Course management method and system based on post cooperation for civil aviation aircraft operation process

Abstract

The application discloses a course management method and system for civil aviation aircraft operation process based on post cooperation, and relates to the technical field of aviation management. The sequence of data transmission in a training post is acquired to construct post relations. A graph structure is constructed according to the post relations, and a post connection matrix is used to represent the post relations. The matching of training posts and training courses is performed by using the relations between the knowledge points of training post requirements and the knowledge points taught in the training courses. The associated training posts are found through post relation detection networks. Therefore, the relations between the training posts and the training courses and the relations between the training posts are used, so that the technical effect of more accurately allocating training courses to training posts is achieved.

Description

Technical Field

[0001] This invention relates to the field of aviation management technology, specifically to a course management method and system for civil aviation aircraft operation processes based on job collaboration. Background Technology

[0002] Civil aviation aircraft operation is a highly complex collaborative process involving multiple positions, departments, and systems. A complete flight operation (from block to block) typically encompasses multiple roles, including flight crew (pilots), cabin crew, operations control center (dispatchers), aircraft maintenance personnel, ground support personnel (loading, unloading, refueling), and air traffic control (ATC). With the continuous growth of civil aviation traffic and the popularization of the "large-scale operation" concept, airlines have increasingly higher requirements for operational efficiency and safety redundancy. Any operational error in a single position or the distortion of information transmission between positions can lead to flight delays, incorrect routes, manifest errors, or even unsafe incidents. Therefore, ensuring that personnel in each position are not only proficient in their own job skills but also have a deep understanding of the collaborative logic with other positions throughout the entire process has become crucial for improving the safety and efficiency of civil aviation operations.

[0003] Existing job-specific training results in personnel only being familiar with their own specific roles, lacking a direct understanding of the work content of upstream and downstream positions. When non-standard processes or information asymmetry occur in actual operations, communication barriers and operational gaps can easily arise between positions. Training courses specifically for civil aviation aircraft operation procedures should not only consider the corresponding knowledge that each position should learn, but also enable other positions that interact with that position to learn that knowledge, facilitating mutual understanding during interactions.

[0004] Therefore, optimization algorithms or deep learning methods are used to match the courses that should be learned for each position. Although the above-mentioned existing technologies meet the basic training needs to a certain extent, they are still not accurate enough in terms of course allocation when it comes to the core operational feature of "job collaboration".

[0005] Therefore, we need to develop a course management method that can break down job barriers and support collaborative interaction among multiple roles. Summary of the Invention

[0006] The purpose of this invention is to provide a course management method and system for civil aviation aircraft operation processes based on job collaboration, so as to solve the above-mentioned problems existing in the prior art.

[0007] In a first aspect, embodiments of the present invention provide a course management method for civil aviation aircraft operation processes based on job collaboration, including:

[0008] Obtain m training positions and n training courses at multiple time points; the training courses represent courses that need to be learned during the aircraft operation process; the training positions represent positions that require training.

[0009] Based on m training positions, obtain m position relationships; each position relationship includes the current position, the position's parent node vector, and the position's child node vector; the position's parent node vector contains the training positions that send data to the current position; the position's child node vector contains the training positions that receive data from the current position.

[0010] Based on the job relationships, a graph structure is constructed to obtain the job connectivity matrix; multiple job connectivity matrices are obtained corresponding to multiple time points.

[0011] Based on the m training positions and n training courses, a training relationship is constructed to obtain the course association value; the course association value represents the association value between the training position and the training course.

[0012] By using a job relationship detection network and based on multiple job connectivity matrices, a job association vector is obtained; the job association vector represents the degree of association between multiple training jobs.

[0013] Based on the job association vector and course association value, m sets of course jobs are obtained; the course job set represents multiple training jobs that need to be attended for one training course.

[0014] Optionally, the step of constructing training relationships based on the m training positions and n training courses to obtain course association values ​​includes:

[0015] Based on the training positions, obtain a set of keywords for each training position; for each m training positions, obtain a set of m keywords for each training position.

[0016] Based on the training courses, keywords are extracted to obtain a set of training course keywords; n training courses correspond to n sets of training course keywords.

[0017] Based on a set of m training job keywords and a set of n training course keywords, keyword matching is performed to obtain m*n course association values; the course association values ​​represent the degree of matching between the knowledge required for the training job and the knowledge taught in the training course.

[0018] Optionally, the step of obtaining a job association vector through a job relationship detection network based on multiple job connectivity matrices includes:

[0019] The job relationship detection network includes a first feature extraction network, a graph convolutional network, and a temporal convolutional network;

[0020] Based on the job connectivity matrix at multiple time points, feature extraction is performed to obtain multiple first job association vectors; the first job association vectors represent the relationship between training jobs.

[0021] Based on the job connectivity matrix at multiple time points, feature extraction is performed to obtain the second job association vector; the second job association vector represents the degree of interaction between different training jobs over time.

[0022] Based on the second job association vector and multiple first job association vectors, feature fusion is performed to obtain the job association vector; the job association vector represents the relationship between multiple training jobs at multiple time points.

[0023] Optionally, feature extraction is performed based on the job connectivity matrix at multiple time points to obtain multiple first job association vectors, including:

[0024] The values ​​in the set of training job keywords are input into the first feature extraction network to obtain the first feature vector; the first feature vector represents the features of the points in the graph structure corresponding to the job connectivity matrix.

[0025] The job connectivity matrix and the first feature vector are input into a graph convolutional network to obtain the first job association vector;

[0026] Multiple time points correspond to multiple primary job position association vectors.

[0027] Optionally, the feature extraction based on the job connectivity matrix at multiple time points to obtain the second job association vector includes:

[0028] The values ​​of multiple columns in the same row of the job connectivity matrix are summed to obtain the job interaction vector; the job interaction vector represents the number of other training positions that a training position interacts with at a given time point.

[0029] Multiple job interaction vectors are obtained corresponding to multiple time points;

[0030] The job interaction vectors at multiple time points are input into a temporal convolutional network to detect changes in the degree of job interaction, thereby obtaining a second job association vector.

[0031] Optionally, the step of obtaining a set of m course positions based on the job position association vector and the course association value includes:

[0032] If the course association value is greater than the matching threshold, the training positions corresponding to the training courses will be used to construct a first course position set; the first course position set contains multiple training positions corresponding to the training courses; one training course corresponds to one first course position set;

[0033] The job association vector is input into a discriminant network to detect the degree of association between training jobs, resulting in a job association set; the job association set contains m*(m-1) / 2 job association values; the job association value represents the association between two training jobs;

[0034] Obtain m sets of associated training positions; one training position corresponds to one set of associated training positions.

[0035] If the job association value is greater than the job association threshold, the two training jobs corresponding to the job association value will be designated as the first training job and the second training job.

[0036] Add the second training position to the set of associated training positions corresponding to the first training position;

[0037] Add the first training position to the set of associated training positions corresponding to the second training position;

[0038] To obtain the course job set, add elements from the associated training job set corresponding to the training job in the first course job set;

[0039] For each m training position, a set of m course positions is obtained.

[0040] Optionally, the step of performing keyword matching based on m sets of training job keywords and n sets of training course keywords to obtain m*n course association values ​​includes:

[0041] Obtain p training job keywords and q training course keywords; the training job keywords are elements in the training job keyword set; the training course keywords are elements in the training course keyword set.

[0042] Based on the keywords of the training positions and the key points of the training courses, the first course association value is obtained through a similarity measurement algorithm;

[0043] For each p training job keyword and q training course keyword, p*q first course association values ​​are obtained.

[0044] Based on p*q first course association values, m*n course association values ​​are obtained.

[0045] Optionally, obtaining m*n course association values ​​based on p*q first course association values ​​includes:

[0046] Multiple first course association values ​​that exceed the course association threshold are used as second course association values; the second course association value represents the degree of association between the matched training position keywords and training course keywords;

[0047] If the quotient of the number of the second course association values ​​divided by the number of the first course association values ​​is greater than the number threshold, the average of the multiple second course association values ​​is calculated to obtain the course association value.

[0048] One set of keywords for training positions and one set of keywords for training courses correspond to one course association value;

[0049] Given m sets of keywords for training positions and n sets of keywords for training courses, we obtain m*n course association values.

[0050] Optionally, the step of constructing a graph structure based on the job relationships to obtain a job connectivity matrix includes:

[0051] Based on the training positions, construct a position connectivity matrix; the position connectivity matrix represents a connectivity matrix with training positions as indices;

[0052] The initial values ​​of the job connectivity matrix are all 0;

[0053] In the job connectivity matrix, the value of the position corresponding to the job relationship is set to 1; the job connectivity matrix represents the interactive training jobs.

[0054] Secondly, embodiments of the present invention provide a course management system for civil aviation aircraft operation processes based on job collaboration, including:

[0055] The acquisition module is used to acquire m training positions and n training courses at multiple time points; the training courses represent the courses that need to be learned during the aircraft operation process; the training positions represent the positions that need to be trained.

[0056] The job relationship module is used to obtain m job relationships based on m training jobs. The job relationship includes the current job, the job parent node vector, and the job child node vector. The job parent node vector contains the training jobs that send data to the current job. The job child node vector contains the training jobs that receive data from the current job.

[0057] The graph structure module is used to construct a graph structure based on the job relationships to obtain a job connectivity matrix; multiple job connectivity matrices are obtained corresponding to multiple time points.

[0058] The course association module is used to construct training relationships based on the m training positions and n training courses, and obtain course association values; the course association values ​​represent the association values ​​between training positions and training courses.

[0059] The job association module 1 is used to obtain a job association vector based on multiple job connectivity matrices through a job relationship detection network; the job association vector represents the degree of association between multiple training jobs.

[0060] The discrimination module is used to obtain m sets of course positions based on the job association vector and course association value; the course position set represents multiple training positions that need to be attended for one training course.

[0061] Compared with the prior art, the embodiments of the present invention achieve the following beneficial effects:

[0062] This invention also provides a course management method and system for civil aviation aircraft operation processes based on job collaboration.

[0063] In this invention, the order of data transmission in training positions is used to construct position relationships. A graph structure is then built based on these relationships, using a position connectivity matrix to represent them. The relationships between the knowledge points required for each training position and those taught in the training courses are used to match training positions with training courses. Through a position relationship detection network, positions prone to errors in aircraft operation procedures are identified as associated training positions. This approach, utilizing the relationships between training positions and training courses, and the relationships between training positions themselves, achieves a more accurate technical effect in allocating training courses to training positions. Attached Figure Description

[0064] Figure 1 This is a flowchart of a course management method for civil aviation aircraft operation processes based on job collaboration, provided in an embodiment of the present invention. Detailed Implementation

[0065] The present invention will now be described in detail with reference to the accompanying drawings.

[0066] This invention provides a course management method and system for civil aviation aircraft operation processes based on job collaboration, which allocates courses to personnel in multiple training positions.

[0067] Example 1:

[0068] like Figure 1 As shown, this embodiment of the invention provides a course management method for civil aviation aircraft operation processes based on job collaboration, the method comprising:

[0069] S101: Obtain m training positions and n training courses at multiple time points; the training courses represent courses that need to be learned during the aircraft operation process; the training positions represent positions that need to be trained.

[0070] Where m and n are positive integers.

[0071] S102: Based on m training positions, obtain m position relationships; the position relationship includes the current position, the position parent node vector, and the position child node vector; the position parent node vector contains the training positions that send data to the current position; the position child node vector contains the training positions that receive data from the current position.

[0072] The current position refers to the training position currently being judged.

[0073] One training position corresponds to one job relationship.

[0074] The job relationship refers to whether multiple training jobs interact at a single point in time.

[0075] S103: Based on the job relationships, construct a graph structure to obtain a job connectivity matrix; obtain multiple job connectivity matrices corresponding to multiple time points.

[0076] S104: Based on the m training positions and n training courses, construct training relationships to obtain course association values; the course association values ​​represent the association values ​​between training positions and training courses.

[0077] S105: Using a job relationship detection network, based on multiple job connectivity matrices, a job association vector is obtained; the job association vector represents the degree of association between multiple training jobs.

[0078] S106: Based on the job association vector and course association value, obtain m sets of course jobs; the set of course jobs represents the multiple training jobs that need to be attended for one training course.

[0079] Optionally, the step of constructing training relationships based on the m training positions and n training courses to obtain course association values ​​includes:

[0080] Based on the training positions, obtain a set of keywords for each training position; for each m training positions, obtain a set of m keywords for each training position.

[0081] The elements in the training job keyword set represent keywords related to the knowledge that the training job requires to learn.

[0082] Based on the training courses, keywords are extracted to obtain a set of training course keywords; m training courses correspond to m sets of training course keywords.

[0083] In this process, a one-dimensional convolutional neural network (1D CNN) is used for keyword extraction. This 1D CNN is a network trained using labeled data. The labeled data represents keywords related to the training course.

[0084] Specifically, statements describing the course content in the training course are input into a one-dimensional convolutional neural network (1D CNN) for keyword extraction, resulting in a set of training course keywords.

[0085] Based on a set of m training job keywords and a set of n training course keywords, keyword matching is performed to obtain m*n course association values; the course association values ​​represent the degree of matching between the knowledge required for the training job and the knowledge taught in the training course.

[0086] Optionally, the step of obtaining a job association vector through a job relationship detection network based on multiple job connectivity matrices includes:

[0087] The job relationship detection network includes a first feature extraction network, a graph convolutional network, and a temporal convolutional network.

[0088] Based on the job connectivity matrix at multiple time points, feature extraction is performed to obtain multiple first job association vectors; the first job association vectors represent the relationship of knowledge data that should be mastered between the associated training jobs.

[0089] Based on the job connectivity matrix at multiple time points, feature extraction is performed to obtain a second job association vector; the second job association vector represents the degree of interaction between different training jobs over time.

[0090] Based on the second job association vector and multiple first job association vectors, feature fusion is performed to obtain the first job association vector.

[0091] In this embodiment, the second job-related vector and multiple first job-related vectors are input into a fusion neural network to obtain the job-related vector.

[0092] The fusion neural network is a fully connected neural network (FCN).

[0093] Optionally, feature extraction is performed based on the job connectivity matrix at multiple time points to obtain multiple first job association vectors, including:

[0094] The values ​​in the set of training job keywords are input into the first feature extraction network to obtain the first feature vector; the first feature vector represents the features of the points in the graph structure corresponding to the job connectivity matrix.

[0095] The first feature extraction network is a fully connected neural network (FCN).

[0096] The job connectivity matrix and training data are input into a graph convolutional network (GCN) to obtain a first job association vector; the first job association vector represents the relationship between the knowledge data that should be mastered between the associated training jobs.

[0097] Multiple time points correspond to multiple primary job position association vectors.

[0098] Optionally, obtaining the first job association vector based on the training positions and the job connectivity matrix includes:

[0099] The training data is input into a first feature extraction network to obtain a first feature vector; the first feature vector represents the features of the points in the graph structure corresponding to the job connectivity matrix.

[0100] The first feature extraction network is a fully connected neural network (FCN).

[0101] The job connectivity matrix and the first feature vector are input into a graph convolutional network (GCN) to obtain a first job association vector. The first job association vector represents the relationship between the knowledge data that should be mastered between the associated training jobs.

[0102] Multiple time points correspond to multiple primary job position association vectors.

[0103] The above method is used because the interaction between training positions is different at each point in time.

[0104] Optionally, the feature extraction based on the job connectivity matrix at multiple time points to obtain the second job association vector includes:

[0105] The values ​​of multiple columns in the same row of the job connectivity matrix are summed to obtain the job interaction vector; the job interaction vector represents the number of other training positions that a training position interacts with at a given time point.

[0106] Multiple job interaction vectors are obtained corresponding to multiple time points;

[0107] The job interaction vectors at multiple time points are input into a temporal convolutional network (TCN) to detect changes in the degree of job interaction, thereby obtaining a second job association vector.

[0108] Optionally, the step of obtaining a set of m course positions based on the job position association vector and the course association value includes:

[0109] If the course association value is greater than the matching threshold, the training positions corresponding to the training courses will be used to construct a first course position set; the first course position set contains multiple training positions corresponding to the training courses; one training course corresponds to one first course position set.

[0110] In this embodiment, the matching threshold is 0.8.

[0111] If the course association value is greater than the matching threshold, the multiple training positions corresponding to one training course will be combined into the first course position set.

[0112] The job association vector is input into a discriminant network to detect the degree of association between training jobs, resulting in a job association set; the job association set contains m*(m-1) / 2 job association values; the job association value represents the association between two training jobs.

[0113] The discriminant network is a fully connected neural network (FCN).

[0114] The discriminant network contains m*(m-1) / 2 output neurons.

[0115] During training, the status of training positions in historical aircraft operation procedures is used for identification. If an error occurs in the aircraft operation procedure, the pairwise job association value between the erroneous training positions is set to 1, while the pairwise job association value between other training positions is set to 0. This data is used as annotation data for training. A job association value of 1 indicates that the training positions are associated. A job association value of 0 indicates that the training positions are not associated.

[0116] Get m sets of associated training positions; one training position corresponds to one set of associated training positions.

[0117] The set of associated training positions is initially an empty set.

[0118] If the job association value is greater than the job association threshold, the two training jobs corresponding to the job association value will be designated as the first training job and the second training job.

[0119] In this embodiment, the job-related threshold is set to 0.8.

[0120] Add the second training position to the set of associated training positions corresponding to the first training position.

[0121] Add the first training position to the set of associated training positions corresponding to the second training position.

[0122] To obtain the course job set, add elements from the associated training job set corresponding to the training job in the first course job set.

[0123] Using the above method, one training position associated with another training position can be added to the same set of first course positions.

[0124] For each m training position, a set of m course positions is obtained.

[0125] Optionally, the step of obtaining a set of m course positions based on the job position association vector and the course association value includes:

[0126] If the matching value is greater than the matching threshold, the training courses are matched with the training positions to obtain multiple sets of first course positions; the first set of first course positions includes multiple training positions matched with the knowledge-matched training courses.

[0127] In this embodiment, the matching threshold is 0.8.

[0128] The job association vector is input into a discriminant network to detect the degree of association between training jobs, resulting in a job association set; the job association set contains m*(m-1) / 2 job association values; the job association value represents the association between two training jobs;

[0129] If the job association value is greater than the job association threshold, m sets of associated training jobs are obtained; one set of associated training jobs corresponds to one training job.

[0130] To obtain the course job set, add elements from the associated training job set corresponding to the training job in the first course job set;

[0131] For each m training position, a set of m course positions is obtained.

[0132] Optionally, the step of performing keyword matching based on m sets of training job keywords and n sets of training course keywords to obtain m*n course association values ​​includes:

[0133] Obtain p training job keywords and q training course keywords; the training job keywords are elements in the training job keyword set; the training course keywords are elements in the training course keyword set.

[0134] Where p and q are positive integers.

[0135] Based on the keywords of the training positions and the key points of the training courses, a first course association value is obtained through a similarity measurement algorithm.

[0136] In this embodiment, the similarity measurement algorithm is the Euclidean distance algorithm.

[0137] For each p training job keyword and q training course keyword, p*q first course association values ​​are obtained.

[0138] Based on p*q first course association values, m*n course association values ​​are obtained.

[0139] Optionally, obtaining m*n course association values ​​based on p*q first course association values ​​includes:

[0140] Multiple first course association values ​​that exceed the course association threshold are used as second course association values; the second course association value represents the degree of association between the matched training job keywords and training course keywords.

[0141] In this embodiment, the course association threshold is 0.8.

[0142] If the quotient of the number of second course association values ​​divided by the number of first course association values ​​is greater than the quantity threshold, the average of the multiple second course association values ​​is calculated to obtain the course association value.

[0143] In this embodiment, the quantity threshold is 0.85.

[0144] Using the above method, the course association value represents the degree of matching between the knowledge required for the training position and the knowledge taught in the training course.

[0145] One set of keywords for training positions and one set of keywords for training courses correspond to one course association value;

[0146] Given m sets of keywords for training positions and n sets of keywords for training courses, we obtain m*n course association values.

[0147] Optionally, the step of constructing a graph structure based on the job relationships to obtain a job connectivity matrix includes:

[0148] Based on the training positions, construct a position connectivity matrix; the position connectivity matrix represents a connectivity matrix with training positions as subscripts.

[0149] In this context, the subscripts corresponding to the same rows and columns in the job connectivity matrix represent the same training job.

[0150] The initial value of the job connectivity matrix is ​​all 0.

[0151] In the job connectivity matrix, the value of the position corresponding to the job relationship is set to 1; the job connectivity matrix represents the interactive training jobs.

[0152] Specifically, in the job connectivity matrix, the training matrix corresponding to column index 1 and row index 1 is the first training job; the training matrix corresponding to column index 1 and row index 1 is the second training job. For example, in the job child node vector of the first training job in the second training job, the position corresponding to column index 1 and row index 2 is set to 1. And in the job parent node vector of the second training job in the first training job, the position corresponding to column index 2 and row index 1 is set to 1.

[0153] The first training position and the second training position are different training positions.

[0154] Example 2:

[0155] Based on the above-described course management method for civil aviation aircraft operation processes based on job collaboration, this embodiment of the invention also provides a course management system for civil aviation aircraft operation processes based on job collaboration, the system comprising:

[0156] The acquisition module is used to acquire m training positions and n training courses at multiple time points; the training courses represent the courses that need to be learned during the aircraft operation process; the training positions represent the positions that need to be trained.

[0157] The job relationship module is used to obtain m job relationships based on m training jobs. The job relationship includes the current job, the job parent node vector, and the job child node vector. The job parent node vector contains the training jobs that send data to the current job. The job child node vector contains the training jobs that receive data from the current job.

[0158] The graph structure module is used to construct a graph structure based on the job relationships to obtain a job connectivity matrix; multiple job connectivity matrices are obtained corresponding to multiple time points.

[0159] The course association module is used to construct training relationships based on the m training positions and n training courses, and obtain course association values; the course association values ​​represent the association values ​​between training positions and training courses.

[0160] The job association module 1 is used to obtain a job association vector based on multiple job connectivity matrices through a job relationship detection network; the job association vector represents the degree of association between multiple training jobs.

[0161] The discrimination module is used to obtain m sets of course positions based on the job position association vector and the course association value; the set of course positions represents the multiple training positions that need to be attended for one training course.

[0162] The algorithms and displays provided herein are not inherently related to any particular computer, virtual system, or other device. Various general-purpose systems can also be used in conjunction with the teachings herein. The required structure for constructing such systems is apparent from the above description. Furthermore, this invention is not directed to any particular programming language. It should be understood that the contents of the invention described herein can be implemented using various programming languages, and the above description of specific languages ​​is for the purpose of disclosing the best mode of implementation of the invention.

[0163] Numerous specific details are set forth in the specification provided herein. However, it will be understood that embodiments of the invention may be practiced without these specific details. In some instances, well-known methods, structures, and techniques have not been shown in detail so as not to obscure the understanding of this specification.

[0164] The various component embodiments of the present invention can be implemented in hardware, or as software modules running on one or more processors, or a combination thereof. Those skilled in the art will understand that microprocessors or digital signal processors (DSPs) can be used in practice to implement some or all of the functions of some or all of the components in the apparatus according to embodiments of the present invention. The present invention can also be implemented as a device or apparatus program (e.g., a computer program and computer program product) for performing part or all of the methods described herein. Such programs implementing the present invention can be stored on a computer-readable medium or can be in the form of one or more signals. Such signals can be downloaded from an Internet website, provided on a carrier signal, or provided in any other form.

Claims

1. A course management method for civil aviation aircraft operation processes based on job collaboration, characterized in that, include: Obtain m training positions and n training courses at multiple time points; The training courses refer to the courses that need to be learned during the aircraft operation process; The training positions refer to those positions that require training; Based on m training positions, obtain m position relationships; each position relationship includes the current position, the position's parent node vector, and the position's child node vector; the position's parent node vector contains the training positions that send data to the current position; the position's child node vector contains the training positions that receive data from the current position. Based on the job relationships, a graph structure is constructed to obtain the job connectivity matrix; Multiple job connectivity matrices are obtained corresponding to multiple time points; Based on the m training positions and n training courses, a training relationship is constructed to obtain the course association value; the course association value represents the association value between the training position and the training course. By using a job relationship detection network and based on multiple job connectivity matrices, a job association vector is obtained; the job association vector represents the degree of association between multiple training jobs. Based on the job association vector and course association value, m sets of course jobs are obtained; the course job set represents multiple training jobs that need to be attended for one training course.

2. The course management method for civil aviation aircraft operation processes based on job collaboration as described in claim 1, characterized in that, The process of constructing training relationships based on the m training positions and n training courses to obtain course association values ​​includes: Based on the training positions, obtain a set of keywords for each training position; for each m training positions, obtain a set of m keywords for each training position. Based on the training courses, keywords are extracted to obtain a set of training course keywords; n training courses correspond to n sets of training course keywords. Based on a set of m training job keywords and a set of n training course keywords, keyword matching is performed to obtain m*n course association values; the course association values ​​represent the degree of matching between the knowledge required for the training job and the knowledge taught in the training course.

3. The course management method for civil aviation aircraft operation processes based on job collaboration as described in claim 2, characterized in that, The method of obtaining job association vectors through a job relationship detection network based on multiple job connectivity matrices includes: The job relationship detection network includes a first feature extraction network, a graph convolutional network, and a temporal convolutional network; Based on the job connectivity matrix at multiple time points, feature extraction is performed to obtain multiple first job association vectors; the first job association vectors represent the relationship between training jobs. Based on the job connectivity matrix at multiple time points, feature extraction is performed to obtain the second job association vector; the second job association vector represents the degree of interaction between different training jobs over time. Based on the second job association vector and multiple first job association vectors, feature fusion is performed to obtain the job association vector; the job association vector represents the relationship between multiple training jobs at multiple time points.

4. The course management method for civil aviation aircraft operation processes based on job collaboration as described in claim 3, characterized in that, The job connectivity matrix based on multiple time points is used for feature extraction to obtain multiple first job association vectors, including: The values ​​in the set of training job keywords are input into the first feature extraction network to obtain the first feature vector; the first feature vector represents the features of the points in the graph structure corresponding to the job connectivity matrix. The job connectivity matrix and the first feature vector are input into a graph convolutional network to obtain the first job association vector; Multiple time points correspond to multiple primary job position association vectors.

5. The course management method for civil aviation aircraft operation processes based on job collaboration as described in claim 3, characterized in that, The second job association vector is obtained by extracting features from the job connectivity matrix based on multiple time points, including: The values ​​of multiple columns in the same row of the job connectivity matrix are summed to obtain the job interaction vector; the job interaction vector represents the number of other training positions that a training position interacts with at a given time point. Multiple job interaction vectors are obtained corresponding to multiple time points; The job interaction vectors at multiple time points are input into a temporal convolutional network to detect changes in the degree of job interaction, thereby obtaining a second job association vector.

6. The course management method for civil aviation aircraft operation processes based on job collaboration as described in claim 1, characterized in that, The process of obtaining a set of m course positions based on the job association vector and course association value includes: If the course association value is greater than the matching threshold, the training positions corresponding to the training courses will be used to construct a first course position set; the first course position set contains multiple training positions corresponding to the training courses; one training course corresponds to one first course position set; The job association vector is input into a discriminant network to detect the degree of association between training jobs, resulting in a job association set; the job association set contains m*(m-1) / 2 job association values; the job association value represents the association between two training jobs; Obtain m sets of associated training positions; one training position corresponds to one set of associated training positions. If the job association value is greater than the job association threshold, the two training jobs corresponding to the job association value will be designated as the first training job and the second training job. Add the second training position to the set of associated training positions corresponding to the first training position; Add the first training position to the set of associated training positions corresponding to the second training position; To obtain the course job set, add elements from the associated training job set corresponding to the training job in the first course job set; For each m training position, a set of m course positions is obtained.

7. The course management method for civil aviation aircraft operation processes based on job collaboration as described in claim 2, characterized in that, The process involves keyword matching based on a set of m training job keywords and a set of n training course keywords to obtain m*n course association values, including: Obtain p training job keywords and q training course keywords; the training job keywords are elements in the training job keyword set; the training course keywords are elements in the training course keyword set. Based on the keywords of the training positions and the key points of the training courses, the first course association value is obtained through a similarity measurement algorithm; For each p training job keyword and q training course keyword, p*q first course association values ​​are obtained. Based on p*q first course association values, m*n course association values ​​are obtained.

8. The course management method for civil aviation aircraft operation processes based on job collaboration as described in claim 7, characterized in that, The process of obtaining m*n course association values ​​based on p*q first course association values ​​includes: Multiple first course association values ​​that exceed the course association threshold are used as second course association values; the second course association value represents the degree of association between the matched training position keywords and training course keywords; If the quotient of the number of the second course association values ​​divided by the number of the first course association values ​​is greater than the number threshold, the average of the multiple second course association values ​​is calculated to obtain the course association value. One set of keywords for training positions and one set of keywords for training courses correspond to one course association value; Given m sets of keywords for training positions and n sets of keywords for training courses, we obtain m*n course association values.

9. The course management method for civil aviation aircraft operation processes based on job collaboration as described in claim 1, characterized in that, The process of constructing a graph structure based on the job relationships to obtain a job connectivity matrix includes: Based on the training positions, construct a position connectivity matrix; the position connectivity matrix represents a connectivity matrix with training positions as indices; The initial values ​​of the job connectivity matrix are all 0; In the job connectivity matrix, the value of the position corresponding to the job relationship is set to 1; the job connectivity matrix represents the interactive training jobs.

10. A course management system for civil aviation aircraft operation processes based on job collaboration, characterized in that, include: The acquisition module is used to acquire m training positions and n training courses at multiple time points; The training courses refer to the courses that need to be learned during the aircraft operation process; The training positions refer to those positions that require training; The job relationship module is used to obtain m job relationships based on m training jobs. The job relationship includes the current job, the job parent node vector, and the job child node vector. The job parent node vector contains the training jobs that send data to the current job. The job child node vector contains the training jobs that receive data from the current job. The graph structure module is used to construct a graph structure based on the job relationships to obtain a job connectivity matrix; Multiple job connectivity matrices are obtained corresponding to multiple time points; The course association module is used to construct training relationships based on the m training positions and n training courses, and obtain course association values; the course association values ​​represent the association values ​​between training positions and training courses. The job association module 1 is used to obtain a job association vector based on multiple job connectivity matrices through a job relationship detection network; the job association vector represents the degree of association between multiple training jobs. The discrimination module is used to obtain m sets of course positions based on the job position association vector and the course association value; the set of course positions represents the multiple training positions that need to be attended for one training course.

Images