminal. The auxiliary troubleshooting method for the air traffic control equipment comprises the following steps: acquiring fault information, and acquiring a processing suggestion from a knowledge base according to the fault information; feature quantity extraction is carried out on the fault information and the processing suggestion, a fault text representation vector is obtained from the fault information, and a suggestion text representation vector is obtained from the processing suggestion; comparing the fault text representation vector with the suggestion text representation vector to obtain the semantic similarity of the fault text representation vector and the suggestion text representation vector; and forming a processing decision suggestion according to the semantic similarity, and performing troubleshooting on the air traffic control equipment according to the processing decision suggestion. According to the technical scheme, the fault processing efficiency can be effectively improved, and maintenance of the air traffic control equipment can be rapidly completed.

Description

Technical Field

[0001] The present invention relates to the technical field of air traffic control maintenance, and particularly to an auxiliary fault troubleshooting method for air traffic control equipment and an operation and maintenance terminal. Background Art

[0002] With the rapid development of the national economy, the civil aviation traffic volume has been showing a continuous growth trend. To ensure the safe operation and efficient operation of flights, stricter requirements are put forward for the stability of air traffic control equipment and the timeliness of handling equipment failures.

[0003] However, at present, when operation and maintenance personnel troubleshoot air traffic control equipment, they usually rely on their own experience to handle it, resulting in low efficiency in handling faults and difficulty in quickly completing the maintenance of air traffic control equipment. Summary of the Invention

[0004] Aiming at the defects in the prior art, the present invention provides an auxiliary fault troubleshooting method for air traffic control equipment, which can effectively improve the fault handling efficiency and quickly complete the maintenance of air traffic control equipment.

[0005] An auxiliary fault troubleshooting method for air traffic control equipment provided by the present application, the auxiliary fault troubleshooting method for air traffic control equipment includes:

[0006] Obtain fault information, and obtain handling suggestions from a knowledge base according to the fault information;

[0007] Extract feature quantities from the fault information and the handling suggestions respectively, obtain a fault text representation vector from the fault information, and obtain a suggestion text representation vector from the handling suggestions;

[0008] Compare the fault text representation vector and the suggestion text representation vector to obtain the semantic similarity between the two;

[0009] Form a handling decision suggestion according to the semantic similarity, and troubleshoot the air traffic control equipment according to the handling decision suggestion.

[0010] In one aspect, before the step of obtaining fault information, it includes:

[0011] Construct a knowledge base, which includes an expert database, a typical case database and an experience database.

[0012] In one aspect, the step of forming a handling decision suggestion according to the semantic similarity includes:

[0013] Sort in descending order according to the magnitude of the semantic similarity;

[0014] Obtain the top N handling suggestions corresponding to the semantic similarity to form a handling decision suggestion, where N is an integer greater than or equal to 1.

[0015] In one aspect, the fault information includes the device name;

[0016] The step of obtaining a processing suggestion from the knowledge base based on the fault information includes:

[0017] Using the device name as an index, screening out processing suggestions from the knowledge base based on the fault phenomenon and / or fault code, and the fault information further includes the fault phenomenon and / or the fault code.

[0018] In one aspect, the step of respectively extracting feature quantities from the fault information and the processing suggestion includes:

[0019] Using a pre-trained tokenizer to perform tokenization on the fault information and the processing suggestion;

[0020] Using a pre-trained language model to respectively extract features from the fault information and the processing suggestion after tokenization.

[0021] In one aspect, the step of comparing the fault text representation vector and the suggestion text representation vector to obtain the semantic similarity between the two includes:

[0022] Calculating the cosine similarity between the fault text representation vector and the suggestion text representation vector item by item to obtain the semantic similarity between the two.

[0023] In one aspect, defining the fault text representation vector as V query , the suggestion text representation vector n is the number of relevant troubleshooting solution plans corresponding in the knowledge base, then it satisfies:

[0024]

[0025] w n is the knowledge weight of the knowledge base to which the nth relevant processing suggestion belongs, is the text semantic similarity, is the semantic similarity obtained by weighting the text semantic similarity.

[0026] In one aspect, the knowledge weight of the expert database is W 1 , the knowledge weight of the troubleshooting solution plan in the typical case database is W 2 , the knowledge weight of the troubleshooting solution plan in the experience database is W 3 , then it satisfies: W 1 >W 2 >W 3 .

[0027] In one aspect, the step of forming a processing decision recommendation based on the semantic similarity includes:

[0028] Compare the semantic similarity with a preset standard similarity, and combine the corresponding processing recommendations greater than or equal to the standard similarity to form a processing decision.

[0029] In addition, to solve the above problems, the present application also provides an operation and maintenance terminal, which includes:

[0030] A data entry module, which is used to obtain fault information and obtain processing recommendations from a knowledge base according to the fault information;

[0031] An extraction module, which is used to extract feature quantities from the fault information and the processing recommendations respectively, obtain a fault text representation vector from the fault information, and obtain a recommendation text representation vector from the processing recommendations;

[0032] A comparison module, which is used to compare the fault text representation vector and the recommendation text representation vector to obtain the semantic similarity between the two;

[0033] An auxiliary decision-making module, which is used to sort according to the magnitude of the semantic similarity to form a processing decision recommendation, and troubleshoot air traffic control equipment according to the processing decision recommendation.

[0034] The beneficial effects of the present invention are reflected in: when troubleshooting, processing recommendations are extracted from the knowledge base through fault information. Feature quantities are extracted from the fault information to obtain a fault text representation vector, and feature quantities are extracted from the processing recommendations to obtain a recommendation text representation vector. Then, the fault text representation vector and the recommendation text representation vector are compared to obtain the semantic similarity between the two, and relevant processing recommendations are obtained according to the semantic similarity to form a processing decision recommendation. Maintenance personnel troubleshoot air traffic control equipment according to the processing decision recommendation. It can be seen that the technical solution of the present application can automatically provide processing decision recommendations based on fault information, thereby improving the fault handling efficiency and quickly completing the maintenance of air traffic control equipment. Description of the Drawings

[0035] In order to more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the following will briefly introduce the drawings required for use in the description of the specific embodiments or the prior art. In all the drawings, similar elements or parts are generally identified by similar reference numerals. In the drawings, the elements or parts are not necessarily drawn to scale.

[0036] Figure 1 It is a schematic flow chart of the steps of the auxiliary fault troubleshooting method for the air traffic control equipment of the present application;

[0037] Figure 2 It is a schematic diagram of the process steps of the component knowledge base in the auxiliary fault troubleshooting method for the air traffic control equipment of this application;

[0038] Figure 3 It is a schematic diagram of the process steps of forming a processing decision suggestion in the auxiliary fault troubleshooting method for the air traffic control equipment of this application;

[0039] Figure 4 It is a schematic diagram of the process steps of obtaining a processing suggestion in the auxiliary fault troubleshooting method for the air traffic control equipment of this application;

[0040] Figure 5 It is a schematic diagram of the process steps of feature extraction in the auxiliary fault troubleshooting method for the air traffic control equipment of this application;

[0041] Figure 6 It is a schematic diagram of the process steps of semantic similarity calculation in the auxiliary fault troubleshooting method for the air traffic control equipment of this application;

[0042] Figure 7 It is a schematic diagram of the process steps of another embodiment of semantic similarity calculation in the auxiliary fault troubleshooting method for the air traffic control equipment of this application;

[0043] Figure 8 It is a schematic diagram of the content of the knowledge base in the auxiliary fault troubleshooting method for the air traffic control equipment of this application;

[0044] Figure 9 It is a schematic diagram of the expert database in the auxiliary fault troubleshooting method for the air traffic control equipment of this application;

[0045] Figure 10 It is a schematic diagram of the typical case database in the auxiliary fault troubleshooting method for the air traffic control equipment of this application;

[0046] Figure 11 It is a schematic diagram of the experience database in the auxiliary fault troubleshooting method for the air traffic control equipment of this application;

[0047] Figure 12 It is a schematic diagram of the decision suggestion in the auxiliary fault troubleshooting method for the air traffic control equipment of this application;

[0048] Figure 13 It is a schematic diagram of the module structure of the operation and maintenance terminal of this application.

[0049] Description of the drawings: 10. Operation and maintenance terminal; 100. Data entry module; 200. Extraction module; 300. Comparison module; 400. Auxiliary decision-making module. Detailed implementation manners

[0050] The embodiments of the technical solutions of the present invention will be described in detail below with reference to the accompanying drawings. The following embodiments are only used to more clearly illustrate the technical solutions of the present invention, so they are only examples and cannot be used to limit the protection scope of the present invention.

[0051] It should be noted that unless otherwise specified, the technical terms or scientific terms used in this application should have the ordinary meanings understood by those skilled in the art to which the present invention belongs.

[0052] As Figure 1 shown, this application provides an auxiliary troubleshooting method for air traffic control equipment. The auxiliary troubleshooting method for air traffic control equipment includes:

[0053] Step S10, obtain fault information, and obtain processing suggestions from the knowledge base according to the fault information; the fault information is mainly the specific manifestation content of the fault of the air traffic control equipment. A large number of processing suggestions are stored in the knowledge base, and the processing suggestions recorded in the knowledge base can be listed through the fault information.

[0054] Step S20, extract feature quantities from the fault information and the processing suggestions respectively, obtain a fault text representation vector from the fault information, and obtain a suggestion text representation vector from the processing suggestions; in mathematics, a vector (also known as an Euclidean vector, geometric vector) refers to a quantity with magnitude and direction. It can be visually represented as a line segment with an arrow. The direction pointed by the arrow represents the direction of the vector; the length of the line segment represents the magnitude of the vector.

[0055] In this application, the feature quantity is data for measuring and describing a certain text object. These data can be numerical values, words or other forms of information, and are used to reveal the essential features or attributes of the text. The feature quantity can also be understood as obtaining key information from the text.

[0056] Step S30, compare the fault text representation vector and the suggestion text representation vector to obtain the semantic similarity between the two; if the suggestion text representation vector can solve the fault corresponding to the fault text representation vector, the similarity between the two is close.

[0057] Step S40, form a processing decision suggestion according to the semantic similarity, and perform fault troubleshooting on the air traffic control equipment according to the processing decision suggestion. The closer the semantic similarity is, the more likely it is to solve the fault. In this application, the processing decision suggestion may include multiple processing suggestions, and maintenance personnel can select according to the magnitude of the semantic similarity.

[0058] In the technical solution of this embodiment, when troubleshooting, processing suggestions are extracted from the knowledge base based on the fault information. Feature quantities are extracted from the fault information to obtain a fault text representation vector, and feature quantities are extracted from the processing suggestions to obtain a suggestion text representation vector. Then, the fault text representation vector and the suggestion text representation vector are compared to obtain the semantic similarity between the two, and relevant processing suggestions are obtained based on the semantic similarity to form a processing decision suggestion. The maintenance personnel perform troubleshooting on the air traffic control equipment according to the processing decision suggestion. It can be seen that the technical solution of this application can automatically provide processing decision suggestions based on the fault information, thereby improving the fault handling efficiency and quickly completing the maintenance of the air traffic control equipment.

[0059] In this application, the fault processing suggestions most similar to the fault to be processed are quickly retrieved and pushed from the knowledge base. The processing suggestions may include emergency disposal procedures, which can play an auxiliary role, quickly locate the fault location, improve the fault handling time, and reduce the downtime caused by air traffic control equipment failures, which is of great significance for ensuring flight safety and improving operation and maintenance efficiency.

[0060] As Figure 2 and Figure 8 shown, in an embodiment of this application, before the step of obtaining fault information, it includes:

[0061] Step S01, construct a knowledge base, which includes an expert database, a typical case database, and an experience database.

[0062] As Figure 9 shown, among them, the expert database is the equipment fault solution provided by the equipment manufacturer or operation and maintenance experts. The solution can be entered in the expert database interface, including the unit name of the operation and maintenance expert, the equipment system name, the processing procedure, the fault phenomenon, and the fault code.

[0063] As Figure 10 shown, the typical case database is the solution formed by experienced operation and maintenance experts through typical case analysis. The case name, the unit of the operation and maintenance expert, the equipment system name, the fault time (start date and end date), the disposal process, the fault phenomenon, the fault cause, and the fault impact can be entered in the typical case database. And an addition option is also set to add content as needed.

[0064] As Figure 11 shown, the experience database is the solution formed by front-line operation and maintenance personnel based on actual operation and maintenance experience. The unit name of the operation and maintenance personnel, the equipment system name, the fault occurrence time, the processing procedure, the fault location, the fault phenomenon, the fault cause, and the fault code are included in the experience database.

[0065] As Figure 3As shown, in an embodiment of the present application, the steps of forming a processing decision suggestion according to semantic similarity include:

[0066] Step S410, perform a descending order sorting according to the magnitude of the semantic similarity; by means of the descending order sorting, it can be arranged in sequence from high to low according to the magnitude of the semantic similarity.

[0067] Step S420, obtain the processing suggestions corresponding to the top N semantic similarities to form a processing decision suggestion, where N is an integer greater than or equal to 1. The processing suggestions with a more forward arrangement position can generally solve the faults more pertinently. For example, if N is equal to 3, then the first 3 processing suggestions can be selected, and the maintenance personnel can perform maintenance according to these three processing suggestions. N can also be equal to 5, or equal to 10, etc. The value of N can be adjusted according to needs, such as automatically selecting the number of corresponding processing suggestions to be displayed according to the amount of displayed content.

[0068] As Figure 4 shown, in an embodiment of the present application, the fault information includes the device name; the air traffic control device is a set of complex management systems, usually including multiple supporting devices. After a fault occurs in the air traffic control device, the scope can be narrowed down according to the specific faulty device name. The steps of obtaining processing suggestions from the knowledge base based on the fault information include:

[0069] Step S110, using the device name as an index, screen out the processing suggestions from the knowledge base based on the fault phenomenon and / or fault code, and the fault information also includes the fault phenomenon and / or fault code. The possible faults can be quickly retrieved through the device name, and then the processing suggestions can be further screened out in combination with the fault phenomenon or the fault code. It can also be a way of combining the fault phenomenon and the fault code to jointly determine the processing suggestions.

[0070] As Figure 5 shown, in an embodiment of the present application, the steps of respectively extracting feature quantities from the fault information and the processing suggestions include:

[0071] Step S210, use a pre-trained word segmenter to perform word segmentation processing on the fault information and the processing suggestions; the word segmenter can decompose a sentence into smaller units. For example, a complete sentence can be segmented into a combination of words, and through the word segmenter, it can be divided into nouns, verbs, adjectives, etc. Further, the words after word segmentation can be converted into a digital ID sequence, and the digital ID sequences of each word can be combined to obtain the digital ID sequence of this sentence. [CLS] and [SEP] tags can also be added, aiming to divide the entire text into a combination of several complete sentences. Operations such as filling or truncating the text information can also be performed, thereby generating a more representative digital ID sequence.

[0072] Step S220: Use a pre-trained language model to extract features from the tokenized fault information and handling suggestions respectively. The pre-trained language model can be embedded in the tokenizer. The pre-trained language model has been pre-trained and can quickly extract feature quantities from the fault information and handling suggestions. In this embodiment, the pre-trained language model can be BERT (Bidirectional Encoder Representations from Transformers). BERT can understand the context information of the language through pre-training of deep bidirectional representations.

[0073] As Figure 6 shown, in an embodiment of the present application, the step of comparing the fault text representation vector and the suggestion text representation vector to obtain the semantic similarity between the two includes:

[0074] Step S310: Calculate the cosine similarity between the fault text representation vector and the suggestion text representation vector one by one to obtain the semantic similarity between the two. It can be seen from this that the semantic similarity in this embodiment refers to the cosine similarity. Generally speaking, the range of the cosine similarity is [1, -1], that is, between 1 and -1. When the cosine similarity is 1, it means the two are the same. When the cosine similarity is -1, it means the two are opposite.

[0075] In an embodiment of the present application, define the fault text representation vector as V query , the suggestion text representation vector n is the number of relevant troubleshooting solution plans corresponding in the knowledge base, then it satisfies:

[0076]

[0077] w n is the knowledge weight of the knowledge base to which the nth relevant handling suggestion belongs, is the text semantic similarity, is the semantic similarity obtained by weighting the text semantic similarity. The weighted semantic similarity calculated can more accurately reflect the similarity between the fault text representation vector and the suggestion text representation vector, improving the credibility and reliability of the data.

[0078] Furthermore, the knowledge weight of the expert database is W 1 , the knowledge weight of the troubleshooting solution plan in the typical case database is W 2 , the knowledge weight of the troubleshooting solution plan in the experience database is W 3 , then it satisfies: W 1 > W 2 > W 3It can be seen from this that the weight of the expert database is the largest, followed by the weight of the typical case database, and finally the weight of the experience database. In this way, when calculating the semantic similarity, the coefficient of the knowledge weight can be multiplied on the basis of the text semantic similarity. For example, if this processing suggestion comes from the expert database, then multiply by W on the basis of the text semantic similarity 1 , so that the ranking from the expert database is more forward. If this processing suggestion comes from the experience database, then multiply by W 3 on the basis of the text semantic similarity, so that the ranking from the experience database is relatively backward.

[0079] Through this knowledge weight allocation method, the speed of troubleshooting can be improved and more reliable processing suggestions can be provided. For example, W 1 can be 50%, W 2 can be 30%, W 3 can be 20%. W 1 can also be 55%, W 2 can be 25%, W 3 can be 20%; or, W 1 is 45%, W 2 is 30%, W 3 is 25%.

[0080] As Figure 12 shown, the processing suggestion can include content such as fault phenomenon, fault cause, processing flow, emergency flow, weighted similarity, etc. An AI suggestion option can also be added to provide processing suggestions through artificial intelligence.

[0081] As Figure 7 shown, in an embodiment of the present application, the steps of forming a processing decision suggestion according to the semantic similarity include:

[0082] Step S401, comparing the semantic similarity with a preset standard similarity, and combining the corresponding processing suggestions that are greater than or equal to the standard similarity to form a processing decision suggestion. The standard similarity can be adjusted and set as needed, and the semantic similarity obtained through calculation.

[0083] As Figure 13 shown, the present application also provides an operation and maintenance terminal 10. The operation and maintenance terminal 10 includes: a data entry module 100, an extraction module 200, a comparison module 300, and an auxiliary decision-making module 400. The operation and maintenance terminal 10 may also include a display screen to visually display the processing decision suggestion on the display screen. The display screen can be a touch screen and also serve the function of data entry.

[0084] The data entry module 100 is used to obtain fault information and obtain processing suggestions from the knowledge base according to the fault information;

[0085] The extraction module 200 is used to extract feature quantities from the fault information and the handling suggestions respectively, obtain the fault text representation vector from the fault information, and obtain the suggestion text representation vector from the handling suggestions;

[0086] The comparison module 300 is used to compare the fault text representation vector and the suggestion text representation vector to obtain the semantic similarity between the two;

[0087] The auxiliary decision-making module 400 is used to sort according to the magnitude of the semantic similarity to form a handling decision suggestion, and troubleshoot the air traffic control equipment according to the handling decision suggestion.

[0088] In the technical solution of this embodiment, when troubleshooting, the fault information is input through the data entry module 100, and the handling suggestion is extracted from the knowledge base through the fault information. The extraction module 200 extracts the feature quantity of the fault information to obtain the fault text representation vector, extracts the feature quantity of the handling suggestion to obtain the suggestion text representation vector. Then, the comparison module 300 compares the fault text representation vector and the suggestion text representation vector to obtain the semantic similarity between the two. The auxiliary decision-making module 400 obtains the relevant handling suggestion according to the semantic similarity to form a handling decision suggestion. The maintenance personnel troubleshoot the air traffic control equipment according to the handling decision suggestion. It can be seen that the technical solution of this application can automatically provide handling decision suggestions based on the fault information, thereby improving the fault handling efficiency and quickly completing the maintenance of the air traffic control equipment.

[0089] For the specific embodiments and beneficial effects of the operation and maintenance terminal in this application, please refer to the above-mentioned auxiliary fault troubleshooting method for air traffic control equipment, which will not be elaborated here.

[0090] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that: they can still modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements on some or all of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the scope of the technical solutions of the embodiments of the present invention, and they should all be covered within the scope of the claims and the description of the present invention.

Claims

1. An auxiliary troubleshooting method for air traffic control equipment, characterized in that: The auxiliary troubleshooting method of the air traffic control equipment includes: Obtaining fault information, and obtaining processing suggestions from a knowledge base based on the fault information; Extracting features from the fault information and the processing suggestion respectively, obtaining a fault text representation vector from the fault information, and obtaining a suggestion text representation vector from the processing suggestion; Comparing the fault text representation vector with the suggestion text representation vector to obtain the semantic similarity between the two; A processing decision suggestion is formed according to the semantic similarity, and air traffic control equipment is troubleshooted according to the processing decision suggestion.

2. The auxiliary troubleshooting method for air traffic control equipment according to claim 1, characterized in that: Before the step of obtaining fault information, the method includes: A knowledge base is constructed, which includes an expert database, a typical case database and an experience database.

3. The auxiliary troubleshooting method for air traffic control equipment according to claim 1, characterized in that: The step of forming a processing decision suggestion according to the semantic similarity comprises: Sorting in descending order according to the magnitude of the semantic similarity; The processing suggestions corresponding to the first N semantic similarities are obtained to form a processing decision suggestion, where N is an integer greater than or equal to 1.

4. The auxiliary troubleshooting method for air traffic control equipment according to claim 1, characterized in that: The fault information includes a device name; The step of obtaining processing suggestions from a knowledge base based on the fault information includes: Taking the device name as an index, processing suggestions are screened out from the knowledge base based on the fault phenomenon and / or the fault code, and the fault information also includes the fault phenomenon and / or the fault code.

5. The auxiliary troubleshooting method for air traffic control equipment according to claim 1, characterized in that: The step of extracting feature quantities from the fault information and the processing suggestion respectively comprises: Using a pre-trained word segmenter to perform word segmentation processing on the fault information and the processing suggestion; Use the pre-trained language model to extract features from the fault information and processing suggestions after word segmentation.

6. The auxiliary troubleshooting method for air traffic control equipment according to claim 2, characterized in that: The step of comparing the fault text representation vector and the suggestion text representation vector to obtain the semantic similarity between the two includes: The cosine similarity between the fault text representation vector and the suggestion text representation vector is calculated one by one to obtain the semantic similarity between the two.

7. The auxiliary troubleshooting method for air traffic control equipment according to claim 6, characterized in that: Define the fault text representation vector as V query , the proposed text representation vector n is the number of corresponding recommended troubleshooting solutions in the knowledge base, then: w n is the knowledge weight of the knowledge base to which the nth related processing suggestion belongs, is the text semantic similarity, The semantic similarity is obtained by weighting the semantic similarity of the text.

8. The auxiliary troubleshooting method for air traffic control equipment according to claim 7, characterized in that: The knowledge weight of the expert database is W1, the knowledge weight of the troubleshooting solutions of the typical case database is W2, and the knowledge weight of the troubleshooting solutions of the experience database is W3, then: W1>W2>W3.

9. The auxiliary troubleshooting method for air traffic control equipment according to claim 1, characterized in that: The step of forming a processing decision suggestion according to the semantic similarity comprises: The semantic similarity is compared with a preset standard similarity, and corresponding processing suggestions that are greater than or equal to the standard similarity are combined to form a processing decision.

10. An operation and maintenance terminal, characterized in that: The operation and maintenance terminal includes: A data entry module, the data entry module is used to obtain fault information, and obtain processing suggestions from a knowledge base based on the fault information; An extraction module, the extraction module is used to extract features from the fault information and the processing suggestion respectively, obtain a fault text representation vector from the fault information, and obtain a suggestion text representation vector from the processing suggestion; A comparison module, the comparison module is used to compare the fault text representation vector and the suggestion text representation vector to obtain the semantic similarity between the two; An auxiliary decision-making module is used to sort according to the size of the semantic similarity, form a processing decision suggestion, and troubleshoot the air traffic control equipment according to the processing decision suggestion.

Images