Subway tunnel external construction type detection method, device, equipment and medium

Abstract

The invention relates to a subway tunnel external construction type detection method, device and equipment and a medium, and the method comprises the steps: obtaining historical grating vibration data, current grating vibration data and construction condition types outside a subway tunnel, and setting an initial construction type detection model; determining a type weight coefficient of each type in the construction condition types according to the historical grating vibration data and the current grating vibration data; time-frequency feature extraction and labeling classification are carried out on historical grating vibration data according to the construction condition type to obtain an initial set, so that the complexity of the data is reduced; training the initial construction type detection model according to the initial set to obtain a target construction type detection model; according to the target construction type detection model and all the type weight coefficients of the construction working condition type, type detection is carried out on the current grating vibration data, so that a more accurate external construction type is obtained, the accuracy of the external construction type is improved, and the situation of misinformation is reduced.

Description

Technical Field

[0001] The present invention relates to the technical field of subway tunnel external construction type detection, and in particular to a subway tunnel external construction type detection method, device, equipment and medium. Background Art

[0002] Drilling construction above the subway may cause ground instability, deformation or cracking of subway tunnels, derailment of trains or damage to vehicle structures, endangering the lives of passengers and the normal operation of the subway. To avoid the above problems, detailed geological surveys and risk assessments are required before construction, and advanced construction technologies and equipment are used to ensure the safety and accuracy of construction operations. During the construction process, an effective monitoring and early warning system should be set up to monitor the status of the strata, tunnels and tracks in real time, and to promptly detect and deal with potential abnormalities to prevent safety accidents. There is currently a subway tunnel external intrusion monitoring technology based on fiber grating arrays. By laying a distributed fiber optic sensor network on the inner wall of the subway tunnel, various vibration signals generated by mechanical operations above the tunnel are sensed. The fiber grating array sensing system analyzes and identifies the collected vibration signals, identifies equipment such as drilling rigs that have potential hazards to subway safety according to the vibration characteristics of different machines and equipment, and locates their specific positions on the tunnel. This full-time and full-domain monitoring technology can achieve non-contact and non-destructive monitoring with a wider monitoring range, higher accuracy and larger capacity. Once a hidden danger is found, the system can promptly alarm and take investigation and disposal measures, thereby improving the safety of the subway operation process.

[0003] At present, a method combining traditional methods and artificial intelligence is used to identify vibration signals, which can accurately determine the type of construction machinery and effectively reduce the occurrence of subway safety accidents. However, this method also has many problems, such as large amount of original vibration data, high complexity, complex data processing, multiple alarms, inaccurate alarms, and many false alarms.

[0004] Therefore, it is urgent to propose a method, device, equipment and medium for detecting the external construction type of subway tunnels to solve the technical problems existing in the prior art of inspecting subway safety accidents through vibration signals, which leads to multiple alarms, inaccurate alarms and many false alarms due to large data volume, high complexity and complex data processing. Summary of the invention

[0005] In view of this, it is necessary to provide a method, device, equipment and medium for detecting the external construction type of a subway tunnel, so as to solve the technical problems existing in the prior art of inspecting subway safety accidents through vibration signals, which lead to many alarms, inaccurate alarms and many false alarms due to large data volume, high complexity and complex data processing.

[0006] In order to solve the above problems, the present invention provides a method for detecting the external construction type of a subway tunnel, comprising: Obtain historical grating vibration data, current grating vibration data and construction condition type outside the subway tunnel, and set up an initial construction type detection model; Determining a type weight coefficient of each type of the construction condition type according to the historical grating vibration data and the current grating vibration data; Extracting time-frequency features and labeling classification of the historical grating vibration data according to the type of construction conditions to obtain an initial set; Training the initial construction type detection model according to the initial set to obtain a target construction type detection model; The current grating vibration data is subjected to type detection according to the target construction type detection model and all type weight coefficients of the construction condition type to obtain an external construction type.

[0007] In a possible implementation, the extracting time-frequency features and labeling classification of the historical grating vibration data according to the construction condition type to obtain an initial set includes: Extracting time-frequency features of the historical grating vibration data to obtain a time-frequency energy graph; The time-frequency energy diagram is labeled and classified according to the construction condition type to obtain an initial set.

[0008] In a possible implementation, extracting time-frequency features from the historical grating vibration data to obtain a time-frequency energy graph includes: Reading the historical grating vibration data to obtain an original time-domain vibration signal; Preprocessing the original time-domain vibration signal to obtain an initial time-domain vibration signal; Energy conversion is performed on the initial time-domain vibration signal to obtain a time-frequency energy diagram.

[0009] In a possible implementation, the time-frequency energy graph is labeled and classified according to the construction condition type to obtain an initial set, including: The time-frequency energy graph is segmented according to a preset sliding window to obtain a plurality of sample data of a fixed size; Labeling the plurality of sample data according to the construction condition type to obtain target sample data; The target sample data is classified to obtain an initial set.

[0010] In a possible implementation, the initial set includes a training set, a validation set, and a test set; and training the initial construction type detection model according to the initial set to obtain a target construction type detection model includes: Training and verifying the initial construction type detection model according to the training set and the verification set to obtain a training construction type detection model; Testing the training construction type detection model according to the test set to obtain a test result; When the test result meets the preset conditions, the training construction type detection model is determined to be the target construction type detection model.

[0011] In a possible implementation, when the test result satisfies a preset condition, determining the training construction type detection model as a target construction type detection model includes: Performing weight calculation on the test results to obtain a weight value; Determine whether the weight value is greater than a preset threshold; If not, the training set is updated and optimized according to the test results, and the target construction type detection model is trained again; If so, it is determined that the test result meets the preset condition.

[0012] In a possible implementation, the performing type detection on the current grating vibration data according to the target construction type detection model and all type weight coefficients of the construction condition type to obtain the external construction type includes: Performing type detection on the current grating vibration data according to the target construction type detection model to obtain an initial external construction type; Compare the initial external construction type with the construction condition type to obtain a corresponding target type weight coefficient; When the target type weight coefficient is greater than a preset weight threshold, the initial external construction type is determined to be an external construction type.

[0013] On the other hand, the present invention also provides a subway tunnel external construction type detection device, comprising: A data acquisition module is used to obtain historical grating vibration data, current grating vibration data and construction condition type outside the subway tunnel, and set an initial construction type detection model; A type calculation module, used for determining a type weight coefficient of each type of the construction condition type according to the historical grating vibration data and the current grating vibration data; A set determination module, used for extracting time-frequency features and labeling and classifying the historical grating vibration data according to the type of construction conditions to obtain an initial set; A model training module, used for training the initial construction type detection model according to the initial set to obtain a target construction type detection model; The type detection module is used to perform type detection on the current grating vibration data according to the target construction type detection model and all type weight coefficients of the construction condition type to obtain the external construction type.

[0014] On the other hand, an embodiment of the present invention discloses an electronic device, comprising: a processor, a memory, and a computer program stored in the memory and capable of running on the processor, wherein when the computer program is executed by the processor, each step of the above-mentioned subway tunnel external construction type detection method embodiment is implemented.

[0015] On the other hand, an embodiment of the present invention discloses a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the various steps of the above-mentioned subway tunnel external construction type detection method embodiment are implemented.

[0016] The beneficial effects of the present invention are as follows: historical grating vibration data, current grating vibration data and construction condition types outside a subway tunnel are obtained, and an initial construction type detection model is set; the type weight coefficient of each type in the construction condition type is determined according to the historical grating vibration data and the current grating vibration data; the historical grating vibration data is subjected to time-frequency feature extraction and annotation classification according to the construction condition type to obtain an initial set; the initial construction type detection model is trained according to the initial set to obtain a target construction type detection model; the current grating vibration data is subjected to type detection according to the target construction type detection model and all type weight coefficients of the construction condition type to obtain an external construction type; the present invention can obtain a more accurate initial set of each type through time-frequency feature extraction and annotation classification, thereby reducing the complexity of the data; the current grating vibration data can also be detected through the target construction type detection model, and the detection result can be further determined through all type weight coefficients of the construction condition type, thereby obtaining a more accurate external construction type, improving the accuracy of the external construction type, and reducing the possibility of false alarms. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 A schematic flow chart of an embodiment of a method for detecting the external construction type of a subway tunnel provided by the present invention; Figure 2 A schematic diagram of a flow chart of an embodiment of obtaining a time-frequency energy diagram provided by the present invention; Figure 3 A schematic diagram of a flow chart of an embodiment of obtaining an initial set provided by the present invention; Figure 4 For the present invention Figure 1 A schematic flow chart of an embodiment of step S104; Figure 5 A schematic structural diagram of an embodiment of a device for detecting the external construction type of a subway tunnel provided by the present invention; Figure 6 A schematic structural diagram of an embodiment of an electronic device provided by the present invention. DETAILED DESCRIPTION

[0018] The preferred embodiments of the present invention are described in detail below in conjunction with the accompanying drawings, wherein the accompanying drawings constitute a part of this application and are used together with the embodiments of the present invention to illustrate the principles of the present invention, but are not used to limit the scope of the present invention.

[0019] like Figure 1 As shown, a specific embodiment of the present invention discloses a method for detecting the external construction type of a subway tunnel, comprising: S101, obtaining historical grating vibration data, current grating vibration data and construction condition type outside the subway tunnel, and setting an initial construction type detection model; S102, determining a type weight coefficient of each type of construction condition type according to historical grating vibration data and current grating vibration data; S103, extracting time-frequency features and labeling and classifying the historical grating vibration data according to the construction condition type to obtain an initial set; S104, training the initial construction type detection model according to the initial set to obtain a target construction type detection model; S105 , performing type detection on the current grating vibration data according to the target construction type detection model and all type weight coefficients of the construction condition type to obtain the external construction type.

[0020] It should be understood that the historical grating vibration data of the subway tunnel obtained in step S101 can be the previous historical grating vibration data obtained based on the array grating sensor, or can be the historical grating vibration data stored in the storage medium.

[0021] In a specific embodiment of the present invention, an array grating sensor arranged outside a subway tunnel can be used to collect grating vibration data at different positions in real time, so that grating vibration data at different time periods can be obtained, and then historical grating vibration data can be obtained, and the data can be saved to multiple server nodes through a distributed storage system. The historical grating vibration data stored in history can also be obtained by other means. The specific acquisition method can be set according to actual conditions, and the embodiment of the present invention is not limited here; the current grating vibration data of the current time period can also be collected, and the construction condition type can also be set. The construction condition type can include normal working conditions, drilling, demolition and other types. The initial construction type detection model can be a network model that combines a multi-layer Transformer and a sliding attention mechanism module and GCN, which is used to realize end-to-end data processing and model recognition processes, and can realize automatic recognition of normal working conditions, drilling, demolition and other working conditions, thereby improving the intelligent level of subway safety operation management. The type weight coefficient of each type in the construction condition type can be calculated based on the historical grating vibration data and the current grating vibration data. Specifically: Under normal working conditions, the type weight coefficient is P x , then the type weight coefficient under normal working conditions is as shown in formula (1): (1) In the formula, is the type weight coefficient of normal working conditions in the current detection period, n Expressed as the total number of historical grating vibration data, is the current grating vibration data, For the i Historical grating vibration data corresponding to each construction condition type.

[0022] Similarly, other types of historical grating vibration data and current grating vibration data are input into the weight function of formula (1), and the type weight coefficient of the corresponding type is output. Then, the time-frequency feature extraction and annotation classification of the historical grating vibration data can be performed according to the construction condition type to obtain the initial set; then, the initial construction type detection model can be trained according to the initial set to obtain the target construction type detection model; and the current grating vibration data can also be type-detected according to the target construction type detection model and all type weight coefficients of the construction condition type to obtain the external construction type.

[0023] Compared with the prior art, the present embodiment provides the method of obtaining historical grating vibration data, current grating vibration data and construction condition types outside the subway tunnel, and setting an initial construction type detection model; determining the type weight coefficient of each type in the construction condition type according to the historical grating vibration data and the current grating vibration data; extracting time-frequency features and labeling classification of the historical grating vibration data according to the construction condition type to obtain an initial set; training the initial construction type detection model according to the initial set to obtain a target construction type detection model; performing type detection on the current grating vibration data according to the target construction type detection model and all type weight coefficients of the construction condition type to obtain the external construction type; the present invention can obtain a more accurate initial set of each type through time-frequency feature extraction and labeling classification, thereby reducing the complexity of the data; the current grating vibration data can also be detected through the target construction type detection model, and the detection result can be further determined through all type weight coefficients of the construction condition type, so that a more accurate external construction type can be obtained, the accuracy of the external construction type is improved, and the false alarm situation is reduced.

[0024] In some embodiments of the present invention, step S103 includes: Extract time-frequency features from historical grating vibration data to obtain a time-frequency energy diagram; The time-frequency energy diagrams are labeled and classified according to the construction condition types to obtain the initial set.

[0025] In a specific embodiment of the present invention, after acquiring the historical grating vibration data, the time-frequency features of the historical grating vibration data can be extracted to obtain a time-frequency-energy diagram, that is, a time-frequency energy diagram, and then the construction condition types can be marked one by one on the time-frequency energy diagram to obtain an initial set.

[0026] In some embodiments of the present invention, the time-frequency feature extraction is performed on the historical grating vibration data to obtain a time-frequency energy diagram, such as Figure 2 As shown, including: S201, reading historical grating vibration data to obtain an original time domain vibration signal; S202, preprocessing the original time domain vibration signal to obtain an initial time domain vibration signal; S203, performing energy conversion on the initial time-domain vibration signal to obtain a time-frequency energy graph.

[0027] In a specific embodiment of the present invention, the collected historical grating vibration data can be read from a distributed storage system, and then the historical grating vibration data can be read to extract the original time domain vibration signal. Then, the original time domain vibration signal can be preprocessed, such as denoising, de-redundancy and other operations to obtain an initial time domain vibration signal. The specific preprocessing process can be set according to the actual situation, and the embodiment of the present invention is not limited here. Then, each vibration signal in the initial time domain vibration signal can be converted into energy using the Hilbert-Huang transform (HHT) technology, so that a time-frequency-energy graph is drawn according to the conversion results of all vibration signals to obtain a time-frequency energy graph. Compared with traditional STFT and CWT, the time-frequency-energy graph generated by HHT can simultaneously show the characteristics of the signal in the three dimensions of time-frequency-energy, providing more abundant and valuable input features for subsequent pattern recognition.

[0028] In some embodiments of the present invention, the time-frequency energy graph is labeled and classified according to the construction condition type to obtain an initial set, such as Figure 3 As shown, including: S301, dividing the time-frequency energy graph according to a preset sliding window to obtain a plurality of sample data of a fixed size; S302, labeling multiple sample data according to the construction condition type to obtain target sample data; S303: Classify the target sample data to obtain an initial set.

[0029] In a specific embodiment of the present invention, a preset sliding window can be set, and the sliding window technology can be used to divide the time-frequency-energy graph generated by HHT into a large number of sample data of fixed size. Then, the target sample data can be classified by manual labeling or semi-supervised learning. For example, the construction working condition type may include normal working condition, drilling, demolition and other types, and the target sample data is labeled with normal working condition, drilling, demolition and other labels. Then, the target sample data can be divided according to a preset ratio to obtain an initial set. The initial set may include a training set, a validation set and a test set, and the preset ratio may be 8:1:1.

[0030] In some embodiments of the present invention, the initial set includes a training set, a validation set, and a test set; Figure 4 As shown, step S104 includes: S401, training and verifying the initial construction type detection model according to the training set and the verification set to obtain a training construction type detection model; S402, testing the trained construction type detection model according to the test set to obtain a test result; S403: When the test result meets the preset conditions, determine the training construction type detection model as the target construction type detection model.

[0031] In a specific embodiment of the present invention, a network model combining a multi-layer Transformer and a sliding attention mechanism module and a graph convolutional neural network (GCN) is designed, that is, an initial construction type detection model, and a two-dimensional spectrum data graph is classified. The network combines a 4-layer Transformer and a sliding attention mechanism module, and a two-layer GCN network, wherein the Transformer can extract long-distance sequence feature information in the time dimension and the frequency dimension respectively, and can perform feature extraction and object recognition more quickly and accurately for long-time series data, and can achieve full parallel calculation compared to the recurrent neural network, greatly accelerating the model training speed and reasoning efficiency. The sliding attention mechanism module can use the sliding window method in the current image object to extract specific information from each window, and combine the information in all windows, and can selectively focus on important information in the input, improving the performance and generalization ability of the model. The graph convolutional neural network can perform a fast calculation on global information, focusing on extracting global features of the image. This method of combining global and local features can effectively improve the accuracy of model recognition.

[0032] The initial construction type detection model can be first trained through the training set, and then the trained model can be verified through the verification set. After the verification is completed, the trained construction type detection model can be obtained, and then the trained construction type detection model can be tested through the test set to obtain the test results. The test results may include the results of the model's recognition and classification of images, such as the recognized images and types.

[0033] In some embodiments of the present invention, step S403 includes: Perform weight calculation on the test results to obtain the weight value; Determine whether the weight value is greater than a preset threshold; If not, the training set is updated and optimized according to the test results, and the target construction type detection model is trained again; If so, it is determined that the test result meets the preset conditions.

[0034] In a specific embodiment of the present invention, the test results can then be weighted, for example, by calculating through the cross entropy loss function, the labels in the test results, and the edge values ​​of the identified images, to obtain the identified images, which are automatically saved in the corresponding label folders. The specific calculation process can be set according to the actual situation, and the embodiment of the present invention is not limited here. It can then be determined whether the weight value is greater than the preset threshold value, and the specific preset threshold value can be set according to the actual situation. If not, the training set can be expanded through the images of the label folder in the label folder test results, so as to update and optimize, iteratively optimize the performance of the classification model, and then the process of training the target construction type detection model through the optimized training set, so as to cycle until the weight value is greater than the preset threshold value, determine that the test results meet the preset conditions, and determine that the training construction type detection model is the target construction type detection model.

[0035] In some embodiments of the present invention, step S105 includes: Perform type detection on the current grating vibration data according to the target construction type detection model to obtain the initial external construction type; Compare the initial external construction type with the construction condition type to obtain the corresponding target type weight coefficient; When the target type weight coefficient is greater than a preset weight threshold, the initial external construction type is determined to be the external construction type.

[0036] In a specific embodiment of the present invention, after obtaining the target construction type detection model, the current grating vibration data can be input into the target construction type detection model, and the target construction type detection model can output the initial external construction type, and then the initial external construction type can be compared with the construction condition type, so as to obtain the target type weight coefficient corresponding to the initial external construction type. For example, if the initial external construction type is a normal condition, the target type weight coefficient of the normal condition in the construction condition type can be determined, and then it can be determined whether the target type weight coefficient is greater than a preset weight threshold, wherein the preset weight threshold can be set according to actual conditions, and the embodiment of the present invention is not limited here. If yes, the initial external construction type is determined to be an external construction type, and if not, the grating vibration data is re-acquired and detected again.

[0037] The embodiment of the present invention adopts Hilbert-Huang transform to generate time-frequency-energy diagram, enriches the time-frequency feature representation of vibration signal, and has stronger non-stationary signal analysis capability combined with time-frequency analysis compared with short-time Fourier transform and wavelet transform. A network model combining multi-layer Transformer, sliding attention mechanism module and GCN is also proposed, which realizes automatic identification and high-precision classification of different working conditions, and improves the intelligent level of subway safety monitoring. The online monitoring and model iteration optimization method is adopted to ensure that the classification model can continuously adapt to the changes in the actual operating environment, and improve the reliability and practicality of the overall solution. In general, the embodiment of the present invention proposes an automatic vibration data classification method combining multi-layer Transformer, sliding attention mechanism module and GCN for the specific application scenario of external intrusion in subway tunnels. It has innovative designs in data acquisition, feature extraction, model training and online application, greatly improves the accuracy of alarm, effectively improves the application level of fiber Bragg grating technology in the field of subway safety monitoring, and provides effective technical support for improving the level of subway safety operation management.

[0038] In order to better implement the subway tunnel external construction type detection method in the embodiment of the present invention, based on the subway tunnel external construction type detection method, the embodiment of the present invention also provides a subway tunnel external construction type detection device, such as Figure 5 As shown, the subway tunnel external construction type detection device 500 includes: The data acquisition module 501 is used to acquire the historical grating vibration data, current grating vibration data and construction condition type outside the subway tunnel, and set the initial construction type detection model; A type calculation module 502 is used to determine a type weight coefficient of each type of construction condition type according to historical grating vibration data and current grating vibration data; A set determination module 503 is used to extract time-frequency features and classify the historical grating vibration data according to the type of construction conditions to obtain an initial set; A model training module 504 is used to train the initial construction type detection model according to the initial set to obtain a target construction type detection model; The type detection module 505 is used to perform type detection on the current grating vibration data according to the target construction type detection model and all type weight coefficients of the construction condition type to obtain the external construction type.

[0039] The subway tunnel external construction type detection device 500 provided in the above embodiment can implement the technical solution described in the above subway tunnel external construction type detection method embodiment. The specific implementation principles of the above modules or units can refer to the corresponding contents in the above subway tunnel external construction type detection method embodiment, which will not be repeated here.

[0040] like Figure 6 As shown, the present invention also provides an electronic device 600. The electronic device 600 includes a processor 601, a memory 602 and a display 603. Figure 6 Only some components of the electronic device 600 are shown, but it should be understood that it is not required to implement all of the components shown, and more or fewer components may be implemented instead.

[0041] In some embodiments, the memory 602 may be an internal storage unit of the electronic device 600, such as a hard disk or memory of the electronic device 600. In other embodiments, the memory 602 may also be an external storage device of the electronic device 600, such as a plug-in hard disk, a smart media card (SMC), a secure digital (SD) card, a flash card, etc., equipped on the electronic device 600.

[0042] Furthermore, the memory 602 may include both an internal storage unit of the electronic device 600 and an external storage device. The memory 602 is used to store application software installed in the electronic device 600 and various data.

[0043] In some embodiments, the processor 601 may be a central processing unit (CPU), a microprocessor or other data processing chip, used to run program codes or process data stored in the memory 602, such as the subway tunnel external construction type detection method of the present invention.

[0044] In some embodiments, the display 603 may be an LED display, a liquid crystal display, a touch-sensitive liquid crystal display, an OLED (Organic Light-Emitting Diode) touch device, etc. The display 603 is used to display information of the electronic device 600 and to display a visual user interface. The components 601-603 of the electronic device 600 communicate with each other through a system bus.

[0045] In some embodiments of the present invention, when the processor 601 executes the subway tunnel external construction type detection program in the memory 602, the following steps may be implemented: Obtain historical grating vibration data, current grating vibration data and construction condition type outside the subway tunnel, and set up an initial construction type detection model; Determine the type weight coefficient of each type of construction condition type according to the historical grating vibration data and the current grating vibration data; According to the construction condition type, the time-frequency features of the historical grating vibration data are extracted and labeled and classified to obtain the initial set; The initial construction type detection model is trained according to the initial set to obtain the target construction type detection model; According to the target construction type detection model and all type weight coefficients of the construction condition type, the current grating vibration data is type detected to obtain the external construction type.

[0046] It should be understood that: when the processor 601 executes the subway tunnel external construction type detection program in the memory 602, in addition to the above functions, other functions can also be implemented. For details, please refer to the description of the corresponding method embodiment above.

[0047] Furthermore, the embodiment of the present invention does not specifically limit the type of the electronic device 600 mentioned, and the electronic device 600 may be a portable electronic device such as a mobile phone, a tablet computer, a personal digital assistant (PDA), a wearable device, a laptop computer, etc. Exemplary embodiments of portable electronic devices include but are not limited to portable electronic devices equipped with IOS, Android, Microsoft or other operating systems. The above-mentioned portable electronic devices may also be other portable electronic devices, such as a laptop computer with a touch-sensitive surface (e.g., a touch panel). It should also be understood that in some other embodiments of the present invention, the electronic device 600 may not be a portable electronic device, but a desktop computer with a touch-sensitive surface (e.g., a touch panel).

[0048] Correspondingly, an embodiment of the present application also provides a computer-readable storage medium, which is used to store computer-readable programs or instructions. When the program or instructions are executed by a processor, it can implement the steps or functions of the subway tunnel external construction type detection method provided by the above-mentioned method embodiments.

[0049] Those skilled in the art will appreciate that all or part of the processes of the above-mentioned embodiments can be implemented by instructing related hardware (such as a processor, a controller, etc.) through a computer program, and the computer program can be stored in a computer-readable storage medium, wherein the computer-readable storage medium is a disk, an optical disk, a read-only storage memory, or a random access memory, etc.

[0050] The above is a detailed introduction to the subway tunnel external construction type detection method, device, equipment and medium provided by the present invention. Specific examples are used in this article to illustrate the principles and implementation methods of the present invention. The description of the above embodiments is only used to help understand the method of the present invention and its core idea; at the same time, for technical personnel in this field, according to the idea of ​​the present invention, there will be changes in the specific implementation method and application scope. In summary, the content of this specification should not be understood as a limitation on the present invention.

Claims

1. A method for detecting the external construction type of a subway tunnel, characterized in that: include: Obtain historical grating vibration data, current grating vibration data and construction condition type outside the subway tunnel, and set up an initial construction type detection model; Determining a type weight coefficient of each type of the construction condition type according to the historical grating vibration data and the current grating vibration data; Extracting time-frequency features and labeling classification of the historical grating vibration data according to the type of construction conditions to obtain an initial set; Training the initial construction type detection model according to the initial set to obtain a target construction type detection model; The current grating vibration data is subjected to type detection according to the target construction type detection model and all type weight coefficients of the construction condition type to obtain an external construction type.

2. The method for detecting the external construction type of a subway tunnel according to claim 1, characterized in that: The extracting time-frequency features and labeling classification of the historical grating vibration data according to the construction condition type to obtain an initial set includes: Extracting time-frequency features of the historical grating vibration data to obtain a time-frequency energy graph; The time-frequency energy diagram is labeled and classified according to the construction condition type to obtain an initial set.

3. The method for detecting the external construction type of a subway tunnel according to claim 2, characterized in that: The step of extracting time-frequency features from the historical grating vibration data to obtain a time-frequency energy graph includes: Reading the historical grating vibration data to obtain an original time-domain vibration signal; Preprocessing the original time-domain vibration signal to obtain an initial time-domain vibration signal; Energy conversion is performed on the initial time-domain vibration signal to obtain a time-frequency energy diagram.

4. The method for detecting the external construction type of a subway tunnel according to claim 2, characterized in that: The time-frequency energy graph is labeled and classified according to the construction condition type to obtain an initial set, including: The time-frequency energy graph is segmented according to a preset sliding window to obtain a plurality of sample data of a fixed size; Labeling the plurality of sample data according to the construction condition type to obtain target sample data; The target sample data is classified to obtain an initial set.

5. The method for detecting the external construction type of a subway tunnel according to claim 1, characterized in that: The initial set includes a training set, a validation set and a test set; the initial construction type detection model is trained according to the initial set to obtain a target construction type detection model, including: Training and verifying the initial construction type detection model according to the training set and the verification set to obtain a training construction type detection model; Testing the training construction type detection model according to the test set to obtain a test result; When the test result meets the preset conditions, the training construction type detection model is determined to be the target construction type detection model.

6. The method for detecting the external construction type of a subway tunnel according to claim 5, characterized in that: When the test result meets the preset condition, determining the training construction type detection model as the target construction type detection model includes: Performing weight calculation on the test results to obtain a weight value; Determine whether the weight value is greater than a preset threshold; If not, the training set is updated and optimized according to the test results, and the target construction type detection model is trained again; If so, it is determined that the test result meets the preset condition.

7. The method for detecting the external construction type of a subway tunnel according to claim 1, characterized in that: The method of performing type detection on the current grating vibration data according to the target construction type detection model and all type weight coefficients of the construction condition type to obtain the external construction type includes: Performing type detection on the current grating vibration data according to the target construction type detection model to obtain an initial external construction type; Compare the initial external construction type with the construction condition type to obtain a corresponding target type weight coefficient; When the target type weight coefficient is greater than a preset weight threshold, the initial external construction type is determined to be an external construction type.

8. A device for detecting the external construction type of a subway tunnel, characterized in that: include: A data acquisition module is used to obtain historical grating vibration data, current grating vibration data and construction condition type outside the subway tunnel, and set an initial construction type detection model; A type calculation module, used for determining a type weight coefficient of each type of the construction condition type according to the historical grating vibration data and the current grating vibration data; A set determination module, used for extracting time-frequency features and labeling and classifying the historical grating vibration data according to the type of construction conditions to obtain an initial set; A model training module, used for training the initial construction type detection model according to the initial set to obtain a target construction type detection model; The type detection module is used to perform type detection on the current grating vibration data according to the target construction type detection model and all type weight coefficients of the construction condition type to obtain the external construction type.

9. An electronic device, characterized in that: include: A processor, a memory, and a computer program stored in the memory and capable of running on the processor, wherein when the computer program is executed by the processor, the steps of the method for detecting the external construction type of a subway tunnel as described in any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the method for detecting the external construction type of a subway tunnel as described in any one of claims 1 to 7 are implemented.

Images