Cable damage early warning methods, devices and early warning systems

Abstract

The cable damage early warning method, device, and system provided in this application include: collecting vibration signals from the ground surface, determining the corresponding feature vector set based on the vibration signals, then calling a preset sound classification model and inputting the feature vector set into the sound classification model to obtain sound information corresponding to the feature vector set; when the sound information belongs to the construction type, triggering an alarm voice and sending the sound information to each terminal in a preset terminal list. By setting up an early warning system in the soil above the underground cable, the cable damage early warning device in the early warning system can collect vibration signals from the ground surface to provide early warning in conjunction with the sound classification model. The warning is delivered through on-site voice alerts and online notifications to maintenance personnel, achieving a better warning effect and minimizing external damage. It also reduces the number of manual inspections and lowers maintenance costs.

Description

Technical Field

[0001] This application relates to the field of power grid safety technology, and in particular to a method, device and early warning system for cable damage warning. Background Technology

[0002] With the rapid development of urban construction, large-scale projects such as municipal construction, road excavation, and outdoor operations have been launched. While these projects have brought prosperity and development to cities, they have also brought some challenges. In particular, during the operation of cable lines, because cable lines need to be laid over long distances and their operation is closely connected with the surrounding environment, accidents caused by external forces damaging cables occur frequently, thus posing a significant hidden danger to the safe operation of cables.

[0003] Currently, the application of 10KV cable power supply is becoming increasingly widespread. Generally, dedicated transformer users are connected to the power supply through a switch room or overhead line, and the power is supplied to the user's substation via cable. The incoming cables for dedicated transformer users are often laid using methods such as pipe jacking or direct burial. In response to cable damage accidents caused by external forces, cable identification signs or cable stakes are set up in the cable corridor to provide warnings on-site. However, this method is not only costly to maintain, but also cannot provide early warning of external damage events. Summary of the Invention

[0004] The purpose of this application is to at least address one of the aforementioned technical defects, particularly the technical defect in the prior art of using cable markers or cable stakes in cable corridors for warning purposes, which is not only costly to maintain but also fails to provide early warning of external damage events.

[0005] In a first aspect, this application provides a cable damage early warning method, applied to a cable damage early warning device in an early warning system, wherein the early warning system is placed in the soil above the underground cable; the method includes:

[0006] Collect vibration signals from the Earth's surface;

[0007] Based on the vibration signal, determine the corresponding feature vector set;

[0008] A preset sound classification model is invoked, and the feature vector set is input into the sound classification model to obtain the sound information corresponding to the feature vector set;

[0009] When the sound information is related to construction, a warning voice is triggered, and the sound information is sent to each terminal in the preset terminal list.

[0010] In one embodiment, the method further includes:

[0011] If no vibration signal is collected from the ground surface within a preset time interval, the low-power operation mode is activated.

[0012] In one embodiment, the step of determining the corresponding feature vector set based on the vibration signal includes:

[0013] The vibration signal is separated to obtain the background sound signal and the destruction sound signal;

[0014] Time-frequency analysis is performed on the background sound signal and the destructive sound signal respectively to extract the energy distribution features of the background sound signal and the destructive sound signal, and a feature vector set is generated based on the energy distribution features of the background sound signal and the destructive sound signal.

[0015] In one embodiment, the step of separating the vibration signal includes:

[0016] The vibration signal is preprocessed to obtain the signal to be analyzed;

[0017] Independent component analysis is performed on the signal to be analyzed to separate the background sound signal and the destructive sound signal from the signal to be analyzed.

[0018] In one embodiment, the step of generating the feature vector set includes:

[0019] Based on the energy distribution characteristics of the background sound signal, the energy value of the background sound signal in each frequency band is determined, so as to generate a first feature vector based on the energy value of the background sound signal in each frequency band;

[0020] Based on the energy distribution characteristics of the destructive sound signal, the energy value of the destructive sound signal in each frequency band is determined, and a second feature vector is generated based on the energy value of the destructive sound signal in each frequency band.

[0021] A feature vector set is formed using the first feature vector and the second feature vector.

[0022] In one embodiment, the sound classification model is trained based on the Transformer neural network architecture.

[0023] In one embodiment, the training process of the sound classification model includes:

[0024] Obtain the training dataset and its corresponding sample dataset; wherein the training dataset includes single-class data samples and combined data samples;

[0025] Determine the Transformer model;

[0026] The Transformer model is iteratively trained based on the training dataset and the sample dataset to obtain the trained Transformer model, and the trained Transformer model is determined as the sound classification model.

[0027] Secondly, this application provides a cable damage early warning device, which includes a wireless module, a field alarm module, a vibration monitoring module, and a soundprint monitoring module;

[0028] The vibration monitoring module is used to collect vibration signals from the ground surface and upload the vibration signals to the acoustic signature monitoring module;

[0029] The voiceprint monitoring module is used to determine the corresponding feature vector set based on the received vibration signal, and call the preset sound classification model to input the feature vector set into the sound classification model to obtain the sound information corresponding to the feature vector set. When the sound information belongs to the construction type, the sound information is sent to the wireless module and the on-site alarm module.

[0030] The on-site alarm module is used to trigger an alarm voice when it receives the sound information;

[0031] The wireless module is used to send the received audio information to each terminal in a preset terminal list.

[0032] In one embodiment, the device further includes a power management module; the power management module is used to activate a low-power operation mode when the vibration monitoring module does not collect a vibration signal within a preset time interval.

[0033] Thirdly, this application provides an early warning system, the system including a housing, a top cover, a vibrating spring, and a cable damage early warning device as described in any of the above embodiments;

[0034] The cable damage warning device is placed inside the housing, and the top cover is used to close the housing to protect the cable damage warning device.

[0035] The cable damage early warning device is used to perform the cable damage early warning method as described in any of the above embodiments;

[0036] The vibrating spring is fixed to the bottom of the outer shell and is used to receive vibration signals from the ground.

[0037] As can be seen from the above technical solutions, the embodiments of this application have the following advantages:

[0038] This application provides a cable damage early warning method, device, and system. The method is applied to a cable damage early warning device within an early warning system, which is placed in the soil above the underground cable. The method includes: collecting vibration signals from the ground surface; determining the corresponding feature vector set based on the vibration signals; then calling a preset sound classification model and inputting the feature vector set into the sound classification model to obtain sound information corresponding to the feature vector set; when the sound information belongs to a construction type, triggering an alarm voice and sending the sound information to various terminals in a preset terminal list. By setting up an early warning system in the soil above the underground cable, the cable damage early warning device in the system can collect vibration signals from the ground surface and combine them with a sound classification model for early warning. The warning is delivered through on-site voice alerts and online notifications to maintenance personnel, achieving a better warning effect. This approach minimizes external damage, notifies maintenance personnel of the cable's on-site environment in advance, reduces the frequency of manual inspections, and lowers maintenance costs. Attached Figure Description

[0039] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0040] Figure 1 A schematic flowchart illustrating a cable damage early warning method provided in an embodiment of this application;

[0041] Figure 2 A flowchart illustrating the steps for determining the corresponding feature vector set based on the vibration signal, as provided in an embodiment of this application;

[0042] Figure 3 This is one of the structural schematic diagrams of a cable damage early warning device provided in the embodiments of this application;

[0043] Figure 4 This is a second schematic diagram of the structure of a cable damage early warning device provided in an embodiment of this application;

[0044] Figure 5 This is a schematic diagram of the structure of an early warning system provided in an embodiment of this application. Detailed Implementation

[0045] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of this application.

[0046] In one embodiment, this application provides a cable damage early warning method. The following embodiments illustrate the application of this method to a cable damage early warning device within an early warning system. It is understood that when cable markers or cable posts are used for warnings, maintenance personnel often only learn of the cable fault after it has already been damaged. Furthermore, a large number of cable markers or cable posts are needed to achieve a slightly better warning effect, resulting in high construction costs. Additionally, regular inspections by maintenance personnel are required during this process. Based on this, this application provides a cable damage early warning method, and places the early warning system implementing this method in the soil above underground cables. Figure 1 As shown, cable damage early warning methods include:

[0047] Step S101: Collect vibration signals from the ground surface.

[0048] Vibration signal refers to the signal generated due to the vibration of an object.

[0049] In this step, because the early warning system is placed in the soil above the underground cables, vibration signals from the ground surface can be collected when construction activities are taking place at the site where the underground cables are located, and these collected vibration signals can be analyzed. Moreover, sound travels fastest through solids, so collecting vibration signals from the ground surface allows for rapid acquisition of vibration signals, thereby improving the response speed of the cable damage early warning method.

[0050] Furthermore, when early warning of underground cables is needed, areas can be divided according to the distribution location of the underground cables, and early warning systems can be set up in the divided areas. Once the early warning system is set up, it can be activated, and at this time, the early warning system will detect and collect vibration signals from the ground in real time.

[0051] It is understood that the density and specific locations of the early warning system can be adjusted and set according to the actual environmental conditions and actual needs, and this application does not impose specific restrictions in this regard. For example, when the accuracy of the early warning is required to be high, the distribution density of the early warning system can be increased to enhance the intensity of the vibration signal.

[0052] Specifically, when underground cables are installed at a site where construction is being carried out, construction noise is generally generated, mainly from the following aspects:

[0053] (1) Mechanical noise: The mechanical noise of various types of construction machinery and equipment on site, such as excavators, pile drivers, cranes, graders and transport vehicles on site, can usually be regarded as point noise sources compared with the entire construction site.

[0054] (2) On-site operation noise: During the construction operation, the collision, cutting, grinding and other noises between construction machinery and building materials or the ground, such as the sound of pile driver hitting the ground, the sound of cutting steel bar by cutting machine and the sound of grinding machine rubbing against the ground surface, the instantaneous sound pressure level of these noises may reach more than 130dB(A).

[0055] (3) Collision noise: noise generated by collisions during handling or vehicle loading and unloading, scattered noise generated by loading and unloading scaffolding, templates, etc., as well as knocking sounds from workers, impact sounds from metal materials, etc., which are usually instantaneous noises.

[0056] (4) Construction worker noise: noise generated by workers talking, shouting, and using hand tools at the construction site.

[0057] Step S102: Determine the corresponding feature vector set based on the vibration signal.

[0058] The feature vector set describes the distribution of the vibration signal at different frequency components.

[0059] In this step, when a vibration signal is acquired, it needs to be processed by separation and component analysis to obtain a feature vector set corresponding to the vibration signal, so as to match the input specification of the sound classification model.

[0060] Step S103: Call the preset sound classification model and input the feature vector set into the sound classification model to obtain the sound information corresponding to the feature vector set.

[0061] Here, sound information refers to classification information and damage level information. Sound classification models are used to analyze sound features in a feature vector set to determine the corresponding classification and damage level information.

[0062] In this step, a pre-trained sound classification model is invoked, and then the feature vector set is input into the sound classification model so that the sound classification model can perform inference based on the feature vector set and output the analysis results of the feature vector set, i.e., sound information.

[0063] Step S104: When the sound information belongs to the construction type, trigger the warning voice and send the sound information to each terminal in the preset terminal list.

[0064] The construction type can be any one or more combinations from the preset construction type list.

[0065] In this step, the classification information from the sound data is first acquired. Based on this classification information, it is determined whether the sound data belongs to any one or more combinations of a preset construction type list. If not, the cable damage warning ends, and ground vibration data continues to be collected. If so, an alarm voice is triggered to alert nearby construction teams to the location of the underground cable and prevent damage. Furthermore, the sound data is sent to various terminals in the terminal list, enabling maintenance personnel to understand the specific construction type (classification information) and damage level (damage grade information) around the underground cable.

[0066] Understandably, the terminal list can be set up by operations and maintenance personnel. For example, the terminal list for different regional early warning systems can be set up according to the person in charge of the corresponding region. When the early warning system detects the presence of a construction team nearby, it can send specific information to the corresponding terminals according to the pre-determined terminal list, thereby achieving the purpose of notifying the operations and maintenance personnel in advance.

[0067] In one example, the communication method for sending voice information is wireless communication, which can be achieved by setting an antenna to send voice information via 4G / 5G communication.

[0068] In another embodiment, the early warning system can also be equipped with a light-emitting device. When the sound information is of the construction type, the on-site prompt measures can be triggered by flashing lights through the light-emitting device in addition to triggering the warning voice.

[0069] This application provides a cable damage early warning method, device, and system. The method is applied to a cable damage early warning device within an early warning system, which is placed in the soil above the underground cable. The method includes: collecting vibration signals from the ground surface; determining the corresponding feature vector set based on the vibration signals; then calling a preset sound classification model and inputting the feature vector set into the sound classification model to obtain sound information corresponding to the feature vector set; when the sound information belongs to a construction type, triggering an alarm voice and sending the sound information to various terminals in a preset terminal list. By setting up an early warning system in the soil above the underground cable, the cable damage early warning device in the system can collect vibration signals from the ground surface and combine them with a sound classification model for early warning. The warning is delivered through on-site voice alerts and online notifications to maintenance personnel, achieving a better warning effect. This approach minimizes external damage, notifies maintenance personnel of the cable's on-site environment in advance, reduces the frequency of manual inspections, and lowers maintenance costs.

[0070] In one embodiment, the cable damage early warning method further includes:

[0071] If no vibration signal is collected from the ground surface within a preset time interval, the low-power operation mode is activated.

[0072] The preset time period refers to a preset time length, such as 40 minutes, 1 hour, or 3 hours.

[0073] In this embodiment, if no vibration signals are collected from the ground surface within a preset time interval, it indicates that there is likely no construction activity at the site of the underground cable. In this case, the power consumption of the cable damage early warning device can be reduced by lowering the operating frequency of some modules, thus entering a low-power operation mode. This allows the device to enter a low-power operation mode when the risk of damage is low, reducing resource consumption and improving system resource utilization while ensuring the safety of the underground cable.

[0074] like Figure 2 As shown, in one embodiment, the step of determining the corresponding feature vector set based on the vibration signal includes:

[0075] Step S201: Separate the vibration signal to obtain the background sound signal and the destruction sound signal.

[0076] Background sound signals refer to sound signals unrelated to construction activities, such as the sound of passing cars, human voices, thunder, and other normal sound signals. Disruptive sound signals refer to sound signals generated by the construction team during construction activities.

[0077] In this step, since the acquired vibration signals can be broadly categorized into background sound and destructive sound, signal separation techniques are used to separate the vibration signals into background sound and destructive sound signals. Specifically, signal separation techniques include, but are not limited to, independent component analysis, deep learning, nonnegative matrix factorization, or adaptive filters. The specific signal separation technique can be selected according to specific needs, and this application does not impose any specific limitations on it.

[0078] Step S202: Perform time-frequency analysis on the background sound signal and the destructive sound signal respectively to extract the energy distribution characteristics of the background sound signal and the destructive sound signal.

[0079] Among them, energy distribution characteristics are used to describe the energy distribution of a signal at different frequencies.

[0080] In this step, methods such as Fourier transform or wavelet transform can be used to convert the background sound signal and the destructive sound signal from the time domain to the frequency domain in order to perform time domain analysis and observe the distribution of the signal at different frequencies.

[0081] Furthermore, in this embodiment, particular attention is paid to the energy distribution characteristics of the signal during time-frequency analysis. Energy distribution characteristics reflect the signal's intensity at different frequencies, which is crucial for identifying and distinguishing different types of sound signals. For example, background sound signals may have higher energy in the low-frequency range, while disruptive sound signals may exhibit more energy concentration in the high-frequency range. By analyzing these energy distribution characteristics, a better understanding of the composition of sound signals can be achieved, providing important information for subsequent sound signal processing and sound recognition tasks.

[0082] Step S203: Generate a feature vector set based on the energy distribution characteristics of the background sound signal and the energy distribution characteristics of the disruptive sound signal.

[0083] In one embodiment, the step of separating the vibration signal includes:

[0084] The vibration signal is preprocessed to obtain the signal to be analyzed;

[0085] Independent component analysis is performed on the signal to be analyzed to separate the background sound signal and the destructive sound signal from the signal to be analyzed.

[0086] In this embodiment, the vibration signal is first centered by subtracting the mean from each variable value. This simplifies the calculation and preprocesses the vibration signal, resulting in the preprocessed vibration signal, which is the signal to be analyzed. This improves the processing performance for subsequent independent cost analysis. Next, independent component analysis is performed on the signal to be analyzed to identify independent components. Background sound and destructive sound signals are separated from the signal to be analyzed, using two categories: background sound and destructive sound.

[0087] Understandably, independent component analysis can effectively identify the independent components in the signal to be analyzed, making it possible to clearly distinguish between the originally mixed background sound signal and the damage sound signal, so as to further analyze the different sound signals and improve the accuracy of the final output sound information, thereby improving the accuracy of cable damage early warning.

[0088] In one example, the independent component analysis method used in this embodiment is the FastICA (Fast Independent Component Analysis) method.

[0089] In one embodiment, the step of generating a feature vector set includes:

[0090] Based on the energy distribution characteristics of the background sound signal, the energy value of the background sound signal in each frequency band is determined, and a first feature vector is generated based on the energy value of the background sound signal in each frequency band.

[0091] Based on the energy distribution characteristics of the disruptive sound signal, the energy value of the disruptive sound signal in each frequency band is determined, and a second feature vector is generated based on the energy value of the disruptive sound signal in each frequency band.

[0092] The first and second eigenvectors form an eigenvector set.

[0093] The first feature vector is used to describe the intensity of the background sound signal in different frequency bands, and the second feature vector is used to describe the intensity of the destructive sound signal in different frequency bands.

[0094] In this embodiment, based on the energy distribution characteristics of the background sound signal and the energy distribution characteristics of the destructive sound signal, the energy values ​​of the background sound signal and the destructive sound signal in each frequency band are calculated respectively, and the energy values ​​in each frequency band are used to form a first feature vector corresponding to the background sound signal and a second feature vector corresponding to the destructive sound signal.

[0095] For example, assuming that the energy distribution characteristics of the background sound signal involve frequency bands a, b, and c, the total energy value A of frequency band a, the total energy value B of frequency band b, and the total energy value C of frequency band c are calculated sequentially according to the energy distribution characteristics. Then, the first feature vector corresponding to the background sound signal is [A, B, C].

[0096] It is understandable that by calculating the energy values ​​of each frequency band through energy distribution characteristics, and then constructing the first feature vector and the second feature vector, the sound classification model can better identify sound information, determine its category information, and thus improve the accuracy of cable damage early warning and reduce the probability of false alarms.

[0097] In one embodiment, the sound classification model is trained based on the Transformer neural network architecture.

[0098] In this embodiment, the Transformer neural network architecture is a deep learning model mainly used to process sequential data. Its core idea is to use the attention mechanism to capture long-distance dependencies in sequential data, and it has good generalization ability.

[0099] In one embodiment, the training process of the sound classification model includes:

[0100] Obtain the training dataset and its corresponding sample dataset; the training dataset includes single-class data samples and combined data samples.

[0101] Determine the Transformer model;

[0102] The Transformer model is iteratively trained based on the training dataset and the sample dataset to obtain the trained Transformer model, which is then identified as the sound classification model.

[0103] Specifically, the training dataset includes multiple sound samples, including single-class samples and combined samples. Single-class samples are sound samples with only one class of information, while combined samples are sound samples with a mixture of two or more class information. The sample dataset includes the sound information of each data sample in the training dataset.

[0104] In this embodiment, the Transformer model is iteratively trained using the training dataset and its corresponding sample dataset until a preset stopping condition is met. At this point, all network parameters of the trained Transformer model are fixed, and the trained Transformer model is identified as the sound classification model. Further, the stopping condition can be reaching a preset number of iterations, or the change in model parameters between two iterations being less than a preset threshold; this application does not impose specific limitations on this.

[0105] In one example, the workflow of the Transformer model in this embodiment is as follows: First, a single fully connected layer is used to process a vector of length 40 into a vector of length 80 for preliminary feature encoding; second, the encoder in the Transformer is used for feature extraction, with a model dimension of 80, using two multi-head attention mechanisms, and internally augmented to 256 bits; finally, two fully connected layers are used for classification, increasing the dimension to 600 to obtain a 5-bit output.

[0106] Understandably, when training a sound classification model using the Transformer architecture, the Transformer model's attention mechanism can capture long-distance dependencies and effectively handle complex patterns in time-series data. At the same time, the Transformer model has parallel processing capabilities, resulting in fast training speed and rapid convergence. Furthermore, the Transformer model has strong generalization ability, making it well-suited for different sound classification tasks and improving the model's accuracy and robustness.

[0107] It should be understood that although the steps in the flowcharts of the embodiments described above are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the embodiments described above may include multiple steps or multiple stages. These steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the steps or stages of other steps.

[0108] The cable damage early warning device provided in the embodiments of this application is described below. The cable damage early warning device described below can be referred to in correspondence with the cable damage early warning method described above.

[0109] like Figure 3 As shown, this application provides a cable damage early warning device 300, which includes a wireless module 301, a field alarm module 302, a vibration monitoring module 303, and a soundprint monitoring module 304.

[0110] The vibration monitoring module 303 is used to collect vibration signals from the ground surface and upload the vibration signals to the acoustic signature monitoring module 304;

[0111] The voiceprint monitoring module 302 is used to determine the corresponding feature vector set based on the received vibration signal, and call the preset sound classification model to input the feature vector set into the sound classification model to obtain the sound information corresponding to the feature vector set. When the sound information belongs to the construction type, the sound information is sent to the wireless module 301 and the on-site alarm module 302.

[0112] The on-site alarm module 302 is used to trigger an alarm voice when it receives sound information;

[0113] The wireless module 301 is used to send the received audio information to each terminal in a preset terminal list.

[0114] The vibration monitoring module 303 includes a vibration sensor for receiving and acquiring vibration signals from the vibrating spring.

[0115] In this embodiment, when the voiceprint monitoring model 302 determines the sound information and classifies it as construction-related, it sends the sound information to the wireless module 301 and the on-site alarm module 302. The on-site alarm module 302 then issues a warning voice alert to construction equipment and personnel, reminding them of the presence of underground cables and prohibiting any damage. Meanwhile, the wireless module 301 can transmit the sound information via its antenna using 4G / 5G communication to the terminal of the corresponding maintenance personnel, thus alerting them to the cable maintenance staff.

[0116] in addition, Figure 3 The connection relationships between the various modules indicate the main data interaction flow and do not constitute a limitation on the specific data interaction in the cable damage early warning device.

[0117] Understandably, by installing an early warning system in the soil above underground cables, the cable damage warning device in the system can collect vibration signals from the ground surface and combine them with a sound classification model to provide early warnings. Furthermore, the warnings can be delivered through on-site voice alerts and online notifications to maintenance personnel, achieving a better warning effect. This approach can minimize external damage, inform maintenance personnel of the cable's on-site environment in advance, reduce the number of manual inspections, and lower maintenance costs.

[0118] like Figure 4 As shown, in one embodiment, the cable damage early warning device 300 further includes a power management module 305; the power management module 305 is used to start a low-power operation mode when the vibration monitoring module 303 does not collect a vibration signal within a preset time interval.

[0119] In one embodiment, the voiceprint monitoring module includes:

[0120] The signal separation submodule is used to separate the vibration signal to obtain the background sound signal and the destruction sound signal;

[0121] The time-frequency analysis submodule is used to perform time-frequency analysis on the background sound signal and the destructive sound signal respectively, so as to extract the energy distribution characteristics of the background sound signal and the destructive sound signal, and generate a feature vector set based on the energy distribution characteristics of the background sound signal and the destructive sound signal.

[0122] In one embodiment, the signal separation submodule includes:

[0123] The preprocessing unit is used to preprocess the vibration signal to obtain the signal to be analyzed;

[0124] The component analysis unit is used to perform independent component analysis on the signal to be analyzed, so as to separate the background sound signal and the destructive sound signal from the signal to be analyzed.

[0125] In one embodiment, the time-frequency analysis submodule includes:

[0126] The first generation unit is used to determine the energy value of the background sound signal in each frequency band according to the energy distribution characteristics of the background sound signal, so as to generate a first feature vector according to the energy value of the background sound signal in each frequency band.

[0127] The second generation unit is used to determine the energy value of the disruptive sound signal in each frequency band based on the energy distribution characteristics of the disruptive sound signal, so as to generate a second feature vector based on the energy value of the disruptive sound signal in each frequency band.

[0128] The vector set generation unit is used to form a feature vector set using the first feature vector and the second feature vector.

[0129] In one embodiment, the cable damage early warning device further includes:

[0130] The dataset acquisition module is used to acquire the training dataset and its corresponding sample dataset; the training dataset includes single-class data samples and combined data samples.

[0131] The model determination module is used to determine the Transformer model;

[0132] The model training module is used to iteratively train the Transformer model based on the training dataset and the sample dataset to obtain the trained Transformer model, and then identify the trained Transformer model as a sound classification model.

[0133] The division of modules in the above-described cable damage early warning device is merely illustrative. In other embodiments, the cable damage early warning device can be divided into different modules as needed to complete all or part of its functions. Each module in the above-described cable damage early warning device can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in a computer device in hardware form, or stored in the memory of a computer device in software form, so that the processor can call and execute the operations corresponding to each module.

[0134] like Figure 5 As shown, this application provides an early warning system, which includes a housing 1, a top cover 2, a vibrating spring 4, and a cable damage early warning device 3 as described in any of the above embodiments;

[0135] The cable damage warning device 3 is placed inside the housing 1, and the top cover 2 is used to close the housing 1 to protect the cable damage warning device 3.

[0136] The cable damage early warning device 3 is used to perform the cable damage early warning method as described in any of the above embodiments;

[0137] The vibrating spring 4 is fixed to the bottom of the outer casing 1 and is used to receive vibration signals from the ground.

[0138] Those skilled in the art will understand that Figure 5 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the early warning system to which the present application is applied. For example, the vibrating spring 3 can be replaced with any device capable of vibration response. A specific early warning system may include more or fewer components than shown in the figure, or combine certain components, or have different component arrangements.

[0139] Finally, it should be noted that in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising a…" does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element. In this document, the singular forms "a," "an," and "the" may also include the plural forms unless the context clearly indicates otherwise. It should also be understood that the terms “comprising / including” or “having” specify the presence of the stated features, wholes, steps, operations, components, parts or combinations thereof, but do not exclude the possibility of the presence or addition of one or more other features, wholes, steps, operations, components, parts or combinations thereof. Meanwhile, the term “and / or” as used in this specification includes any and all combinations of the associated listed items.

[0140] The various embodiments in this specification are described in a progressive manner. Each embodiment focuses on the differences from other embodiments. The various embodiments can be combined as needed, and the same or similar parts can be referred to each other.

[0141] The above description of the disclosed embodiments enables those skilled in the art to make or use this application. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of this application. Therefore, this application is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for early warning of cable damage, characterized in that, A cable damage early warning device is used in an early warning system, wherein the early warning system is placed in the soil above the underground cable; The method includes: Collect vibration signals from the Earth's surface; The vibration signal is separated to obtain the background sound signal and the destruction sound signal; Time-frequency analysis is performed on the background sound signal and the destructive sound signal respectively to extract the energy distribution features of the background sound signal and the destructive sound signal, and a feature vector set is generated based on the energy distribution features of the background sound signal and the destructive sound signal. A preset sound classification model is invoked, and the feature vector set is input into the sound classification model to obtain the sound information corresponding to the feature vector set; When the sound information is related to construction, a warning voice is triggered, and the sound information is sent to each terminal in the preset terminal list.

2. The cable damage early warning method according to claim 1, characterized in that, The method further includes: If no vibration signal is collected from the ground surface within a preset time interval, the low-power operation mode is activated.

3. The cable damage early warning method according to claim 2, characterized in that, The step of separating the vibration signal includes: The vibration signal is preprocessed to obtain the signal to be analyzed; Independent component analysis is performed on the signal to be analyzed to separate the background sound signal and the destructive sound signal from the signal to be analyzed.

4. The cable damage early warning method according to claim 2, characterized in that, The step of generating the feature vector set includes: Based on the energy distribution characteristics of the background sound signal, the energy value of the background sound signal in each frequency band is determined, so as to generate a first feature vector based on the energy value of the background sound signal in each frequency band; Based on the energy distribution characteristics of the destructive sound signal, the energy value of the destructive sound signal in each frequency band is determined, and a second feature vector is generated based on the energy value of the destructive sound signal in each frequency band. A feature vector set is formed using the first feature vector and the second feature vector.

5. The cable damage early warning method according to any one of claims 1 to 4, characterized in that, The sound classification model is trained based on the Transformer neural network architecture.

6. The cable damage early warning method according to claim 5, characterized in that, The training process of the sound classification model includes: Obtain the training dataset and its corresponding sample dataset; wherein the training dataset includes single-class data samples and combined data samples; Determine the Transformer model; The Transformer model is iteratively trained based on the training dataset and the sample dataset to obtain the trained Transformer model, and the trained Transformer model is determined as the sound classification model.

7. A cable damage early warning device, characterized in that, The device includes a wireless module, a field alarm module, a vibration monitoring module, and a voiceprint monitoring module; The vibration monitoring module is used to collect vibration signals from the ground surface and upload the vibration signals to the acoustic signature monitoring module; The voiceprint monitoring module is used to separate the vibration signal to obtain a background sound signal and a destructive sound signal. Time-frequency analysis is performed on the background sound signal and the destructive sound signal respectively to extract energy distribution characteristics. Based on the energy distribution characteristics of the background sound signal and the destructive sound signal, a feature vector set is generated. A preset sound classification model is invoked, and the feature vector set is input into the sound classification model to obtain sound information corresponding to the feature vector set. Furthermore, when the sound information belongs to the construction type, the sound information is sent to the wireless module and the on-site alarm module. The on-site alarm module is used to trigger an alarm voice when it receives the sound information; The wireless module is used to send the received audio information to each terminal in a preset terminal list.

8. The cable damage early warning device according to claim 7, characterized in that, The device also includes a power management module; the power management module is used to start a low-power operation mode when the vibration monitoring module does not collect a vibration signal within a preset time interval.

9. An early warning system, characterized in that, The system includes a housing, a top cover, a vibrating spring, and a cable damage early warning device as described in claim 7 or 8; The cable damage warning device is placed inside the housing, and the top cover is used to close the housing to protect the cable damage warning device. The cable damage early warning device is used to perform the cable damage early warning method as described in any one of claims 1 to 6; The vibrating spring is fixed to the bottom of the outer shell and is used to receive vibration signals from the ground.

Images