Distribution network vibration source type information generation method, device, electronic equipment and medium
Through the vibrating fiber sensor network, the distribution network vibration source signal is collected and processed, and combined with time domain and frequency domain analysis methods and deep learning, the distribution network vibration source type information is generated, which solves the problem of data processing delay and recognition accuracy in the existing technology. Problem, improving monitoring efficiency and stability.
Patent Information
- Application Number
- CN202410841639.6
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-06-27
- Publication Date
- 2024-11-19
- Estimated Expiration
- 2044-06-27
AI Technical Summary
In the prior art, when the monitoring system based on vibrating fiber generates the distribution network vibration source type information, there are problems such as data processing delay, identification accuracy and stability, which are mainly due to artificial empirical subjectivity and frequency domain analysis method. Vibration data distortion.
Vibration fiber sensor network is used to collect vibration signals in real time, and through signal conversion, preprocessing, feature extraction and classification processing, combined with time domain analysis method and frequency domain analysis method, deep learning method is used to generate vibration source type information of distribution networks to reduce data distortion. Improve identification accuracy and stability.
It improves the monitoring efficiency of distribution network, reduces the delay of monitoring results, enhances the accuracy and stability of identification and classification of various source types, and realizes rapid positioning and maintenance.
Smart Images

Figure CN118395356B_ABST
Abstract
Description
Technical Field
[0001] Embodiments of the present disclosure relate to the field of computer technology, and in particular to a method, device, electronic device, and medium for generating distribution network vibration source type information. Background Art
[0002] Identifying the type of vibration source in the distribution network can monitor the operating status of the distribution network in real time and promptly detect potential faults or anomalies in the distribution network. At present, the identification of the type of vibration source in the distribution network mainly relies on manual experience or frequency domain analysis to analyze the vibration data generated by the monitoring system of the vibration optical fiber.
[0003] However, in practice, when the above method is used to generate distribution network vibration source type information, the following technical problems often occur:
[0004] The monitoring system based on vibration optical fiber will generate a large amount of vibration data. Efficient and accurate processing and analysis of a large amount of vibration data requires a lot of manpower and a lot of time, resulting in a large delay in obtaining analysis results. Due to the variety of vibration source types and the strong subjectivity of manual experience, the accuracy of identifying and classifying various vibration source types is low, resulting in low monitoring efficiency of the distribution network. Analyzing vibration data only through frequency domain analysis method, due to the need for discrete Fourier transform, there are many cases of vibration data distortion, resulting in low stability in identifying and classifying various vibration source types.
[0005] The above information disclosed in this Background section is only for enhancement of understanding of the background of the inventive concept and therefore it may contain information that does not form the prior art that is already known in this country to a person of ordinary skill in the art. Summary of the invention
[0006] The content of this disclosure is used to introduce concepts in a brief form, which will be described in detail in the detailed implementation section below. The content of this disclosure is not intended to identify the key features or essential features of the technical solution claimed for protection, nor is it intended to limit the scope of the technical solution claimed for protection.
[0007] Some embodiments of the present disclosure propose a method, device, electronic device and computer-readable medium for generating distribution network vibration source type information to solve one or more of the technical problems mentioned in the above background technology section.
[0008] In a first aspect, some embodiments of the present disclosure provide a method for generating distribution network vibration source type information, the method comprising: obtaining each vibration signal information of the target distribution network, wherein the above-mentioned each vibration signal information is collected in real time through a vibration optical fiber sensor network; performing signal conversion processing on the above-mentioned each vibration signal information to obtain each discrete vibration signal information; preprocessing the above-mentioned each discrete vibration signal information to obtain each discrete vibration signal information after preprocessing; performing feature extraction processing on each discrete vibration signal information after preprocessing to obtain each vibration signal feature information; performing vibration source classification processing on the above-mentioned each vibration signal information according to the above-mentioned each vibration signal feature information to obtain each vibration source classification information corresponding to the above-mentioned each vibration signal information; performing vibration source positioning processing on the above-mentioned each vibration signal information to obtain each vibration source position information corresponding to the above-mentioned each vibration signal information; for each vibration source classification information in the above-mentioned each vibration source classification information, generating distribution network vibration source type information according to the vibration source classification information and the vibration source position information corresponding to the above-mentioned each vibration source classification information in the above-mentioned each vibration source position information.
[0009] In a second aspect, some embodiments of the present disclosure provide a distribution network vibration source type information generating device, the device comprising: an acquisition unit, configured to acquire each vibration signal information of a target distribution network, wherein each vibration signal information is collected in real time through a vibration optical fiber sensor network; a conversion unit, configured to perform signal conversion processing on each vibration signal information to obtain each discrete vibration signal information; a preprocessing unit, configured to preprocess each discrete vibration signal information to obtain each discrete vibration signal information after preprocessing; an extraction unit, configured to perform feature extraction processing on each discrete vibration signal information after preprocessing. Obtain characteristic information of each vibration signal; a classification unit is configured to perform vibration source classification processing on each vibration signal information according to the characteristic information of each vibration signal, and obtain each vibration source classification information corresponding to the each vibration signal information; a positioning unit is configured to perform vibration source positioning processing on each vibration signal information, and obtain each vibration source position information corresponding to the each vibration signal information; a generation unit is configured to generate distribution network vibration source type information for each vibration source classification information in the above vibration source classification information according to the vibration source classification information and the vibration source position information corresponding to the above vibration source classification information in the above vibration source position information.
[0010] In a third aspect, some embodiments of the present disclosure provide an electronic device comprising: one or more processors; a storage device on which one or more programs are stored, and when the one or more programs are executed by the one or more processors, the one or more processors implement the method described in any implementation manner of the above-mentioned first aspect.
[0011] In a fourth aspect, some embodiments of the present disclosure provide a computer-readable medium having a computer program stored thereon, wherein when the computer program is executed by a processor, the method described in any implementation manner of the above-mentioned first aspect is implemented.
[0012] The above-mentioned embodiments of the present disclosure have the following beneficial effects: through the distribution network vibration source type information generation method of some embodiments of the present disclosure, the monitoring efficiency of the distribution network is improved, and the delay in obtaining the distribution network monitoring results is reduced. Specifically, the reason for the low monitoring efficiency of the distribution network and the long delay in obtaining the distribution network monitoring results is that the monitoring system based on the vibration optical fiber will generate a large amount of vibration data, and the efficient and accurate processing and analysis of a large amount of vibration data requires a lot of manpower and a lot of time, resulting in a large delay in obtaining the analysis results. Due to the variety of vibration source types and the strong subjectivity of manual experience, the accuracy of identifying and classifying a variety of vibration source types is low, resulting in low monitoring efficiency of the distribution network. Only by analyzing the vibration data through the frequency domain analysis method, due to the need to perform discrete Fourier transform, there are many cases of vibration data distortion, which leads to low stability in identifying and classifying a variety of vibration source types. Based on this, the distribution network vibration source type information generation method of some embodiments of the present disclosure, first, obtains each vibration signal information of the target distribution network, wherein the above-mentioned each vibration signal information is collected in real time through the vibration optical fiber sensor network. Thus, real-time vibration data of the target distribution network can be obtained. Then, the above-mentioned vibration signal information is subjected to signal conversion processing to obtain each discrete vibration signal information. Thus, each discrete vibration signal information can be obtained. Then, the above-mentioned discrete vibration signal information is preprocessed to obtain each discrete vibration signal information after preprocessing. Thus, each discrete vibration signal information after preprocessing can be obtained, which can improve the accuracy of the vibration signal. Next, feature extraction processing is performed on each vibration signal information after preprocessing to obtain each vibration signal feature information. Thus, each vibration signal feature information corresponding to each vibration signal can be obtained. Then, according to each vibration signal feature information, the above-mentioned vibration signal information is subjected to vibration source classification processing to obtain each vibration source classification information corresponding to each vibration signal information. Thus, each vibration source classification information corresponding to each vibration signal information can be obtained. Then, the above-mentioned vibration signal information is subjected to vibration source positioning processing to obtain each vibration source position information corresponding to each vibration signal information. Thus, each vibration source position information corresponding to each vibration signal information can be obtained, and the vibration source position or fault position can be quickly located. Finally, for each of the vibration source classification information, the distribution network vibration source type information is generated according to the vibration source classification information and the vibration source position information corresponding to the vibration source classification information in the vibration source position information. Thus, the vibration source type information of each distribution network can be obtained.Also, the distribution network vibration source type information generation method implemented is to extract and classify the features of each vibration signal by combining time domain analysis, frequency domain analysis and deep learning methods, instead of only using frequency domain analysis to analyze each vibration signal, which reduces the distortion of each vibration signal analysis and improves the accuracy and stability of feature extraction and classification of vibration signals. And according to the characteristic information of each vibration signal, the vibration source position can be quickly located. If there is a fault, it can be quickly isolated and maintenance personnel can be dispatched for maintenance, thereby improving the monitoring efficiency of the distribution network and reducing the delay in obtaining the distribution network monitoring results. BRIEF DESCRIPTION OF THE DRAWINGS
[0013] The above and other features, advantages and aspects of the embodiments of the present disclosure will become more apparent with reference to the following detailed description in conjunction with the accompanying drawings. Throughout the accompanying drawings, the same or similar reference numerals represent the same or similar elements. It should be understood that the drawings are schematic and that components and elements are not necessarily drawn to scale.
[0014] Figure 1 is a flow chart of some embodiments of the method for generating distribution network vibration source type information according to the present disclosure;
[0015] Figure 2 It is a structural schematic diagram of some embodiments of the distribution network vibration source type information generating device according to the present disclosure;
[0016] Figure 3 It is a schematic diagram of the structure of an electronic device suitable for implementing some embodiments of the present disclosure. DETAILED DESCRIPTION
[0017] Embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. Although certain embodiments of the present disclosure are shown in the accompanying drawings, it should be understood that the present disclosure can be implemented in various forms and should not be construed as being limited to the embodiments set forth herein. On the contrary, these embodiments are provided to provide a more thorough and complete understanding of the present disclosure. It should be understood that the drawings and embodiments of the present disclosure are only for exemplary purposes and are not intended to limit the scope of protection of the present disclosure.
[0018] It should also be noted that, for ease of description, only the parts related to the invention are shown in the drawings. In the absence of conflict, the embodiments and features in the embodiments of the present disclosure can be combined with each other.
[0019] It should be noted that the concepts such as "first" and "second" mentioned in the present disclosure are only used to distinguish different devices, modules or units, and are not used to limit the order or interdependence of the functions performed by these devices, modules or units.
[0020] It should be noted that the modifications of "one" and "plurality" mentioned in the present disclosure are illustrative rather than restrictive, and those skilled in the art should understand that unless otherwise clearly indicated in the context, it should be understood as "one or more".
[0021] The names of the messages or information exchanged between multiple devices in the embodiments of the present disclosure are only used for illustrative purposes and are not used to limit the scope of these messages or information.
[0022] The present disclosure will be described in detail below with reference to the accompanying drawings and in conjunction with embodiments.
[0023] Figure 1 The flowchart 100 of some embodiments of the method for generating distribution network vibration source type information according to the present disclosure is shown. The method for generating distribution network vibration source type information comprises the following steps:
[0024] Step 101, obtaining each vibration signal information of the target distribution network.
[0025] In some embodiments, the execution subject (e.g., a computing device) of the distribution network vibration source type information generation method can obtain various vibration signal information of the target distribution network. Wherein, the above-mentioned various vibration signal information is collected in real time through a vibration optical fiber sensor network. The above-mentioned vibration optical fiber sensor network can represent various vibration optical fiber sensors deployed in the above-mentioned target distribution network environment. The above-mentioned various vibration signal information can be various continuous analog signals. In practice, the above-mentioned execution subject can obtain various vibration signal information of the target distribution network from the vibration optical fiber sensor network.
[0026] Step 102, performing signal conversion processing on each vibration signal information to obtain each discrete vibration signal information.
[0027] In some embodiments, the execution subject may perform signal conversion processing on each of the vibration signal information to obtain each discrete vibration signal information. The discrete vibration signal information may be a discrete electrical signal. The discrete vibration signal information in each of the discrete vibration signal information may include time information, voltage value, and current value. In practice, the execution subject may perform signal conversion processing on each of the vibration signal information in various ways to obtain each discrete vibration signal information.
[0028] In some optional implementations of some embodiments, the execution subject may perform signal conversion processing on each of the vibration signal information to obtain each discrete vibration signal information through the following steps:
[0029] The first step is to convert the vibration signal information into electrical signal information. The electrical signal information may be digital signals. In practice, the execution subject may convert the vibration signal information into electrical signal information through a photoelectric converter.
[0030] In the second step, for each of the above electrical signal information, the following steps are performed:
[0031] First sub-step: sampling the above electrical signal information according to a preset time interval to obtain each sampled signal information. The above sampled signal information can be each electrical signal information segment obtained by equally dividing the above electrical signal information. Here, the specific setting of the above preset time interval is not limited.
[0032] The second sub-step: quantize the above-mentioned individual sampling signal information according to the preset quantization level information sequence, and obtain the individual quantized signal information as discrete vibration signal information. Wherein, the above-mentioned preset quantization level information sequence can be the individual signal thresholds for quantizing the above-mentioned individual sampling signal information. In practice, for each of the above-mentioned individual sampling signal information, in response to determining that the peak value of the electrical signal included in the above-mentioned sampling information is greater than the first preset quantization level information, and less than or equal to the second preset quantization level information, the above-mentioned second preset quantization level information is determined as the quantized signal information corresponding to the above-mentioned sampling signal information. Wherein, the above-mentioned first preset quantization level information and the second preset quantization level information can be any two adjacent preset quantization level information of the above-mentioned preset quantization level information sequence. The above-mentioned second preset quantization level information is greater than the above-mentioned first preset quantization level information.
[0033] In some optional implementations of some embodiments, the execution subject may perform sampling processing on the electrical signal information according to a preset time interval to obtain each sampled signal information through the following steps:
[0034] The first step is to determine the signal start time corresponding to the above electrical signal information as the target start time.
[0035] In the second step, based on the target start time, the following sampling steps are performed:
[0036] The first sampling step is to determine the sum of the target start time and the preset time interval as the sampling time, wherein the sampling time may be the signal end time of the sampled signal information.
[0037] The second sampling step, in response to determining that the sampling time is less than the target end time, determines the electrical signal information between the target start time and the sampling time in the electrical signal information as the sampling signal information, and takes the sampling time as the target start time, and executes the sampling step again. The target time can be the signal end time corresponding to the electrical signal information plus the preset time interval. It should be noted that the target start time and the target end time can be timestamps in seconds, and the preset time interval can be a time interval in seconds, so the target start time and the target end time can be directly calculated with the preset time interval.
[0038] Step 103, preprocessing each discrete vibration signal information to obtain each discrete vibration signal information after preprocessing.
[0039] In some embodiments, the execution subject may preprocess the discrete vibration signal information to obtain the preprocessed discrete vibration signal information. In practice, the execution subject may preprocess the discrete vibration signal information to obtain the preprocessed discrete vibration signal information in various ways.
[0040] In some optional implementations of some embodiments, the execution subject may preprocess the discrete vibration signal information to obtain the preprocessed discrete vibration signal information through the following steps:
[0041] For each discrete vibration signal information in the above discrete vibration signal information, the following steps are performed:
[0042] The first step is to determine the time corresponding to the first quantized signal information included in the above discrete vibration signal information as the segmentation start time.
[0043] The second step is to perform the following segmentation steps based on the segmentation start time:
[0044] The first sub-step is to determine the sum of the segmentation start time and the preset sliding window information as the segmentation time, wherein the preset sliding window information may represent a time interval.
[0045] The second sub-step, in response to determining that the above-mentioned split time satisfies the preset split time condition, determines each sampling signal information between the above-mentioned split start time and the above-mentioned split time in the above-mentioned discrete vibration signal information as sub-discrete vibration signal information, and uses the above-mentioned split time as the split start time, and executes the above-mentioned split step again. Among them, the above-mentioned preset split time condition can be that the above-mentioned split time is less than the split end time. The above-mentioned split end time can be the signal end time corresponding to the above-mentioned discrete vibration signal information plus the time of the above-mentioned preset sliding window information. It should be noted that the above-mentioned split start time and the above-mentioned split end time can be timestamps in seconds, and the above-mentioned preset sliding window information can be a time interval in seconds. Therefore, the above-mentioned split start time and the above-mentioned split end time can be directly calculated with the above-mentioned preset sliding window information.
[0046] The third step is to input each obtained sub-discrete vibration signal information into a pre-trained vibration signal denoising model to obtain each sub-vibration signal information after denoising. The vibration signal denoising model can be a neural network model trained with the vibration signal information as input and the denoised vibration signal information as output. For example, the neural network model can be a CNN model. The neural network model can include at least one convolutional layer, at least one pooling layer, one flattening layer and at least one fully connected layer. The output layer of the model has the same length as the input vibration signal information so that a noise-free signal can be reconstructed.
[0047] The fourth step is to determine each sub-discrete vibration signal information after denoising as the pre-processed discrete vibration signal information.
[0048] Step 104, performing feature extraction processing on each vibration signal information after preprocessing to obtain feature information of each vibration signal.
[0049] In some embodiments, the above-mentioned execution subject may perform feature extraction processing on each vibration signal information after preprocessing to obtain each vibration signal feature information. Among them, the above-mentioned each vibration signal feature information may characterize the characteristics of the above-mentioned each vibration signal information. The vibration signal feature information in the above-mentioned each vibration signal feature information may include various features. The above-mentioned each feature may include but is not limited to multiple of the following: frequency, period, phase, amplitude, each frequency domain component, the intensity corresponding to each frequency domain component, and abstract feature information. The above-mentioned abstract feature information may be a feature vector extracted by a neural network model. In practice, the above-mentioned execution subject may adopt various methods to perform feature extraction processing on each vibration signal information after preprocessing to obtain each vibration signal feature information.
[0050] In some optional implementations of some embodiments, the execution subject may perform feature extraction processing on each preprocessed vibration signal information to obtain each vibration signal feature information through the following steps:
[0051] For each of the vibration signal information above, perform the following steps:
[0052] The first step is to perform time domain analysis on the vibration signal information to obtain time domain analysis result information. The time domain analysis result information can characterize the time domain characteristics of the vibration signal information. The time domain analysis result information can include but is not limited to multiple items of the following: frequency, period, amplitude, and phase. In practice, the execution entity can perform time domain analysis on the vibration signal information through a time domain analysis algorithm to obtain time domain analysis result information. The time domain analysis algorithm can be but is not limited to one of the following: time domain analysis based on wavelet transform, time domain analysis based on empirical mode decomposition.
[0053] The second step is to perform frequency domain analysis on the vibration signal information to obtain frequency domain analysis result information. Among them, the frequency domain analysis result information can characterize the frequency domain characteristics of the vibration signal information. The frequency domain analysis result information can include but is not limited to multiple items of the following: each frequency domain component, and the intensity corresponding to each frequency domain component. In practice, the execution entity can perform frequency domain analysis on the vibration signal information through a frequency domain analysis algorithm to obtain frequency domain analysis result information. The frequency domain analysis algorithm can be but is not limited to one of the following: continuous Fourier transform, discrete Fourier transform, and fast Fourier transform.
[0054] The third step is to input the vibration signal information into a pre-trained vibration signal extraction model to obtain the abstract feature information corresponding to the vibration signal. The vibration signal extraction model can be a neural network model trained with the vibration signal information as input and the abstract feature information corresponding to the vibration signal as the expected output. For example, the neural network model can be a convolutional neural network model or various recurrent neural network models.
[0055] The fourth step is to combine the time domain analysis result information, the frequency domain analysis result information and the abstract feature information into vibration signal feature information. In practice, the execution entity may add the time domain analysis result information, the frequency domain analysis result information and the abstract feature information to a preset feature information set as vibration signal feature information.
[0056] Step 105, performing vibration source classification processing on each vibration signal information according to each vibration signal characteristic information, and obtaining each vibration source classification information corresponding to each vibration signal information.
[0057] In some embodiments, the execution subject may perform vibration source classification processing on each vibration signal information according to the vibration signal characteristic information, and obtain vibration source classification information corresponding to each vibration signal information. Among them, each vibration source classification information may characterize the vibration causes causing each vibration signal information. The vibration source classification information in each vibration source classification information may be, but is not limited to, one of the following: excavation by an excavator, passing of a road roller, passing of a large vehicle, manual excavation, and earthquake. In practice, the execution subject may perform vibration source classification processing on each vibration signal information according to the vibration signal characteristic information in various ways, and obtain vibration source classification information corresponding to each vibration signal information.
[0058] In some optional implementations of some embodiments, the execution subject may perform vibration source classification processing on each vibration signal information according to each vibration signal feature information through the following steps to obtain each vibration source classification information corresponding to each vibration signal information:
[0059] For each of the vibration signal information above, perform the following steps:
[0060] The first step is to perform feature screening on the vibration signal feature information to obtain the filtered vibration signal feature information. In practice, the execution subject may determine each feature in the vibration signal feature information corresponding to the input of the pre-trained vibration signal classification model as the filtered vibration signal feature information.
[0061] In the second step, the filtered vibration signal feature information is input into a pre-trained vibration signal classification model to obtain vibration source classification information. The vibration signal classification model can characterize the correspondence between the vibration signal feature information and the vibration source classification information. The vibration signal classification model can be a neural network model trained with the vibration signal feature information as input and the vibration source classification information corresponding to the vibration signal feature information as the expected output. For example, the neural network model can be a CNN model.
[0062] Next, when the vibration signal classification model disclosed in the present invention is used to classify vibration signal information, the following technical problems further exist:
[0063] Since there may be redundant vibration signal features in each vibration signal feature, selecting all vibration signal features to train the vibration signal classification model will lead to overfitting and weak generalization ability of the model. Moreover, when all vibration signal features are involved in model training, the complexity of the model increases and the convergence speed of the model is slow.
[0064] In some optional implementations of some embodiments, the above-mentioned execution subject may train and obtain the vibration signal classification model through the following steps:
[0065] The first step is to obtain a historical vibration signal data set. The historical vibration signal data set may be each discrete vibration signal information of the target distribution network recorded in history. The historical vibration signal data in the historical vibration signal data set may include historical discrete vibration signal information and vibration source classification labels. The historical discrete vibration signal information may be discrete vibration signal information of the target distribution network recorded in history. The vibration source classification label may characterize the vibration source classification information corresponding to the historical discrete vibration signal information. The vibration source classification label may be, but not limited to, one of the following: excavator excavation, road roller passing, large vehicle passing, manual excavation, earthquake. In practice, the execution subject may obtain the historical vibration signal data set from an associated storage device. The storage device may be, but not limited to, one of the following: a database, a disk, a mobile hard disk.
[0066] In the second step, for each historical vibration signal data in the above historical vibration signal data set, the historical discrete vibration signal information included in the above historical vibration signal data is subjected to initial feature extraction to obtain each initial feature information corresponding to the above historical vibration signal data. Among them, the initial feature information in each initial feature information may correspond to the initial feature field in each initial feature field one by one. The above initial feature information may be the vibration signal feature information of the above historical vibration signal data. The above initial feature information may include but is not limited to multiple items of the following: frequency, period, phase, amplitude, each frequency domain component, the intensity corresponding to each frequency domain component, and abstract feature information. The above initial feature fields may be the names of each field corresponding to the above initial feature information. For example, each initial feature field may include but is not limited to multiple items of the following: frequency field, period field, phase field, amplitude field, each frequency domain component field, the intensity field corresponding to each frequency domain component, and abstract feature information field. Here, for each historical vibration signal data in the above historical vibration signal data set, the detailed steps of performing initial feature extraction on the discrete vibration signal information included in the above historical vibration signal data can refer to step 104, which will not be repeated here.
[0067] In the third step, for each historical vibration signal data in the above historical vibration signal data set, the following steps are performed:
[0068] In the first sub-step, each initial feature information of the above historical vibration signal data is determined as sample feature information.
[0069] The second sub-step is to determine the vibration source classification label included in the above historical vibration signal data as the sample label.
[0070] The third sub-step is to determine the sample feature information and the sample label as a sample.
[0071] The fourth step is to determine the determined samples as a sample set.
[0072] Step 5: Based on the sample set and the initial neural network, perform the following training steps:
[0073] In the first training step, the sample feature information of at least one sample in the sample set is input into the initial neural network model to obtain a predicted label corresponding to the at least one sample. The initial state of the initial neural network may include initial weights corresponding to the initial feature information. The predicted label of the at least one sample may be the vibration source classification information corresponding to the sample feature information of the at least one sample predicted by the initial neural network model.
[0074] The second training step is to determine the error information between the predicted label corresponding to the at least one sample and the sample label of the at least one sample. The error information can characterize the degree of difference between the predicted label of the at least one sample and the sample label of the at least one sample. In practice, first, for each sample in the at least one sample, the execution subject can determine the cosine similarity between the sample label of the sample and the predicted label of the sample as the similarity information to obtain each similarity information. The value range of each similarity information can be [-1,1]. Then, for each similarity information in the above similarity information, 1 minus the value of the similarity information can be determined as the difference information to obtain each difference information. Afterwards, the average value of each difference information can be determined as the error information between the predicted label corresponding to the at least one sample and the sample label of the at least one sample.
[0075] The third training step is to determine whether the initial neural network has reached the optimization goal based on the error information. The optimization goal may be that the error information is less than a preset error value. The specific setting of the preset error value is not limited here. In practice, in response to determining that the error information is less than the preset error value, the execution subject may determine that the initial neural network has reached the optimization goal. In response to determining that the error information is greater than or equal to the preset error value, the execution subject may determine that the initial neural network has not reached the optimization goal.
[0076] In the fourth training step, in response to determining that the initial neural network does not achieve the above optimization goal, the following steps are performed:
[0077] In the first sub-step, based on the error information, each weight gradient information of each initial feature information in the sample feature information corresponding to the at least one sample is determined as a weight gradient information sequence. The weight gradient information sequence can represent each weight gradient of each initial feature information. In practice, the execution subject can use a back propagation algorithm to determine each weight gradient information of each initial feature information in the sample feature information corresponding to the at least one sample based on the error information.
[0078] The second sub-step is to update the initial weights corresponding to the initial feature information in the initial neural network and the network parameters of the initial neural network according to the weight gradient information of the initial feature information in the sample feature information corresponding to the at least one sample, and to perform the training step again based on the sample set and the initial neural network using a sample set composed of unused samples. In practice, the execution subject may use a gradient descent method or an Adam algorithm to update the initial weights corresponding to the initial feature information in the initial neural network and the network parameters of the initial neural network.
[0079] The fifth training step, in response to determining that the initial neural network has achieved the above-mentioned optimization goal, determines the importance scores of each vibration signal classification corresponding to each initial feature field according to the obtained weight gradient information sequences. Among them, the above-mentioned importance scores of each vibration signal classification can characterize the degree of influence of each initial feature field on the model output. In practice, first, for each of the above-mentioned initial feature fields, the above-mentioned execution entity can determine the average value of each weight gradient information corresponding to the above-mentioned initial feature field in the above-mentioned weight gradient information sequences as the average gradient information, and obtain each average gradient information. Then, the above-mentioned each average gradient information can be determined as the importance score of each vibration signal classification corresponding to the above-mentioned each initial feature field.
[0080] In the sixth step, according to the importance scores of the above-mentioned vibration signal classifications, feature selection processing is performed on the above-mentioned initial feature fields to obtain the target feature fields. Among them, the above-mentioned target feature fields can be the initial feature fields that contribute more to the output of the initial neural network model. In practice, the above-mentioned execution entity can determine the initial feature fields whose vibration signal classification importance scores in the above-mentioned initial feature fields are greater than the preset importance score threshold as the target feature fields. Here, the specific setting of the above-mentioned preset importance score threshold is not limited.
[0081] Step 7: Generate a target sample set based on the above target feature fields and the above historical vibration signal data set. The above target sample set may be samples of each initial feature information corresponding to the above target feature fields. In practice, first, for each historical vibration signal data in the above historical vibration signal data set, the execution subject may perform the following steps:
[0082] In the first sub-step, each initial feature information corresponding to each target feature field in each initial feature information of the historical vibration signal data is determined as target sample feature information.
[0083] The second sub-step is to determine the vibration source classification label included in the above historical vibration signal data as the target sample label.
[0084] The third sub-step is to determine the target sample feature information and the target sample label as the target sample.
[0085] Then, the determined target samples may be determined as a target sample set.
[0086] In the eighth step, each target sample in the target sample set is used as input, and each target sample label corresponding to each target sample is used as the expected output to train a vibration signal classification model. In practice, the execution subject can train the vibration signal classification model through back propagation and gradient descent algorithms.
[0087] The first to eighth steps of the above technical solution and their related contents serve as an inventive point of the embodiments of the present disclosure, and solve the second technical problem mentioned in the background technology: "Since there may be redundant vibration signal features in each vibration signal feature, all vibration signal features are selected to train the vibration signal classification model, resulting in overfitting and weak generalization ability of the model, and when all vibration signal features are involved in model training, the complexity of the model increases, and the convergence speed of the model is slow." The factors that lead to overfitting and weak generalization ability of the model, large complexity of the model, and slow convergence speed of the model are often as follows: Since there may be redundant vibration signal features in each vibration signal feature, all vibration signal features are selected to train the vibration signal classification model, resulting in overfitting and weak generalization ability of the model, and when all vibration signal features are involved in model training, the complexity of the model increases, and the convergence speed of the model is slow. If the above factors are solved, the effect of improving the generalization ability and convergence speed of the model and reducing the complexity and overfitting of the model can be achieved. In order to achieve this effect, the present disclosure introduces a method of training an initial classification model using all vibration signal features and recording the weight gradient changes of all vibration signal features during the initial classification model training process, thereby removing irrelevant or redundant features and retaining the features that contribute most to the model prediction results. The vibration signal classification model is retrained based on the retained features, thereby reducing the overfitting of the model and improving the generalization ability of the model. The reduction of features can reduce the complexity of the model and improve the convergence speed of the model.
[0088] Step 106, performing vibration source positioning processing on each vibration signal information to obtain each vibration source position information corresponding to each vibration signal information.
[0089] In some embodiments, the execution subject may perform source location processing on each of the vibration signal information to obtain source location information corresponding to each of the vibration signal information. The source location information in each of the vibration source location information may characterize the distribution network location where vibration occurs. In practice, the execution subject may use a time domain analysis method to perform source location processing on each of the vibration signal information to obtain source location information corresponding to each of the vibration signal information.
[0090] Next, when the vibration source positioning process of each of the above-mentioned vibration signal information is performed using the present disclosure, the following technical problems further exist:
[0091] The time domain analysis method has a weak anti-interference ability. When there is a lot of noise and interference in the vibration signal data, the positioning accuracy is reduced. For complex vibration signals, the time domain analysis method may find it difficult to accurately extract key features, resulting in low accuracy in vibration source positioning.
[0092] In some optional implementations of some embodiments, the execution subject may perform vibration source positioning processing on each of the vibration signal information to obtain each vibration source position information corresponding to each of the vibration signal information through the following steps:
[0093] The first step is to obtain the time interval information of each scattered signal reception and the scattered signal information corresponding to the above-mentioned vibration signal information. The scattered signal information in the above-mentioned scattered signal information includes the phase information corresponding to the above-mentioned scattered signal reception time interval information. In practice, the above-mentioned execution subject can obtain the time interval information of each scattered signal reception and the scattered signal information corresponding to the above-mentioned vibration signal information from the associated phase-sensitive optical time domain reflectometry device.
[0094] In the second step, for each piece of scattered signal information in the above scattered signal information, the following steps are performed:
[0095] In the first sub-step, in response to determining that there is phase information satisfying a preset abnormal condition among the phase information included in the scattered signal information, the scattered signal receiving time interval information corresponding to the phase information satisfying the preset abnormal condition among the phase information is determined as the target time interval information. The preset abnormal condition may be that the difference between the first preset value phase adjacent to the phase information and the second preset value phase adjacent to the phase information is greater than the preset difference. Here, the specific setting of the preset value and the preset difference is not limited.
[0096] The second sub-step is to determine the vibration source distance information according to the above-mentioned target time interval information and the preset parameter information. The above-mentioned preset parameter information may be the refractive index of the optical fiber in the above-mentioned vibration optical fiber sensor network. In practice, firstly, the above-mentioned execution subject may determine the quotient of the preset speed and the above-mentioned preset parameter information as the target speed. Then, the result of multiplying the above-mentioned target speed and the above-mentioned target time interval information may be determined as the vibration source distance information. The above-mentioned preset speed may be the propagation speed of light in a vacuum. The above-mentioned target speed may be the propagation speed of light in the optical fiber of the above-mentioned vibration optical fiber sensor network.
[0097] The third sub-step is to determine the above-mentioned vibration source distance information as vibration source position information.
[0098] The first step to the second step of the above technical solution and its related contents serve as an inventive point of the embodiment of the present disclosure, which solves the second technical problem mentioned in the background technology: "The time domain analysis method has weak anti-interference ability. When there are more noise and interference in the vibration signal data, the positioning accuracy is reduced. For complex vibration signals, the time domain analysis method may be difficult to accurately extract key features, resulting in low accuracy in vibration source positioning." The factors that lead to low accuracy in vibration source positioning are often as follows: The time domain analysis method has weak anti-interference ability. When there are more noise and interference in the vibration signal data, the positioning accuracy is reduced. For complex vibration signals, the time domain analysis method may be difficult to accurately extract key features. If the above factors are solved, the effect of improving the accuracy of vibration source positioning can be achieved. In order to achieve this effect, the present disclosure introduces phase-sensitive optical time domain reflection technology, which uses backscattered light in optical fiber to detect vibration signals, and determines the position of the vibration source by measuring the phase change of light. The sensitivity and resolution of vibration signal detection are high. Since the phase information of light is used for positioning, the position of the vibration signal can be more accurately identified, thereby improving the accuracy of vibration source positioning.
[0099] Step 107: for each vibration source classification information in each vibration source classification information, generate distribution network vibration source type information according to the vibration source classification information and the vibration source position information corresponding to the vibration source classification information in each vibration source position information.
[0100] In some embodiments, the execution subject may generate distribution network vibration source type information for each of the vibration source classification information, according to the vibration source classification information and the vibration source location information corresponding to the vibration source classification information in the vibration source location information. The distribution network vibration source type information may characterize the cause and location of the target distribution network vibration. In practice, for each of the vibration source classification information, the execution subject may determine the vibration source classification information and the vibration source location information corresponding to the vibration source classification information in the vibration source location information as the distribution network vibration source type information.
[0101] Optionally, first, the execution subject may also determine whether the generated distribution network vibration source type information satisfies a preset warning condition. The preset warning condition may be that the vibration source classification information included in the distribution network vibration source type information is earthquake or excavation by an excavator.
[0102] Then, in response to determining that there is distribution network vibration source type information that meets the preset warning condition among the above-mentioned distribution network vibration source type information, each distribution network vibration source type information that meets the above-mentioned preset warning condition among the above-mentioned distribution network vibration source type information is determined as each abnormal distribution network vibration source type information. Among them, the above-mentioned each abnormal distribution network vibration source type information can be each distribution network vibration source type information whose vibration source classification information is earthquake or excavation by excavator.
[0103] Afterwards, each early warning prompt information corresponding to the above-mentioned each abnormal distribution network vibration source type information is sent to the associated terminal device. The above-mentioned terminal device may be, but not limited to, a computer or a mobile phone. The above-mentioned each early warning prompt information may be prompt information that there is an earthquake or excavation by an excavator at each target location. The above-mentioned each target location may be each vibration source location information included in the above-mentioned each abnormal distribution network vibration source type information. In practice, the above-mentioned execution entity may send each early warning prompt information corresponding to the above-mentioned each abnormal distribution network vibration source type information to the associated terminal device through a communication protocol. The above-mentioned communication protocol may be, but not limited to, one of the following: HTTPS protocol, FTP protocol, WebSocket protocol, TCP / IP protocol family.
[0104] Finally, based on the topological structure information corresponding to the above-mentioned target distribution network, the above-mentioned vibration source type information of each abnormal distribution network is visualized. In practice, the above-mentioned execution entity can highlight the vibration source type information of each abnormal distribution network at the corresponding vibration source position in the topological structure information of the above-mentioned target distribution network. The above-mentioned topological structure information can be a topological map of the above-mentioned target distribution network. The above-mentioned vibration source positions can be the various positions of the vibration source position information included in the vibration source type information of each abnormal distribution network in the topological map of the above-mentioned target distribution network. The above-mentioned highlighting can be a dynamic flashing display.
[0105] The above-mentioned embodiments of the present disclosure have the following beneficial effects: through the distribution network vibration source type information generation method of some embodiments of the present disclosure, the monitoring efficiency of the distribution network is improved, and the delay in obtaining the distribution network monitoring results is reduced. Specifically, the reason for the low monitoring efficiency of the distribution network and the long delay in obtaining the distribution network monitoring results is that the monitoring system based on the vibration optical fiber will generate a large amount of vibration data, and the efficient and accurate processing and analysis of a large amount of vibration data requires a lot of manpower and a lot of time, resulting in a large delay in obtaining the analysis results. Due to the variety of vibration source types and the strong subjectivity of manual experience, the accuracy of identifying and classifying a variety of vibration source types is low, resulting in low monitoring efficiency of the distribution network. Only by analyzing the vibration data through the frequency domain analysis method, due to the need to perform discrete Fourier transform, there are many cases of vibration data distortion, which leads to low stability in identifying and classifying a variety of vibration source types. Based on this, the distribution network vibration source type information generation method of some embodiments of the present disclosure, first, obtains each vibration signal information of the target distribution network, wherein the above-mentioned each vibration signal information is collected in real time through the vibration optical fiber sensor network. Thus, real-time vibration data of the target distribution network can be obtained. Then, the above-mentioned vibration signal information is subjected to signal conversion processing to obtain each discrete vibration signal information. Thus, each discrete vibration signal information can be obtained. Then, the above-mentioned discrete vibration signal information is preprocessed to obtain each discrete vibration signal information after preprocessing. Thus, each discrete vibration signal information after preprocessing can be obtained, which can improve the accuracy of the vibration signal. Next, feature extraction processing is performed on each vibration signal information after preprocessing to obtain each vibration signal feature information. Thus, each vibration signal feature information corresponding to each vibration signal can be obtained. Then, according to each vibration signal feature information, the above-mentioned vibration signal information is subjected to vibration source classification processing to obtain each vibration source classification information corresponding to each vibration signal information. Thus, each vibration source classification information corresponding to each vibration signal information can be obtained. Then, the above-mentioned vibration signal information is subjected to vibration source positioning processing to obtain each vibration source position information corresponding to each vibration signal information. Thus, each vibration source position information corresponding to each vibration signal information can be obtained, and the vibration source position or fault position can be quickly located. Finally, for each of the vibration source classification information, the distribution network vibration source type information is generated according to the vibration source classification information and the vibration source position information corresponding to the vibration source classification information in the vibration source position information. Thus, the vibration source type information of each distribution network can be obtained.Also, the distribution network vibration source type information generation method implemented is to extract and classify the features of each vibration signal by combining time domain analysis, frequency domain analysis and deep learning methods, instead of only using frequency domain analysis to analyze each vibration signal, which reduces the distortion of each vibration signal analysis and improves the accuracy and stability of feature extraction and classification of vibration signals. And according to the characteristic information of each vibration signal, the vibration source position can be quickly located. If there is a fault, it can be quickly isolated and maintenance personnel can be dispatched for maintenance, thereby improving the monitoring efficiency of the distribution network and reducing the delay in obtaining the distribution network monitoring results.
[0106] Further references Figure 2 As an implementation of the methods shown in the above figures, the present disclosure provides some embodiments of a distribution network vibration source type information device, and these device embodiments are Figure 1 Corresponding to the method embodiments shown, the device can be specifically applied to various electronic devices.
[0107] like Figure 2 As shown, the distribution network vibration source type information generating device 200 of some embodiments includes: an acquisition unit 201, a conversion unit 202, a preprocessing unit 203, an extraction unit 204, a classification unit 205, a positioning unit 206 and a generation unit 207. The acquisition unit 201 is configured to acquire each vibration signal information of the target distribution network, wherein the above-mentioned each vibration signal information is collected in real time through a vibration optical fiber sensor network; the conversion unit 202 is configured to perform signal conversion processing on the above-mentioned each vibration signal information to obtain each discrete vibration signal information; the preprocessing unit 203 is configured to preprocess the above-mentioned each discrete vibration signal information to obtain each discrete vibration signal information after preprocessing; the extraction unit 204 is configured to perform feature extraction processing on each vibration signal information after preprocessing to obtain each vibration signal feature information; the classification unit 205 is configured to perform vibration source classification processing on each vibration signal information according to the above-mentioned vibration signal characteristic information, and obtain each vibration source classification information corresponding to the above-mentioned each vibration signal information; the positioning unit 206 is configured to perform vibration source positioning processing on each vibration signal information, and obtain each vibration source position information corresponding to the above-mentioned each vibration signal information; the generating unit 207 is configured to generate distribution network vibration source type information for each vibration source classification information in the above-mentioned each vibration source classification information according to the above-mentioned vibration source classification information and the vibration source position information corresponding to the above-mentioned vibration source classification information in the above-mentioned each vibration source position information.
[0108] It can be understood that the units recorded in the distribution network vibration source type information generating device 200 are similar to the reference Figure 1Therefore, the operations, features and beneficial effects described above for the method are also applicable to the device 200 and the units included therein, and will not be described in detail here.
[0109] Reference below Figure 3 , which shows a structural schematic diagram of an electronic device (eg, a computing device) 300 suitable for implementing some embodiments of the present disclosure. Figure 3 The electronic device shown is only an example and should not bring any limitation to the functions and scope of use of the embodiments of the present disclosure.
[0110] like Figure 3 As shown, the electronic device 300 may include a processing device (e.g., a central processing unit, a graphics processing unit, etc.) 301, which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 302 or a program loaded from a storage device 308 to a random access memory (RAM) 303. In the RAM 303, various programs and data required for the operation of the electronic device 300 are also stored. The processing device 301, the ROM 302, and the RAM 303 are connected to each other via a bus 304. An input / output (I / O) interface 305 is also connected to the bus 304.
[0111] Typically, the following devices may be connected to the I / O interface 305: an input device 306 including, for example, a touch screen, a touch pad, a keyboard, a mouse, a camera, a microphone, an accelerometer, a gyroscope, etc.; an output device 307 including, for example, a liquid crystal display (LCD), a speaker, a vibrator, etc.; a storage device 308 including, for example, a magnetic tape, a hard disk, etc.; and a communication device 309. The communication device 309 may allow the electronic device 300 to communicate with other devices wirelessly or by wire to exchange data. Figure 3 The electronic device 300 is shown with various devices, but it should be understood that it is not required to implement or possess all the devices shown. More or fewer devices may be implemented or possessed instead. Figure 3 Each block shown in the figure may represent one device, or may represent multiple devices as required.
[0112] In particular, according to some embodiments of the present disclosure, the process described above with reference to the flowchart can be implemented as a computer software program. For example, some embodiments of the present disclosure include a computer program product, which includes a computer program carried on a computer-readable medium, and the computer program includes a program code for executing the method shown in the flowchart. In some such embodiments, the computer program can be downloaded and installed from the network through the communication device 309, or installed from the storage device 308, or installed from the ROM 302. When the computer program is executed by the processing device 301, the above-mentioned functions defined in the method of some embodiments of the present disclosure are executed.
[0113] It should be noted that the computer-readable medium recorded in some embodiments of the present disclosure may be a computer-readable signal medium or a computer-readable storage medium or any combination of the above two. The computer-readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or device, or any combination of the above. More specific examples of computer-readable storage media may include, but are not limited to: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In some embodiments of the present disclosure, a computer-readable storage medium may be any tangible medium containing or storing a program, which may be used by or in combination with an instruction execution system, device or device. In some embodiments of the present disclosure, a computer-readable signal medium may include a data signal propagated in a baseband or as part of a carrier wave, which carries a computer-readable program code. Such propagated data signals may take a variety of forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above. Computer readable signal media may also be any computer readable medium other than computer readable storage media, which may send, propagate or transmit a program for use by or in conjunction with an instruction execution system, apparatus or device. The program code contained on the computer readable medium may be transmitted using any appropriate medium, including but not limited to: wires, optical cables, RF (radio frequency), etc., or any suitable combination of the above.
[0114] In some embodiments, the client and the server may communicate using any currently known or future developed network protocol such as HTTP (HyperText Transfer Protocol), and may be interconnected with any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include a local area network ("LAN"), a wide area network ("WAN"), an internet (e.g., the Internet), and a peer-to-peer network (e.g., an ad hoc peer-to-peer network), as well as any currently known or future developed network.
[0115] The computer-readable medium may be included in the electronic device; or it may exist independently without being installed in the electronic device. The computer-readable medium carries one or more programs. When the one or more programs are executed by the electronic device, the electronic device: obtains each vibration signal information of the target distribution network, wherein each vibration signal information is collected in real time through a vibration optical fiber sensor network; performs signal conversion processing on each vibration signal information to obtain each discrete vibration signal information; performs preprocessing on each discrete vibration signal information to obtain each discrete vibration signal information after preprocessing; performs feature extraction processing on each discrete vibration signal information after preprocessing to obtain each vibration signal feature information; performs vibration source classification processing on each vibration signal information according to each vibration signal feature information to obtain each vibration source classification information corresponding to each vibration signal information; performs vibration source positioning processing on each vibration signal information to obtain each vibration source position information corresponding to each vibration signal information; for each vibration source classification information in each vibration source classification information, generates distribution network vibration source type information according to the vibration source classification information and the vibration source position information corresponding to the vibration source classification information in the vibration source position information.
[0116] Computer program code for performing the operations of some embodiments of the present disclosure may be written in one or more programming languages or a combination thereof, including object-oriented programming languages such as Java, Smalltalk, C++, and conventional procedural programming languages such as "C" or similar programming languages. The program code may be executed entirely on the user's computer, partially on the user's computer, as a separate software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (e.g., via the Internet using an Internet service provider).
[0117] The flow chart and block diagram in the accompanying drawings illustrate the possible architecture, function and operation of the system, method and computer program product according to various embodiments of the present disclosure. In this regard, each square box in the flow chart or block diagram can represent a module, a program segment or a part of a code, and the module, the program segment or a part of the code contains one or more executable instructions for realizing the specified logical function. It should also be noted that in some implementations as replacements, the functions marked in the square box can also occur in a sequence different from that marked in the accompanying drawings. For example, two square boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each square box in the block diagram and / or flow chart, and the combination of the square boxes in the block diagram and / or flow chart can be implemented with a dedicated hardware-based system that performs a specified function or operation, or can be implemented with a combination of dedicated hardware and computer instructions.
[0118] The units described in some embodiments of the present disclosure may be implemented by software or by hardware. The described units may also be set in a processor, for example, it may be described as: a processor includes an acquisition unit, a conversion unit, a preprocessing unit, an extraction unit, a classification unit, a positioning unit and a generation unit. The names of these units do not constitute a limitation on the units themselves in some cases. For example, the acquisition unit may also be described as a "unit for acquiring information on various vibration signals of the target distribution network."
[0119] The functions described above herein may be performed at least in part by one or more hardware logic components. For example, without limitation, exemplary types of hardware logic components that may be used include: field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on chips (SOCs), complex programmable logic devices (CPLDs), and the like.
[0120] The above descriptions are only some preferred embodiments of the present disclosure and an explanation of the technical principles used. Those skilled in the art should understand that the scope of the invention involved in the embodiments of the present disclosure is not limited to the technical solutions formed by a specific combination of the above technical features, but should also cover other technical solutions formed by any combination of the above technical features or their equivalent features without departing from the above inventive concept. For example, the above features are replaced with (but not limited to) technical features with similar functions disclosed in the embodiments of the present disclosure.
Claims
1. A method for generating distribution network vibration source type information, comprising: Acquire each vibration signal information of the target distribution network, wherein each vibration signal information is collected in real time through a vibration optical fiber sensor network; Performing signal conversion processing on each vibration signal information to obtain each discrete vibration signal information; Preprocessing each discrete vibration signal information to obtain each discrete vibration signal information after preprocessing, wherein preprocessing each discrete vibration signal information to obtain each discrete vibration signal information after preprocessing includes: For each discrete vibration signal information in the discrete vibration signal information, the following steps are performed: Determine the time corresponding to the first quantized signal information included in the discrete vibration signal information as the segmentation start time; Based on the split start time, perform the following split steps: The sum of the split start time and the preset sliding window information is determined as the split time; In response to determining that the segmentation time satisfies a preset segmentation time condition, each sampled signal information between the segmentation start time and the segmentation time in the discrete vibration signal information is determined as sub-discrete vibration signal information, and the segmentation step is performed again with the segmentation time as the segmentation start time; Inputting the obtained sub-discrete vibration signal information into a pre-trained vibration signal denoising model to obtain denoised sub-discrete vibration signal information; Determine each sub-discrete vibration signal information after denoising as pre-processed discrete vibration signal information; Performing feature extraction processing on each discrete vibration signal information after preprocessing to obtain feature information of each vibration signal; According to the characteristic information of each vibration signal, the vibration source classification processing is performed on the information of each vibration signal to obtain the vibration source classification information corresponding to the information of each vibration signal, wherein, according to the characteristic information of each vibration signal, the vibration source classification processing is performed on the information of each vibration signal to obtain the vibration source classification information corresponding to the information of each vibration signal, including: For each vibration signal information in the vibration signal information, the following steps are performed: Performing feature screening processing on the vibration signal feature information of the vibration signal information to obtain screened vibration signal feature information; Inputting the filtered vibration signal feature information into a pre-trained vibration signal classification model to obtain vibration source classification information; The vibration signal classification model is trained by the following steps: Acquire a historical vibration signal data set, wherein the historical vibration signal data in the historical vibration signal data set includes historical discrete vibration signal information and vibration source classification labels; For each historical vibration signal data in the historical vibration signal data set, initial feature extraction is performed on the historical discrete vibration signal information included in the historical vibration signal data to obtain each initial feature information corresponding to the historical vibration signal data, wherein the initial feature information in each initial feature information corresponds to the initial feature field in each initial feature field in a one-to-one manner; For each historical vibration signal data in the historical vibration signal data set, the following steps are performed: Determining each initial feature information of the historical vibration signal data as sample feature information; Determine the vibration source classification label included in the historical vibration signal data as a sample label; Determine the sample feature information and the sample label as a sample; Determine the determined samples as a sample set; Based on the sample set and the initial neural network, the following training steps are performed: Inputting sample feature information of at least one sample in the sample set into an initial neural network model to obtain a prediction label corresponding to the at least one sample, wherein an initial state of the initial neural network includes initial weights corresponding to respective initial feature information; Determining error information between a predicted label corresponding to the at least one sample and a sample label of the at least one sample; Determining whether the initial neural network achieves the optimization goal according to the error information; In response to determining that the initial neural network does not achieve the optimization goal, performing the following steps: Determine, according to the error information, each piece of weight gradient information corresponding to each piece of initial feature information in the sample feature information of the at least one sample as a weight gradient information sequence; According to each weight gradient information of each initial feature information in the sample feature information corresponding to the at least one sample, each initial weight corresponding to each initial feature information in the initial neural network and the network parameters of the initial neural network are updated, and based on the sample set composed of unused samples, the training step is performed again based on the sample set and the initial neural network; In response to determining that the initial neural network achieves the optimization goal, determining the importance scores of the vibration signal classifications corresponding to the initial feature fields according to the obtained weight gradient information sequences; According to the classification importance scores of the vibration signals, feature selection processing is performed on the initial feature fields to obtain target feature fields; Generate a target sample set according to each target feature field and the historical vibration signal data set; Taking each target sample in the target sample set as input and each target sample label included in each target sample as expected output, training a vibration signal classification model; Performing vibration source positioning processing on each vibration signal information to obtain each vibration source position information corresponding to each vibration signal information; For each vibration source classification information in the various vibration source classification information, distribution network vibration source type information is generated according to the vibration source classification information and the vibration source position information corresponding to the vibration source classification information in the various vibration source position information.
2. The method according to claim 1, wherein: The method further comprises: Determine whether the generated vibration source type information of each distribution network meets the preset warning conditions; In response to determining that there is distribution network vibration source type information that meets the preset warning condition among the various distribution network vibration source type information, determining each distribution network vibration source type information that meets the preset warning condition among the various distribution network vibration source type information as each abnormal distribution network vibration source type information; Sending each warning prompt information corresponding to each abnormal distribution network vibration source type information to the associated terminal device; Based on the topological structure information corresponding to the target distribution network, the vibration source type information of each abnormal distribution network is visualized.
3. The method according to claim 1, wherein: The performing signal conversion processing on each vibration signal information to obtain each discrete vibration signal information includes: Converting each vibration signal information into each electrical signal information; For each piece of electrical signal information, perform the following steps: Sampling the electrical signal information according to a preset time interval to obtain each sampled signal information; According to each preset quantization level information, each sampling signal information is quantized to obtain each quantized signal information as discrete vibration signal information.
4. The method according to claim 1, wherein: The feature extraction process is performed on each discrete vibration signal information after preprocessing to obtain feature information of each vibration signal, including: For each discrete vibration signal information in the discrete vibration signal information, the following steps are performed: Performing time domain analysis on the discrete vibration signal information to obtain time domain analysis result information; Performing frequency domain analysis on the discrete vibration signal information to obtain frequency domain analysis result information; Inputting the discrete vibration signal into a pre-trained vibration signal extraction model to obtain abstract feature information corresponding to the discrete vibration signal; The time domain analysis result information, the frequency domain analysis result information and the abstract feature information are combined into vibration signal feature information.
5. A distribution network vibration source type information generating device, comprising: An acquisition unit is configured to acquire each vibration signal information of the target distribution network, wherein each vibration signal information is collected in real time through a vibration optical fiber sensor network; a conversion unit, configured to perform signal conversion processing on each vibration signal information to obtain each discrete vibration signal information; The preprocessing unit is configured to preprocess the discrete vibration signal information to obtain the preprocessed discrete vibration signal information, wherein the preprocessing of the discrete vibration signal information to obtain the preprocessed discrete vibration signal information includes: for each discrete vibration signal information in the discrete vibration signal information, performing the following steps: determining the time corresponding to the first quantized signal information included in the discrete vibration signal information as the segmentation start time; based on the segmentation start time, performing the following segmentation step: determining the sum of the segmentation start time and the preset sliding window information as the segmentation time; in response to determining that the segmentation time satisfies the preset segmentation time condition, determining each sampling signal information between the segmentation start time and the segmentation time in the discrete vibration signal information as sub-discrete vibration signal information, and performing the segmentation step again with the segmentation time as the segmentation start time; inputting the obtained sub-discrete vibration signal information into a pre-trained vibration signal denoising model to obtain the denoised sub-discrete vibration signal information; and determining the denoised sub-discrete vibration signal information as the preprocessed discrete vibration signal information; An extraction unit is configured to perform feature extraction processing on each discrete vibration signal information after preprocessing to obtain feature information of each vibration signal; The classification unit is configured to perform vibration source classification processing on each vibration signal information according to each vibration signal characteristic information to obtain each vibration source classification information corresponding to each vibration signal information, wherein the performing vibration source classification processing on each vibration signal information according to each vibration signal characteristic information to obtain each vibration source classification information corresponding to each vibration signal information includes: for each vibration signal information in each vibration signal information, performing the following steps: performing feature screening processing on the vibration signal characteristic information of the vibration signal information to obtain filtered vibration signal characteristic information; inputting the filtered vibration signal characteristic information into a pre-trained vibration signal classification model to obtain to vibration source classification information; the vibration signal classification model is trained through the following steps: obtaining a historical vibration signal data set, wherein the historical vibration signal data in the historical vibration signal data set includes historical discrete vibration signal information and vibration source classification labels; for each historical vibration signal data in the historical vibration signal data set, performing initial feature extraction on the historical discrete vibration signal information included in the historical vibration signal data to obtain each initial feature information corresponding to the historical vibration signal data, wherein the initial feature information in each initial feature information corresponds one-to-one to the initial feature field in each initial feature field; for each historical vibration signal data in the historical vibration signal data set, performing The following steps are performed: each initial feature information of the historical vibration signal data is determined as sample feature information; the vibration source classification label included in the historical vibration signal data is determined as a sample label; the sample feature information and the sample label are determined as samples; each determined sample is determined as a sample set; based on the sample set and the initial neural network, the following training steps are performed: the sample feature information of at least one sample in the sample set is input into the initial neural network model to obtain a prediction label corresponding to the at least one sample, wherein the initial state of the initial neural network includes each initial weight corresponding to each initial feature information; the prediction label corresponding to the at least one sample and the sample label of the at least one sample are determined; error information between the signatures; determining whether the initial neural network has achieved the optimization target based on the error information; in response to determining that the initial neural network has not achieved the optimization target, performing the following steps: determining, based on the error information, each weight gradient information of each initial feature information in the sample feature information corresponding to the at least one sample as a weight gradient information sequence; updating each initial weight corresponding to each initial feature information in the initial neural network and the network parameters of the initial neural network based on each weight gradient information of each initial feature information in the sample feature information corresponding to the at least one sample, and performing the training step again based on the sample set and the initial neural network using a sample set composed of unused samples;In response to determining that the initial neural network achieves the optimization goal, determining the importance scores of the vibration signal classifications corresponding to the initial feature fields according to the obtained weight gradient information sequences; performing feature selection processing on the initial feature fields according to the vibration signal classification importance scores to obtain target feature fields; generating a target sample set according to the target feature fields and the historical vibration signal data set; taking the target samples in the target sample set as input and the target sample labels corresponding to the target samples as expected outputs, training to obtain a vibration signal classification model; A positioning unit is configured to perform vibration source positioning processing on each vibration signal information to obtain each vibration source position information corresponding to each vibration signal information; The generating unit is configured to generate distribution network vibration source type information for each vibration source classification information according to the vibration source classification information and the vibration source position information corresponding to the vibration source classification information in the vibration source position information.
6. An electronic device comprising: one or more processors; A storage device for storing one or more programs; When the one or more programs are executed by the one or more processors, the one or more processors implement the method according to any one of claims 1 to 4.
7. A computer-readable storage medium having a computer program stored thereon, wherein: When the computer program is executed by a processor, the method according to any one of claims 1 to 4 is implemented.
