Distributed optical fiber sensing cross-domain event classification method based on model migration

Through passive domain adaptation and pseudo-label optimization strategies, combined with the self-supervised comparison learning framework, the problems of data security, pseudo-label noise and high computing costs in distributed fiber sensing cross-domain event classification are solved, achieving high accuracy and Cross-domain event classification with low computational cost.

CN119046754AActive Publication Date: 2024-11-29JIANGSU JIKE HUIZHI TECH CO LTD

Patent Information

Application Number
CN202411133646.7
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-08-19
Publication Date
2024-11-29
Estimated Expiration
2044-08-19

AI Technical Summary

Technical Problem

The existing distributed fiber sensing event classification technology faces data security and privacy issues, pseudo-label noise issues and high computing costs, especially in the cross-domain migration process, it is difficult to effectively solve domain drift and high computing resource requirements.

Method used

Passive domain domain adaptation method and pseudo-label optimization strategy are adopted to realize cross-domain event classification when source data is inaccessible, reducing pseudo-label noise and reducing computing costs through neighboring sample knowledge aggregation and self-supervised comparison learning framework.

Benefits of technology

It improves the accuracy and model robustness of cross-domain event classification, is suitable for application scenarios with high data privacy and security requirements, reduces computing resource requirements, and enhances classification performance and stability.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119046754A_ABST
    Figure CN119046754A_ABST
Patent Text Reader

Abstract

The invention provides a distributed optical fiber sensing cross-domain event classification method based on model migration, and aims to solve the problem that an event classification method based on knowledge migration in the prior art is inconvenient to transmit and share when data of a source domain and data of a target domain are accessed at the same time. The method mainly comprises the steps of constructing a data set, training a source domain model, training a target domain model, deploying the target domain model and the like. According to the method, cross-domain event classification under the condition that source data cannot be accessed is realized by introducing a passive domain adaptation method; through knowledge aggregation and pseudo tag optimization of neighborhood samples, noise in pseudo tags can be effectively reduced, and more accurate pseudo tags are generated, so that events are classified more accurately; meanwhile, compared with an existing method depending on a complex generative model, the method is further based on neighborhood sample knowledge aggregation and a self-supervised contrast learning framework, a pseudo-label optimization strategy with low calculation cost is achieved, the requirement for calculation resources is reduced, and the method is more suitable for large-scale data processing and real-time monitoring in actual application.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of signal detection, and in particular to a distributed optical fiber sensing cross-domain event classification method based on model migration. Background Art

[0002] In terms of event classification, distributed fiber optic sensing distinguishes different types of events by detecting and identifying features in sensing signals, such as pipeline leak detection, earthquake monitoring, border security, tunnel and bridge structural health monitoring, etc. This requires analysis and processing of a large amount of sensing data, and machine learning and deep learning algorithms are usually used to automate this process.

[0003] The existing distributed fiber optic sensing event classification technology mainly relies on supervised machine learning methods, which require a large amount of labeled training data. In practical applications, obtaining sufficient and accurate labeled data is time-consuming and expensive. In particular, distributed fiber optic sensing systems are usually deployed in different physical environments, which leads to differences in data distribution, the so-called domain shift problem. In order to solve the domain shift problem, researchers proposed domain adaptation, which is a novel method for classifying distributed fiber optic sensing events. The DA technology attempts to solve this problem by transferring knowledge between the source domain (labeled data) and the target domain (unlabeled data). Although this method improves the effect of cross-domain migration to a certain extent, there are still at least the following problems: 1) A major data security issue: Many existing cross-domain event classification methods require access to both source and target domain data. In practical applications, source domain data often involves privacy and security issues, which are not convenient for transmission and sharing. 2) It is also found that there is a pseudo-label noise problem: unsupervised domain adaptation methods usually rely on pseudo-labels to label target domain data. However, there is inevitably noise in pseudo-labels, which will cause the model to be disturbed during training, thereby affecting the classification performance. 3) It is also found that there is a problem of high computational cost: although some methods based on generative models perform well in cross-domain migration, their computational complexity is high, and the training process requires a lot of computing resources, which makes it difficult to promote in practical applications. Therefore, in response to the above pain points, it is necessary to propose further improvement plans for the distributed fiber optic sensing cross-domain event classification method based on knowledge transfer. Summary of the invention

[0004] In response to the shortcomings of the above-mentioned prior art, the present application provides a distributed fiber optic sensing cross-domain event classification method based on model migration. By introducing a passive domain adaptation method, cross-domain event classification is achieved when the source data is inaccessible. At the same time, through the knowledge aggregation and pseudo-label optimization of neighborhood samples, the noise in the pseudo-label can be reduced to more accurately classify events, and a pseudo-label optimization strategy with low computational cost is implemented, reducing the demand for computing resources.

[0005] The technical solution adopted by the present invention is as follows: A distributed optical fiber sensing cross-domain event classification method based on model migration comprises the following steps: S1. Distributed optical fiber sensors are arranged in the monitoring area to construct a labeled source domain vibration event dataset and an unlabeled target domain vibration event dataset; S2, train a source domain model using a labeled source domain vibration dataset; S3, transfer the knowledge of the source domain model to the target domain, and train the target domain model based only on the source domain model and the unlabeled target domain vibration event dataset; S4. Deploy the trained target domain model into the distributed fiber optic sensing system to perform real-time event classification on new data in the target domain.

[0006] Furthermore, the method for training the target domain model in step S3 includes the following steps: A1. Build pseudo labels and optimize them; A2, measure the uncertainty of the pseudo-label and calculate the corresponding weight; A3. In order to reduce the negative impact of incorrect labels on model training, a self-training loss function for target domain data is constructed; A4. Use multi-cycle historical pseudo-label information to exclude negative sample pairs that may belong to the same category; A5. Promote the diversity of classifier predictions through regularization methods and construct an overall loss function.

[0007] Furthermore, in step A1, the trained source domain model is used to generate preliminary pseudo labels for the unlabeled data in the target domain. Based on the neighborhood knowledge aggregation strategy, the pseudo labels are optimized by integrating the information of similar samples in the target domain. For each sample in the target domain, the cosine distance between samples is calculated to find K similar samples in the feature space, and the average score is calculated by the following soft voting strategy calculation formula: ;

[0008] represents the nearest neighbor sample set, represents the probability output, and the superscript c indicates that the averaging operation is performed by class; Use the argmax operation to get the corrected pseudo-label: .

[0009] Furthermore, in step A2, the uncertainty estimation method based on the consistency of neighborhood prediction measures the uncertainty of the pseudo-label by calculating the entropy value of the prediction results of the neighborhood samples, and reweights the classification loss according to the uncertainty. The pseudo-labels with low uncertainty (high confidence) are given higher weights, and the pseudo-labels with high uncertainty (low confidence) are given lower weights. The steps are as follows: For a target domain sample , calculate its average prediction score according to formula (1), which is defined as a probability distribution, and calculate the entropy value of the probability distribution: ;

[0010] Where C is the number of categories, and the entropy value is scaled according to the maximum value. The scaling formula is as follows: ;

[0011] The weight of the sample is calculated by the normalized entropy value. The calculation formula is as follows: .

[0012] Furthermore, in step A3, a negative learning strategy is introduced into the classification loss, that is, when calculating the classification loss, a complementary label is selected instead of the correct label, and the self-training loss function of the target domain data is obtained as follows: ;

[0013] in is a label randomly selected from a label set that does not contain the pseudo label of the current sample, that is, , It is a strongly enhanced image The probability output is The above-mentioned estimate The weight of uncertainty.

[0014] Furthermore, in step A4, the target domain feature space is regularized using a contrastive learning framework to make the positive sample pairs in the feature space close and the negative sample pairs far apart. The prediction results of the past multiple cycles are stored by introducing a time queue, and the negative sample pairs that are marked as the same class in the historical prediction are excluded. The steps are as follows: the pseudo labels of each sample in the past T cycles are stored in the queue, and then all sample pairs that share the same pseudo label at least once in the past T cycles are excluded from the negative sample pair list. Accordingly, the following loss function is optimized: ;

[0015] in is the index set of samples in the storage queue that have never shared the same pseudo-label with the query sample in the past T cycles, q is the current query value, k+ is the positive sample key value, kj is the negative sample key value, is the temperature coefficient.

[0016] Furthermore, in step A5, in order to avoid posterior collapse, that is, the classifier tends to predict all samples as the same class, the following regularization method is used to promote the diversity of classifier predictions: ; And construct the overall loss function: .

[0017] Furthermore, in step S2, the source domain model is a deep neural network including a feature extractor and classifier Two parts.

[0018] Furthermore, in step S1, after the distributed optical fiber sensor collects environmental vibration data, the original vibration data is first filtered and denoised for data preprocessing, and then labeled source domain vibration event data and unlabeled target domain vibration event data are constructed.

[0019] The beneficial effects of the present invention are as follows: 1) Compared with the existing technology, the accuracy of cross-domain event classification is improved; through the knowledge aggregation of neighborhood samples and pseudo-label optimization, the method proposed in the present invention can effectively reduce the noise in pseudo-labels and generate more accurate pseudo-labels. Combined with the self-supervised contrastive learning framework, this method can more accurately classify events in the target domain, improving the overall performance of cross-domain event classification; 2) Compared with the existing technology, the model migration is realized when the source data is inaccessible; the present invention realizes cross-domain event classification when the source data is inaccessible by introducing a passive domain adaptation method. This method is particularly suitable for application scenarios with high data privacy and security requirements, such as underground pipeline intrusion detection and urban geological disaster monitoring; 3) Compared with the existing technology, the computational cost is reduced; compared with the existing methods that rely on complex generative models, the present invention realizes a pseudo-label optimization strategy with low computational cost through neighborhood sample knowledge aggregation and self-supervised contrastive learning framework. This method reduces the demand for computing resources and is more suitable for large-scale data processing and real-time monitoring in practical applications; 4) Compared with the existing technology, the robustness of the model is enhanced; by estimating the uncertainty of the pseudo-label and performing loss reweighting, the method of the present invention can effectively deal with the noise in the pseudo-label and improve the robustness of the model. At the same time, the negative sample exclusion strategy further reduces the negative impact of pseudo-label noise during the training process, making the model more stable in different application scenarios. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] Figure 1 A flow chart of event classification for a distributed optical fiber sensing cross-domain event classification method based on model migration according to the present invention; Figure 2 This is a flowchart of the target domain model training in the present invention. DETAILED DESCRIPTION

[0021] The specific implementation of the present invention will be described below in conjunction with the accompanying drawings.

[0022] The present invention discloses a distributed optical fiber sensing cross-domain event classification method based on model migration, aiming to solve the problems in the prior art of event classification methods based on knowledge migration that are inconvenient to transmit and share when simultaneously accessing source domain and target domain data, as well as the problems of pseudo-label noise and high computational cost of generating models.

[0023] The present invention realizes efficient classification of vibration events in the target domain by training the model in the source domain and migrating it to the target domain. The overall flow chart of the present invention is as follows: Figure 1 shown.

[0024] The main steps of this method are as follows: Specific implementation of step S1, data set construction: Firstly, distributed fiber optic sensors are arranged in the monitoring area to collect environmental vibration data and perform data preprocessing such as filtering and denoising on the original vibration data, so as to construct a labeled source domain vibration event dataset and an unlabeled target domain vibration event dataset.

[0025] Specific implementation of step S2: source domain model training: Use the labeled source domain dataset to train a source domain model, which is a deep neural network including a feature extractor and classifier Two parts.

[0026] Specific implementation of step S3, target domain model training: The knowledge of the source domain model is transferred to the target domain. Only the source domain model and the unlabeled target domain vibration event data are used to train the target domain model. The flowchart of the target domain model training is shown in Figure 2 .

[0027] Among them, Figure 2 As shown, the implementation method of step A1 pseudo-labeling and optimizing it is: Pseudo-label generation: Use the trained source domain model to generate preliminary pseudo-labels for the unlabeled data in the target domain.

[0028] Pseudo-label optimization: Due to the difference in the distribution of target domain data and source domain data, the initial pseudo-label may contain noise. The neighborhood knowledge aggregation strategy is introduced to optimize the pseudo-label by integrating the information of similar samples in the target domain. Specifically, for each sample in the target domain, the cosine distance between samples is calculated to find the K closest samples in the feature space, and the average score is calculated through the soft voting strategy: ; represents the nearest neighbor sample set, Represents the probability output, and the superscript c indicates that the average operation is performed by class. Next, use the argmax operation to get the corrected pseudo-label: .

[0029] Step A2 measures the uncertainty of the pseudo-label and uses it to calculate the corresponding weight implementation method: In order to improve the reliability of the above pseudo-labels, an uncertainty estimation method based on the consistency of neighborhood predictions is adopted. The uncertainty of the pseudo-label is measured by calculating the entropy value of the prediction results of the neighborhood samples, and the classification loss is reweighted according to the uncertainty. Pseudo-labels with low uncertainty (high confidence) are given higher weights, and pseudo-labels with high uncertainty (low confidence) are given lower weights, thereby reducing the impact of noise on model training. The specific steps are as follows: For a target domain sample , calculate the average prediction score according to the formula, which can be regarded as a probability distribution. Then, calculate the entropy value of this probability distribution: ; Where C is the number of categories. The entropy value is also scaled by the maximum value as follows: ; Through the normalized entropy value, the weight of the sample can be calculated: .

[0030] Step A3: Implementation of constructing a self-training loss function: In order to reduce the impact of noisy pseudo-labels, a negative learning strategy is introduced in the classification loss. That is, when calculating the classification loss, complementary labels are selected instead of correct labels, thereby reducing the negative impact of incorrect labels on model training. Therefore, the self-training loss function of the target domain data is as follows: ; in is a label randomly selected from a label set that does not contain the pseudo label of the current sample, that is, , It is a strongly enhanced image The probability output is The above-mentioned estimate The weight of uncertainty.

[0031] Step A4, contrastive learning, uses multi-cycle historical pseudo-label information to exclude negative sample pairs that may belong to the same category: The contrastive learning framework is used to regularize the target domain feature space, that is, to make the positive sample pairs in the feature space as close as possible and the negative sample pairs as far apart as possible, so as to ensure that the samples of the same category are clustered together in the feature space; by introducing a time queue to store the prediction results of multiple cycles in the past and excluding those negative sample pairs that are marked as the same category in historical predictions, the accuracy of pseudo labels can be further improved. The specific steps are as follows: Store the pseudo-labels of each sample in the past T cycles in the queue, and then exclude all sample pairs that share the same pseudo-label at least once in the past T cycles from the negative sample pair list. Accordingly, optimize the following loss function: ; in is the index set of samples in the storage queue that have never shared the same pseudo-label with the query sample in the past T cycles, q is the current query value, k+ is the positive sample key value, kj is the negative sample key value, is the temperature coefficient.

[0032] Step A5, constructing an implementation method of the overall loss function: In order to prevent posterior collapse, that is, the classifier tends to predict all samples as the same class, the following regularization method is used to promote the diversity of classifier predictions: ; Self-training, contrastive learning, regularization and other loss functions constitute the overall loss function: .

[0033] Specific implementation of step S4, target domain model deployment: The trained target domain model is deployed in the distributed fiber optic sensing system for real-time event classification. By classifying the newly collected target domain vibration event data, accurate identification and early warning of underground pipeline intrusion, leakage monitoring, road subsidence, geological disasters and other events can be achieved.

[0034] The above description is an explanation of the present invention, not a limitation of the present invention, and any form of modification may be made within the protection scope of the present invention.

Claims

1. A distributed optical fiber sensing cross-domain event classification method based on model migration, characterized by: The steps include: S1. Distributed optical fiber sensors are arranged in the monitoring area to construct a labeled source domain vibration event dataset and an unlabeled target domain vibration event dataset; S2, train a source domain model using a labeled source domain vibration dataset; S3, transfer the knowledge of the source domain model to the target domain, and train the target domain model based only on the source domain model and the unlabeled target domain vibration event dataset; S4, deploy the trained target domain model into the distributed fiber optic sensing system to perform real-time event classification on new data in the target domain; The method for training the target domain model in step S3 includes the following steps: A1. Build pseudo labels and optimize them; A2, measure the uncertainty of the pseudo-label and calculate the corresponding weight; A3. In order to reduce the negative impact of incorrect labels on model training, a self-training loss function for target domain data is constructed; A4. Use multi-cycle historical pseudo-label information to exclude negative sample pairs that may belong to the same category; A5. Promote the diversity of classifier predictions through regularization methods and construct an overall loss function; In step A2, the uncertainty estimation method based on the consistency of neighborhood predictions measures the uncertainty of pseudo labels by calculating the entropy value of the prediction results of neighborhood samples, and reweights the classification loss according to the uncertainty. Pseudo labels with low uncertainty are given higher weights, and pseudo labels with high uncertainty are given lower weights. The steps are as follows: For a target domain sample , calculate its average prediction score according to the formula, which is defined as a probability distribution, and calculate the entropy of the probability distribution: ; Where C is the number of categories, and the entropy value is scaled according to the maximum value. The scaling formula is as follows: ; The weight of the sample is calculated by the normalized entropy value. The calculation formula is as follows: ; In step A3, a negative learning strategy is introduced into the classification loss, that is, when calculating the classification loss, complementary labels are selected instead of correct labels, and the self-training loss function of the target domain data is obtained as follows: ; in is a label randomly selected from a label set that does not contain the pseudo label of the current sample, that is, , It is a strongly enhanced image The probability output is The above-mentioned estimate The weight of uncertainty.

2. The distributed optical fiber sensing cross-domain event classification method based on model migration according to claim 1 is characterized in that: In step A1, the trained source domain model is used to generate preliminary pseudo labels for the unlabeled data in the target domain. Based on the neighborhood knowledge aggregation strategy, the pseudo labels are optimized by integrating the information of similar samples in the target domain. For each sample in the target domain, the cosine distance between samples is calculated to find K similar samples in the feature space, and the average score is calculated by the following soft voting strategy calculation formula: ; represents the nearest neighbor sample set, represents the probability output, and the superscript c indicates that the averaging operation is performed by class; Use the argmax operation to get the corrected pseudo-label: 。 3. The distributed optical fiber sensing cross-domain event classification method based on model migration according to claim 1 is characterized in that: In step A4, the target domain feature space is regularized using a contrastive learning framework to make positive sample pairs in the feature space close and negative sample pairs far apart. The prediction results of multiple past cycles are stored by introducing a time queue, and negative sample pairs that are marked as the same class in historical predictions are excluded. The steps are as follows: the pseudo labels of each sample in the past T cycles are stored in the queue, and then all sample pairs that share the same pseudo label at least once in the past T cycles are excluded from the negative sample pair list. Accordingly, the following loss function is optimized: ; in is the index set of samples in the storage queue that have never shared the same pseudo-label with the query sample in the past T cycles, q is the current query value, k+ is the positive sample key value, kj is the negative sample key value, is the temperature coefficient.

4. The distributed optical fiber sensing cross-domain event classification method based on model migration according to claim 1, characterized in that: In step A5, in order to avoid posterior collapse, that is, the classifier tends to predict all samples as the same class, the following regularization method is used to promote the diversity of classifier predictions: ; And construct the overall loss function: 。 5. The distributed optical fiber sensing cross-domain event classification method based on model migration according to claim 1, characterized in that: In step S2, the source domain model is a deep neural network, including a feature extractor and classifier Two parts.

6. The distributed optical fiber sensing cross-domain event classification method based on model migration according to claim 1, characterized in that: In the step S1, after the distributed optical fiber sensor collects environmental vibration data, the original vibration data is first filtered and denoised for data preprocessing, and then labeled source domain vibration event data and unlabeled target domain vibration event data are constructed.

Citation Information

Patent Citations

  • Migratable image recognition method and system based on distinguishing confidence level

    CN113435546A

  • Semantic segmentation network model uncertainty quantification method based on evidence reasoning

    CN113780292A

  • Target detection method

    CN117095252A

  • Distributed optical fiber sensing event identification method based on open set unsupervised domain adaptation

    CN118568550A

Cited By

  • Remote sensing data classification method based on transfer learning and label noise filtering

    CN120689757A