Optical fiber temperature measurement method, device and electronic equipment

Through the combination of deep learning model and multi-dimensional convolution kernel parameters, the limitation of fiber winding length in fiber temperature measurement on temperature measurement accuracy and detection distance is solved, effective temperature measurement when it is less than the minimum physical distance, and the fiber temperature measurement is improved flexibility and precision.

CN118936667BActive Publication Date: 2024-12-06HANGZHOU HIKVISION DIGITAL TECHNOLOGY CO LTD
View PDF 2 Cites 0 Cited by

Patent Information

Application Number
CN202411425775.3
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-10-12
Publication Date
2024-12-06
Estimated Expiration
2044-10-12

AI Technical Summary

Technical Problem

In the existing fiber temperature measurement methods, the length requirements of the fiber winding temperature measurement point limit the effective detection distance range of the fiber, resulting in a contradiction between the temperature measurement accuracy and the detection distance.

Method used

By training deep learning models, using the dynamic convolution method of multi-dimensional convolution kernel parameters and pulse signals, training data is generated to compensate for signal attenuation during fiber temperature measurement, reduce the impact of fiber winding length on temperature measurement accuracy, and achieve the following minimum Effective temperature measurement at physical distance of fiber.

Benefits of technology

It effectively avoids signal attenuation defects when the fiber wrap length is less than the minimum physical distance, eliminates the limitations on the detection distance of the fiber wrap length, and improves the temperature measurement accuracy and flexibility of the detection distance.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118936667B_ABST
    Figure CN118936667B_ABST
Patent Text Reader

Abstract

The present application provides an optical fiber temperature measurement method, device and electronic device. This embodiment compensates for the ratio of the light intensity value of the AST signal to the light intensity value of the ST signal measured during the optical fiber temperature measurement process by training a deep learning model, which can avoid the defect that when the length of the above-mentioned optical fiber winding temperature measurement point (referred to as the optical fiber winding length) is less than D_limit, the ratio of the light intensity value of the AST signal to the light intensity value of the ST signal in the optical fiber (referred to as the light intensity ratio) will decay accordingly with the optical fiber winding length, and can also eliminate the limitation of the effective detection distance range of the overall optical fiber caused by the length of the existing optical fiber winding temperature measurement point, such as the object to be measured (such as a machine, etc.), reaching the minimum optical fiber physical distance requirement, and in disguised form achieves the reduction of the length of the optical fiber winding temperature measurement point, such as the object to be measured (such as a machine, etc.), while achieving effective temperature measurement accuracy.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present application relates to optical fiber sensing technology, and in particular to an optical fiber temperature measurement method, device and electronic equipment. Background Art

[0002] Optical fiber has been widely used in the fields of electricity, petroleum, civil engineering, disaster monitoring, etc. in recent years due to its corrosion resistance, high temperature resistance, and anti-interference characteristics. Among them, one of the common uses of optical fiber is to measure the temperature of the temperature measurement point (also known as the object to be measured, such as a machine, etc.) in the above fields.

[0003] In the currently commonly used fiber optic temperature measurement methods, based on the limitations of the fiber optic temperature measurement principle, it is often required that the length of the fiber optic winding temperature measurement point, such as the object to be measured (such as a machine, etc.), reaches the minimum fiber optic physical distance to achieve effective temperature measurement accuracy. However, this requirement that the length of the fiber optic winding temperature measurement point, such as the object to be measured (such as a machine, etc.), reaches the minimum fiber optic physical distance often limits the effective detection distance range of the entire fiber optic. Summary of the invention

[0004] The present application provides an optical fiber temperature measurement method, device and electronic equipment to reduce the physical distance of optical fiber winding during temperature measurement.

[0005] An optical fiber temperature measurement method, the method comprising:

[0006] The multidimensional convolution kernel parameters are determined based on the light intensity ratios at different fiber winding lengths and different test temperatures; if any fiber winding length is less than the set minimum fiber physical distance D_limit, any light intensity ratio It refers to the ratio of the intensity of the anti-Stokes AST signal to the intensity of the Stokes ST signal collected after the laser signal is input into the optical fiber under the premise that the optical fiber winding length is d and the test temperature is t;

[0007] Based on the multidimensional convolution kernel parameters, a convolution result representation is obtained; the convolution result representation includes: a convolution result obtained by performing a convolution operation on the label data representation in each pulse interval using the multidimensional convolution kernel parameters; the label data representation is obtained by performing a convolution operation on the label data representation in each pulse interval using the multidimensional convolution kernel parameters; Insert multiple pulse signals on the line to obtain; each pulse interval refers to the line segment The interval corresponding to each pulse signal inserted on the line segment It is obtained by simulating the ratio of the light intensity value of the AST signal to the light intensity value of the ST signal in a specified coordinate system to decay with distance, wherein the ordinate of the specified coordinate system indicates the ratio of the light intensity value of the AST signal to the light intensity value of the ST signal, and the abscissa indicates the optical fiber winding length or the number of sampling points mapped by the optical fiber winding length;

[0008] By adding noise to the convolution result representation, a training data representation in the specified coordinate system is obtained;

[0009] Based on the training data representation and the label data representation, training data and label data corresponding to the training data are generated to train a deep learning model based on the training data and the label data; the deep learning model is used to compensate for the ratio of the light intensity value of the AST signal to the light intensity value of the ST signal during the optical fiber temperature measurement process to suppress the influence of the optical fiber winding length on the optical fiber temperature measurement.

[0010] An optical fiber temperature measuring device, comprising:

[0011] The determination unit is used to determine the multidimensional convolution kernel parameters according to the light intensity ratios of different fiber winding lengths and different test temperatures; if any fiber winding length is less than the set minimum fiber physical distance D_limit, any light intensity ratio It refers to the ratio of the intensity of the anti-Stokes AST signal to the intensity of the Stokes ST signal collected after the laser signal is input into the optical fiber under the premise that the optical fiber winding length is d and the test temperature is t;

[0012] A processing unit is used to obtain a convolution result representation based on the multidimensional convolution kernel parameters; the convolution result representation includes: a convolution result obtained by performing a convolution operation on the label data representation in each pulse interval using the multidimensional convolution kernel parameters; the label data representation is obtained by performing a convolution operation on the label data representation in each pulse interval; Insert multiple pulse signals on the line to obtain; each pulse interval refers to the line segment The interval corresponding to each pulse signal inserted on the line segment The ratio of the light intensity value of the AST signal to the light intensity value of the ST signal is simulated to attenuate with distance in a specified coordinate system, wherein the ordinate of the specified coordinate system indicates the ratio of the light intensity value of the AST signal to the light intensity value of the ST signal, and the abscissa indicates the winding length of the optical fiber or the number of sampling points mapped by the winding length of the optical fiber; and

[0013] By adding noise to the convolution result representation, a training data representation in the specified coordinate system is obtained; and,

[0014] Based on the training data representation and the label data representation, training data and label data corresponding to the training data are generated to train a deep learning model based on the training data and the label data; the deep learning model is used to compensate for the ratio of the light intensity value of the AST signal to the light intensity value of the ST signal during the optical fiber temperature measurement process to suppress the influence of the optical fiber winding length on the optical fiber temperature measurement.

[0015] An electronic device, comprising:

[0016] processor and machine-readable storage medium;

[0017] The machine-readable storage medium stores computer instructions, and when the computer instructions are executed by a processor, the steps in the above method are implemented.

[0018] It can be seen from the above technical solutions that in the present application, the ratio of the light intensity value of the AST signal to the light intensity value of the ST signal measured during the optical fiber temperature measurement process is compensated by training a deep learning model, which can avoid the defect that when the length of the above-mentioned optical fiber winding temperature measurement point (referred to as the optical fiber winding length) is less than D_limit, the ratio of the light intensity value of the AST signal to the light intensity value of the ST signal in the optical fiber (referred to as the light intensity ratio) will decay accordingly with the optical fiber winding length. It can also eliminate the limitation of the effective detection distance range of the overall optical fiber caused by the length of the existing optical fiber winding temperature measurement point, such as the object to be measured (such as a machine, etc.), reaching the requirement of the minimum optical fiber physical distance, and in disguised form realizes the reduction of the length of the optical fiber winding temperature measurement point, such as the object to be measured (such as a machine, etc.), while achieving effective temperature measurement accuracy;

[0019] Furthermore, in this embodiment, by means of the signal attenuation degree below the D_limit distance (when the length of the optical fiber winding temperature measurement point (referred to as the optical fiber winding length) is less than D_limit, the ratio of the light intensity value of the AST signal in the optical fiber to the light intensity value of the ST signal (referred to as the light intensity ratio) will decay accordingly with the optical fiber winding length), a convolution result representation (subsequently used to generate training data) is generated by a dynamic multi-convolution method, which is equivalent to the final training data fitting the effects of both data smoothing and data attenuation at the same time, so that the generated training data is more in line with the real data distribution;

[0020] Further, in this embodiment, by Insert pulse signal to build , which is more in line with the actual temperature measurement optical fiber system signal, and the restoration effect of the subsequent model is better and more robust. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present disclosure and, together with the description, serve to explain the principles of the present disclosure.

[0022] Figure 1 A schematic diagram of the relationship between the optical fiber winding length and the light intensity ratio is provided for the embodiment of the present application;

[0023] Figure 2 A flow chart of the method provided in the embodiment of the present application;

[0024] Figure 3 A schematic diagram of a light intensity curve provided in an embodiment of the present application;

[0025] Figure 4 A schematic diagram of label data representation and training data representation provided in an embodiment of the present application;

[0026] Figure 5 A model structure diagram provided for an embodiment of the present application;

[0027] Figure 6 Another method flow chart provided for an embodiment of the present application;

[0028] Figure 7 A schematic diagram of the effect provided by the embodiment of the present application;

[0029] Figure 8 A structural diagram of a device provided in an embodiment of the present application;

[0030] Fig. 9 A structural diagram of an electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0031] Exemplary embodiments will be described in detail herein, examples of which are shown in the accompanying drawings. When the following description refers to the drawings, the same numbers in different drawings represent the same or similar elements unless otherwise indicated. The implementations described in the following exemplary embodiments do not represent all implementations consistent with the present application. Instead, they are merely examples of devices and methods consistent with some aspects of the present application as detailed in the appended claims.

[0032] The terms used in this application are for the purpose of describing specific embodiments only and are not intended to limit this application. The singular forms "a", "said" and "the" used in this application and the appended claims are also intended to include plural forms, unless the context clearly indicates other meanings.

[0033] In order to enable those skilled in the art to better understand the technical solutions provided by the embodiments of the present application and to make the above-mentioned purposes, features and advantages of the embodiments of the present application more obvious and understandable, the technical solutions in the embodiments of the present application are further described in detail below in conjunction with the accompanying drawings.

[0034] In specific applications, the principle of optical fiber temperature measurement is mainly: a laser pulse signal is emitted to one end of the optical fiber wrapped around the temperature measurement point (called the head end, and correspondingly, the other end of the optical fiber is the tail end), and the laser pulse signal input into the optical fiber is scattered in the optical fiber to generate an anti-Stokes (AST: Anti-Stokes) signal and a Stokes (ST: Stokes) signal, and the AST signal and the ST signal are collected. The temperature of the temperature measurement point is determined based on the light intensity of the AST signal and the ST signal (for example, the ratio X obtained by dividing the light intensity of the AST signal by the light intensity of the ST signal). For example, based on the set ratio-temperature mapping principle, the temperature mapped by the ratio X is used as the actual temperature of the temperature measurement point.

[0035] In a specific application, the ratio of the light intensity of the AST signal to the light intensity of the ST signal changes with the length of the optical fiber winding around the temperature measurement point, as shown in the following example: Figure 1 As shown. Figure 1 As shown, when the length of the fiber-wrap temperature measurement point, such as the object to be measured (such as a machine, etc.), is less than the set minimum fiber physical distance (recorded as D_limit), the above ratio will decrease, and as the ratio decreases, the mapped temperature will decrease, which means that there is an error in the temperature of the measured temperature measurement point, which is equivalent to reducing the spatial resolution of the entire fiber-optic temperature measurement system. Based on this, in order to ensure the temperature measurement accuracy, as described above, the fiber-optic temperature measurement principle requires that the length of the fiber-wrap temperature measurement point, such as the object to be measured (such as a machine, etc.), is greater than or equal to D_limit.

[0036] As described in the background technology, the optical fiber temperature measurement principle requires that the length of the optical fiber winding temperature measurement point, such as the object to be measured (such as a machine, etc.), be greater than or equal to D_limit, which will limit the effective detection distance range of the entire optical fiber.

[0037] To avoid the above problems, the present application provides the following embodiments: Figure 2 The flowchart shown:

[0038] Participate in construction Figure 2 , Figure 2 This is a flow chart of a method provided in an embodiment of the present application. The method can be applied to electronic devices such as terminals, etc., and this embodiment does not specifically limit it.

[0039] like Figure 2 As shown, the process may include the following steps:

[0040] Step 201, determining the multidimensional convolution kernel parameters according to the light intensity ratios at different optical fiber winding lengths and different test temperatures.

[0041] In this embodiment, the optical fiber winding length involved in step 201 is less than the set minimum optical fiber physical distance D_limit. It refers to the ratio of the light intensity value of the AST signal collected after the laser signal is input into the optical fiber under the premise that the optical fiber winding length is d and the test temperature is t to the light intensity value of the ST signal.

[0042] according to Figure 1 It can be seen that when the length of the fiber winding temperature measurement point (referred to as the fiber winding length) is less than D_limit, the ratio of the light intensity value of the AST signal in the fiber to the light intensity value of the ST signal (referred to as the light intensity ratio) will decay accordingly with the fiber winding length. Therefore, the multidimensional convolution kernel parameter (ker) here can be determined according to the ratio of the light intensity value of the AST signal to the light intensity value of the ST signal when the temperature is measured at each fiber winding length less than D_limit. For example, the winding length of the temperature measuring fiber (the length of the winding temperature measurement point) is set to: 0.5 meters, 1 meter, 2 meters...D_limit. Then, for each fiber winding length, temperature measurement tests are performed in a constant temperature environment, such as different temperatures at equal intervals within the effective measurement temperature range, to collect the corresponding light intensity values ​​of the AST signal and the light intensity values ​​of the ST signal. Then, the multidimensional convolution kernel parameters can be determined based on the light intensity ratios at different fiber winding lengths and different test temperatures.

[0043] As an embodiment, in step 201, the multidimensional convolution kernel parameters are determined based on different optical fiber winding lengths (also referred to as the optical fiber winding lengths used during testing) and light intensity ratios at different test temperatures. There may be multiple implementation methods for specific implementation, such as: first, for each optical fiber winding length, based on the obtained light intensity ratios at different test temperatures for the optical fiber winding length, the weight (also referred to as adjustment weight) of the optical fiber winding length is determined; then, based on the weights of each optical fiber winding length and the obtained convolution kernel related parameter vector, the multidimensional convolution kernel parameters are determined.

[0044] Optionally, for each fiber winding length, based on the obtained light intensity ratio at different test temperatures for the fiber winding length, there are multiple ways to determine the weight of the fiber winding length. Afterwards, based on Calculate the ratio height . Optionally, It can be expressed by the following formula: . Afterwards, according to Calculate the weight corresponding to the fiber winding length d . It can be calculated according to the following formula:

[0045] ;

[0046] Wherein, num_t represents the number of test temperatures at each fiber winding length d.

[0047] Optionally, the above It refers to the average value of the ratio of the light intensity value of the AST signal to the light intensity value of the ST signal collected at different fiber winding lengths at the test temperature t. Or, It can be an empirical value set according to actual needs, and this embodiment does not specifically limit it.

[0048] Finally, based on the above description, the weight corresponding to each optical fiber winding length can be obtained.

[0049] The following describes the above convolution kernel related parameter vector:

[0050] The convolution kernel related parameter vector is generated using the light intensity value at each sampling point involved in the falling edge of the light intensity value curve. Optionally, this embodiment can intercept the falling edge of the light intensity value curve and perform normalization, such as normalizing the light intensity value at each sampling point in the falling edge portion to obtain a k-dimensional one-dimensional vector H; k represents the number of sampling points. Afterwards, the convolution kernel related parameters corresponding to each sampling point are calculated according to the following formula:

[0051] ;

[0052] Where i represents the i-th sampling point, i=0 represents the last sampling point, express The normalized result of the light intensity value corresponding to the i-th sampling point in is: Represents the convolution kernel related parameters corresponding to the i-th sampling point. Finally, based on the convolution kernel related parameters corresponding to each sampling point, the convolution kernel related parameter vector is determined. For example, the convolution kernel related parameters corresponding to each sampling point are organized together in sequence to form a convolution kernel related parameter vector.

[0053] After obtaining the convolution kernel related parameter vector, the multidimensional convolution kernel parameters corresponding to each fiber winding length can be determined based on the weight of each fiber winding length and the obtained convolution kernel related parameter vector. For example, the weight corresponding to each fiber winding length is multiplied by the obtained convolution kernel related parameter vector, and the product is determined as the multidimensional convolution kernel parameter corresponding to the fiber winding length.

[0054] For example, when the fiber winding length is d, the corresponding weight is , then the multidimensional convolution kernel parameter corresponding to the fiber winding length d is It is expressed by the following formula: .

[0055] It should be noted that in specific applications, theoretically, the light intensity curve is as follows Figure 3 As shown in the figure, the actual measured light intensity curve is different from the theoretical light intensity curve. Figure 3As shown. In this embodiment, the above-mentioned light intensity value curve is a curve formed by the light intensity value of the laser signal measured at the tail end of the optical fiber after the laser signal is input at the head end of the optical fiber. It can be seen that in fact, the falling edge of the measured light intensity value curve has a corresponding smoothing effect, which is similar to the convolution operation.

[0056] Step 201 is described above.

[0057] Step 202, based on the multidimensional convolution kernel parameters, a convolution result representation is obtained; the convolution result representation includes: a convolution result obtained by performing a convolution operation on the label data representation in each pulse interval using the multidimensional convolution kernel parameters; the label data representation is ... Insert multiple pulse signals on it; each pulse interval refers to a line segment The interval corresponding to each pulse signal inserted on the line segment It is obtained by simulating the ratio of the light intensity value of the AST signal to the light intensity value of the ST signal in the specified coordinate system by attenuation with distance. The ordinate of the specified coordinate system indicates the ratio of the light intensity value of the AST signal to the light intensity value of the ST signal, and the abscissa indicates the optical fiber winding length or the number of sampling points mapped by the optical fiber winding length.

[0058] Based on the above description, when this step 202 is specifically implemented, it will be based on the following principle: when the length of the fiber winding temperature measurement point (referred to as the fiber winding length) is less than D_limit, the ratio of the light intensity value of the AST signal in the optical fiber to the light intensity value of the ST signal (referred to as the light intensity ratio) will decay accordingly with the fiber winding length, and a line segment in the specified coordinate system in which the light intensity ratio decays as the fiber winding length decreases is simulated. (The line segment The inclination angle is negative The length of is a random value such as .

[0059] Afterwards, in Multiple pulse signals are randomly inserted at different positions of The data is represented by randomly inserting multiple pulse signals at different positions. . Figure 4 By using a negative slope or inclination Insert pulse signal to build , which is more in line with the actual temperature measurement optical fiber system signal, and the restoration effect of the subsequent model is better and more robust.

[0060] Optionally, the height of the pulse signal is inserted With width The ordinate of the designated coordinate system indicates the ratio of the light intensity value of the AST signal to the light intensity value of the ST signal, and the abscissa indicates the fiber winding length or the number of sampling points mapped by the fiber winding length.

[0061] Afterwards, using the above multidimensional convolution kernel parameters, The interval corresponding to each pulse signal inserted in Convolution operation is performed. Specifically, the starting position of any pulse signal is , the end position is . To ensure the continuity of data after the convolution operation, in this embodiment, the interval corresponding to each pulse signal is not from the starting position to the end position of the pulse signal, but the set range is expanded before and after the interval from the starting position to the end position of the pulse signal, such as expanding half of the convolution kernel. That is, the starting position of the pulse interval corresponding to each pulse signal is the difference between the starting position of the pulse signal and the set value, and the end position of the interval is the sum of the starting position of the pulse signal and the pulse width of the pulse signal and the above-mentioned set value. Taking the above-mentioned expansion of half the convolution kernel as an example, the starting coordinates of the interval corresponding to the pulse signal are The ending coordinates are . Afterwards, in [ , ] Perform convolution operation and get Dimensional convolution result The convolution operation is shown below:

[0062] .

[0063] Optionally, in this embodiment, yes Corresponding multi-dimensional convolution kernel parameters. Based on the above description, here Can be the convolution kernel related parameter vector and The weight corresponding to the number of sampling points converted For example, .

[0064] In this embodiment, The weight corresponding to the number of sampling points converted Available in weight set Can be found in . It is expressed by the following formula: 】.in, Indicates the number of sampling points The corresponding weight. is the weight corresponding to the fiber winding length D_limit (corresponding to the number of sampling points Slimit). According to the above formula, The value is 1.

[0065] In this embodiment, for the fiber winding length used in the above test, after obtaining the weight corresponding to each fiber winding length, the fiber winding length can be converted into the corresponding number of sampling points based on the set algorithm, and the weight corresponding to the fiber winding length is recorded as the weight corresponding to the sampling point number in the weight set. .

[0066] It should be noted that In addition to the weights of the sampling points corresponding to the fiber winding lengths used in the above test, the weights of other sampling points are also included. Optionally, the weights of other sampling points can be determined according to the following formula: For example, the If the number of sampling points is not the same as the number of sampling points corresponding to any of the above fiber winding lengths, then the number of sampling points corresponding to the fiber winding lengths used in the above test is selected. The number of adjacent points with the largest value , and The number of adjacent points with the smallest value ,Right now Then use the test results The corresponding weights and The corresponding weights can be obtained by differential calculation The corresponding weights are recorded in the weight set .

[0067] It can be seen that this embodiment uses the signal attenuation degree below the D_limit distance (when the length of the fiber winding temperature measurement point (referred to as the fiber winding length) is less than D_limit, the ratio of the light intensity value of the AST signal in the optical fiber to the light intensity value of the ST signal (referred to as the light intensity ratio) will decay accordingly with the fiber winding length) to generate a convolution result representation (subsequently used to generate training data) through a dynamic multi-convolution method, which is equivalent to the final training data simultaneously fitting the effects of both data smoothing and data attenuation, so that the generated training data is more in line with the actual data distribution.

[0068] Step 202 is described above.

[0069] Step 203, by adding noise to the above convolution result representation, obtain the training data representation in the above specified coordinate system.

[0070] For example, random Gaussian noise is added to the above convolution result to obtain the training data representation . Figure 4The training data representation is also shown as an example .

[0071] Step 204, based on the training data representation and the label data representation, generate training data and label data corresponding to the training data, so as to train a deep learning model based on the training data and the label data; the deep learning model is used to compensate for the ratio of the measured light intensity value of the AST signal to the light intensity value of the ST signal during the optical fiber temperature measurement process, so as to suppress the influence of the optical fiber winding length on the optical fiber temperature measurement.

[0072] Optionally, in this embodiment, in order to eliminate the instability of model training caused by individual data differences, this embodiment represents the label data And the training data representation Perform normalization. Take the light intensity ratio corresponding to each designated sampling point in the training data representation after normalization as training data. Take the light intensity ratio corresponding to each designated sampling point in the label data representation after normalization as label data; the label data corresponding to the same sampling point as the training data is the label data corresponding to the training data.

[0073] Optionally, the normalization of the training data representation and the label data representation may include: for the two matching data representations , , determine the maximum light intensity ratio Ratio to minimum light intensity Optionally, the maximum light intensity ratio here is the maximum light intensity ratio selected from the above training data representation and label data representation. The minimum light intensity ratio here is is the minimum light intensity ratio selected from the above training data representation and label data representation. That is: , , , Then, based on the light intensity ratio corresponding to each designated sampling point in the above training data representation, and the above and the above , determine the training data corresponding to the specified sampling point; and, based on the light intensity ratio of each specified sampling point in the above label data representation, and the above and the above , determine the label data corresponding to the specified sampling point.

[0074] Optionally, the label data corresponding to any specified sampling point can be calculated according to the following formula:

[0075] ;

[0076] in, Indicates the light intensity ratio corresponding to the i-th sampling point in the label data representation, Represents the label data corresponding to the i-th sampling point.

[0077] Optionally, the training data corresponding to any specified sampling point can be calculated according to the following formula:

[0078] ;

[0079] in, represents the light intensity ratio corresponding to the i-th sampling point in the training data representation, Represents the training data corresponding to the i-th sampling point.

[0080] It can be seen that the method of generating training data and label data in this embodiment effectively reduces the large amount of real data required for general supervised learning and reduces work costs; it can also solve the defect that real labels cannot be obtained during supervised learning.

[0081] After obtaining the training data and the label data corresponding to the training data, a deep learning model can be trained based on the training data and the label data corresponding to the training data. Figure 5 As shown in Figure 1, the deep learning model can include three parts: InBlock, N ResBlocks, and OutBlock. InBlock mainly increases the number of data channels, and the data dimension is increased by become For example, for a training data , its data dimension , the training data passes through InBlock, and the data dimension is become The N ResBlocks in the middle maintain the data dimension OutBlock is used to first use Conv1D to transform the data dimension Transformed into , and then pass through the Sigmoid layer to output the data Mapped between (0, 1). In this embodiment, the normalization (LayerNorm) layer in InBlock or ResBlock performs the above normalization to ensure that the model robustness is enhanced. In this embodiment, the Sigmoid layer performs the opposite denormalization operation (also called the inverse normalization process) based on the previous normalization, limiting the output distribution to between 0.14 and 0.71, making the output distribution interval of different inputs more consistent, and in the non-saturated interval of sigmoid, the model converges faster and has better robustness.

[0082] Optionally, in this embodiment, there are multiple loss functions used when training the deep learning model, such as the mean square error MSE loss function, etc., which are not specifically limited in this embodiment. Taking the mean square error MSE loss function as an example, the loss function is, for example: .

[0083] The purpose of the loss function is to expect the output based on each training data As close as possible to the label data corresponding to the training data .

[0084] After a converged deep learning model (referred to as model M) is obtained through the above training, temperature measurement can be performed based on the model M. The overall workflow is as follows: Figure 6 I will not go into details here.

[0085] In this embodiment, through the above-mentioned model M, the ratio of the light intensity value of the AST signal to the light intensity value of the ST signal can be compensated during the optical fiber temperature measurement process to suppress the influence of the optical fiber winding length on the optical fiber temperature measurement. In this way, even if the optical fiber winding length of the optical fiber at the temperature measurement point is less than the set minimum optical fiber physical distance D_limit, the ratio of the light intensity value of the AST signal to the light intensity value of the ST signal measured at this time is compensated through the above-mentioned model M. This can avoid the defect that when the length of the above-mentioned optical fiber winding temperature measurement point (referred to as the optical fiber winding length) is less than D_limit, the ratio of the light intensity value of the AST signal to the light intensity value of the ST signal in the optical fiber (referred to as the light intensity ratio) will decay accordingly with the optical fiber winding length. It can also eliminate the limitation of the effective detection distance range of the overall optical fiber caused by the length of the existing optical fiber winding temperature measurement point, such as the object to be measured (such as a machine, etc.), reaching the requirement of the minimum optical fiber physical distance, and in disguise realizes the reduction of the length of the optical fiber winding temperature measurement point, such as the object to be measured (such as a machine, etc.), while achieving effective temperature measurement accuracy.

[0086] To make the above effects more intuitive, Figure 7 The effect diagrams before and after compensation (that is, before and after the model M is restored) are shown as examples. Figure 7 The four peaks from left to right in the figure represent the measured temperatures when the fiber winding length is 0.5m, 1m, 2m, and D_limit (for example, 3m). Based on the compensation of model M, even if the fiber winding length is not D_limit, such as 1m, the accuracy of the measured temperature is the same as the accuracy of the temperature measured when the fiber winding length is D_limit. In disguise, it reduces the length of the fiber winding temperature measurement point, such as the object to be measured (such as a machine, etc.), while achieving effective temperature measurement accuracy.

[0087] So far, completed Figure 2 The process shown.

[0088] pass Figure 2The process shown realizes compensation of the ratio of the light intensity value of the AST signal to the light intensity value of the ST signal measured in the optical fiber temperature measurement process by training a deep learning model, which can avoid the defect that when the length of the above-mentioned optical fiber winding temperature measurement point (referred to as the optical fiber winding length) is less than D_limit, the ratio of the light intensity value of the AST signal to the light intensity value of the ST signal in the optical fiber (referred to as the light intensity ratio) will decay accordingly with the optical fiber winding length, and can also eliminate the limitation of the effective detection distance range of the overall optical fiber caused by the length of the existing optical fiber winding temperature measurement point, such as the object to be measured (such as a machine, etc.), reaching the requirement of the minimum optical fiber physical distance, and in disguised form realizes the reduction of the length of the optical fiber winding temperature measurement point, such as the object to be measured (such as a machine, etc.), while achieving the effective temperature measurement accuracy;

[0089] Furthermore, in this embodiment, by means of the signal attenuation degree below the D_limit distance (when the length of the optical fiber winding temperature measurement point (referred to as the optical fiber winding length) is less than D_limit, the ratio of the light intensity value of the AST signal in the optical fiber to the light intensity value of the ST signal (referred to as the light intensity ratio) will decay accordingly with the optical fiber winding length), a convolution result representation (subsequently used to generate training data) is generated by a dynamic multi-convolution method, which is equivalent to the final training data fitting the effects of both data smoothing and data attenuation at the same time, so that the generated training data is more in line with the real data distribution;

[0090] Further, in this embodiment, by Insert pulse signal to build , which is more in line with the actual temperature measurement optical fiber system signal, and the restoration effect of the subsequent model is better and more robust.

[0091] Below Figure 6 The process shown is described as follows:

[0092] Participate in construction Figure 6 , Figure 6 Another flow chart provided for the embodiment of the present application. Figure 6 As shown, the process may include:

[0093] Step 601, during the optical fiber temperature measurement process, obtain the collected effective light intensity ratio interval; the effective light intensity ratio interval includes multiple different light intensity ratios, and any light intensity ratio is the ratio of the light intensity value of the AST signal collected within the effective interval of the optical fiber to the light intensity value of the ST signal.

[0094] In specific applications, for any optical fiber, the ST signal intensity value and AST signal intensity value measured at a length starting from the fiber head end and a length before the fiber tail end can be ignored, and the ratio of the ST signal intensity value to the AST signal intensity value measured at each position in the remaining length (i.e., the effective interval of the optical fiber) is determined as the effective intensity ratio. Finally, the effective intensity ratios constitute the effective intensity ratio interval.

[0095] Step 602, based on each light intensity ratio in the effective light intensity ratio interval and in accordance with the method of determining training data or label data, determine the input light intensity ratio.

[0096] Optionally, the effective light intensity ratio interval may be normalized according to the following steps: selecting the light intensity ratio with the largest value and the light intensity ratio with the smallest value from the effective light intensity ratio interval; based on each light intensity ratio in the effective light intensity ratio interval, based on the light intensity ratio, the light intensity ratio with the largest value and the light intensity ratio with the smallest value, determining the input light intensity ratio; for example, calculating the input light intensity ratio according to the following formula:

[0097] ;

[0098] in, represents the i-th light intensity ratio in the effective light intensity ratio interval, represents the i-th input light intensity ratio, Indicates the minimum light intensity ratio selected from the effective light intensity ratio range. Indicates the maximum light intensity ratio selected from the effective light intensity ratio range.

[0099] Step 603, inputting the input light intensity ratio into the above-mentioned deep learning model to obtain the output light intensity ratio.

[0100] Step 604, determining the temperature based on the output light intensity ratio.

[0101] For example, the inverse process of the above normalization process is performed on each output light intensity ratio to obtain a target light intensity ratio, and the temperature is determined based on the target light intensity ratio.

[0102] Optionally, the target light intensity ratio may be calculated according to the following formula:

[0103] ;

[0104] in, Indicates the target light intensity ratio, Indicates any output light intensity ratio.

[0105] After obtaining the target light intensity ratio, temperature mapping is performed to obtain The mapping process here is not within the scope of protection of this patent, so it will not be expanded here. Then output , or, Compared with the set temperature threshold thred, if If it is greater than thred, it is confirmed that the temperature measured by the optical fiber exceeds the range and an alarm is output.

[0106] So far, completed Figure 6 The process shown.

[0107] pass Figure 6 The process shown implements how to apply deep learning models for optical fiber temperature measurement.

[0108] The method provided in the embodiment of the present application is described above. The device provided in the embodiment of the present application is described below:

[0109] See also Figure 8 , Figure 8 This is a diagram of the device structure provided in the embodiment of the present application. Figure 8 As shown, the device may include:

[0110] The determination unit is used to determine the multidimensional convolution kernel parameters according to the light intensity ratios of different fiber winding lengths and different test temperatures; if any fiber winding length is less than the set minimum fiber physical distance D_limit, any light intensity ratio It refers to the ratio of the intensity of the anti-Stokes AST signal to the intensity of the Stokes ST signal collected after the laser signal is input into the optical fiber under the premise that the optical fiber winding length is d and the test temperature is t;

[0111] A processing unit is used to obtain a convolution result representation based on the multidimensional convolution kernel parameters; the convolution result representation includes: a convolution result obtained by performing a convolution operation on the label data representation in each pulse interval using the multidimensional convolution kernel parameters; the label data representation is obtained by performing a convolution operation on the label data representation in each pulse interval; Insert multiple pulse signals on the line to obtain; each pulse interval refers to the line segment The interval corresponding to each pulse signal inserted on the line segment The ratio of the light intensity value of the AST signal to the light intensity value of the ST signal is simulated to attenuate with distance in a specified coordinate system, wherein the ordinate of the specified coordinate system indicates the ratio of the light intensity value of the AST signal to the light intensity value of the ST signal, and the abscissa indicates the winding length of the optical fiber or the number of sampling points mapped by the winding length of the optical fiber; and

[0112] By adding noise to the convolution result representation, a training data representation in the specified coordinate system is obtained; and,

[0113] Based on the training data representation and the label data representation, training data and label data corresponding to the training data are generated to train a deep learning model based on the training data and the label data; the deep learning model is used to compensate for the ratio of the light intensity value of the AST signal to the light intensity value of the ST signal during the optical fiber temperature measurement process to suppress the influence of the optical fiber winding length on the optical fiber temperature measurement.

[0114] Optionally, determining the multidimensional convolution kernel parameters according to the light intensity ratios at different optical fiber winding lengths and different test temperatures includes:

[0115] For each optical fiber winding length, based on the obtained light intensity ratios at different test temperatures under the optical fiber winding length, determine the weight of the optical fiber winding length;

[0116] The multidimensional convolution kernel parameters are determined based on the weight of each optical fiber winding length and the obtained convolution kernel related parameter vector; the convolution kernel related parameter vector is generated using the light intensity value at each sampling point involved in the falling edge of the light intensity value curve; the light intensity value curve is a curve formed by the light intensity value of the laser signal measured at the tail end of the optical fiber after the laser signal is input at the head end of the optical fiber.

[0117] Optionally, the convolution kernel related parameter vector is generated by the following steps:

[0118] Normalizing the light intensity values ​​at each sampling point involved in the falling edge of the light intensity value curve to obtain a k-dimensional one-dimensional vector H, where k represents the number of sampling points;

[0119] The convolution kernel parameters corresponding to each sampling point are calculated according to the following formula:

[0120] ;

[0121] Where i represents the i-th sampling point, i=0 represents the last sampling point, express The normalized result of the light intensity value corresponding to the i-th sampling point in is: Represents the convolution kernel related parameters corresponding to the i-th sampling point;

[0122] Based on the convolution kernel related parameters corresponding to each sampling point, the convolution kernel related parameter vector is determined.

[0123] Optionally, the pulse interval corresponding to each pulse signal is determined based on the starting position and pulse width of the pulse signal;

[0124] The interval start position of the pulse interval corresponding to each pulse signal is the difference between the start position and the set value, and the interval end position is the sum of the start position, the pulse width and the set value, so as to ensure the continuity of the convolution operation.

[0125] Optionally, based on the training data representation and the label data representation, generating the training data and the label data corresponding to the training data includes:

[0126] Normalizing the training data representation and the label data representation to eliminate the impact of data differences on deep learning model training;

[0127] The light intensity ratio corresponding to each designated sampling point in the normalized training data representation is used as training data;

[0128] The light intensity ratios corresponding to each designated sampling point in the normalized label data representation are used as label data; the label data corresponding to the same sampling point as the training data is the label data corresponding to the training data.

[0129] Optionally, the normalizing the training data representation and the label data representation includes:

[0130] Determine a maximum light intensity ratio value having a maximum value from the training data representation and the label data representation The ratio of the minimum light intensity to the minimum value ;

[0131] Based on the light intensity ratio corresponding to each designated sampling point in the training data representation, and the and stated , determine the training data corresponding to the specified sampling point;

[0132] Based on the light intensity ratio of each designated sampling point in the label data representation, and the and stated , determine the label data corresponding to the specified sampling point.

[0133] Optionally, the processing unit further obtains an effective light intensity ratio interval that has been collected during the optical fiber temperature measurement process; the effective light intensity ratio interval includes a plurality of different light intensity ratios, and any light intensity ratio is a ratio of the light intensity value of the AST signal collected within the effective interval of the optical fiber to the light intensity value of the ST signal;

[0134] Based on each light intensity ratio in the effective light intensity ratio interval, and in accordance with the method of determining the training data or the label data, determine the input light intensity ratio;

[0135] Inputting the input light intensity ratio into the deep learning model to obtain an output light intensity ratio;

[0136] The temperature is determined based on the output light intensity ratio.

[0137] Optionally, the determining the input light intensity ratio based on each light intensity ratio in the effective light intensity ratio interval and in a manner of determining training data or label data includes:

[0138] The effective light intensity ratio interval is normalized according to the following steps: selecting a light intensity ratio with a maximum value and a light intensity ratio with a minimum value from the effective light intensity ratio interval; based on each light intensity ratio in the effective light intensity ratio interval, based on the light intensity ratio, the light intensity ratio with a maximum value and the light intensity ratio with a minimum value, determining an input light intensity ratio;

[0139] Determining the temperature based on the output light intensity ratio comprises:

[0140] The inverse process of the normalization process is performed on each of the output light intensity ratios to obtain a target light intensity ratio, and the temperature is determined based on the target light intensity ratio.

[0141] So far, completed Figure 8 The device shown.

[0142] The present application also provides Figure 8 The hardware structure of the device shown is described. Fig. 9 , Fig. 9 This is a diagram of the hardware structure (electronic device) provided in the embodiment of the present application. Fig. 9 As shown, the hardware structure may include: a processor and a machine-readable storage medium.

[0143] The machine-readable storage medium stores computer instructions, and when the computer instructions are executed by a processor, the method disclosed in the above example of the present application can be implemented.

[0144] Based on the same application concept as the above method, an embodiment of the present application also provides a machine-readable storage medium, on which a number of computer instructions are stored. When the computer instructions are executed by a processor, the method disclosed in the above example of the present application can be implemented.

[0145] Exemplarily, the above-mentioned machine-readable storage medium can be any electronic, magnetic, optical or other physical storage device, which can contain or store information, such as executable instructions, data, etc. For example, the machine-readable storage medium can be: RAM (Radom Access Memory), volatile memory, non-volatile memory, flash memory, storage drive (such as hard disk drive), solid state drive, any type of storage disk (such as optical disk, DVD, etc.), or similar storage medium, or a combination thereof.

[0146] For the convenience of description, the above device is described in various units according to their functions. Of course, when implementing the present application, the functions of each unit can be implemented in the same or multiple software and / or hardware.

[0147] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the embodiments of the present application may adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program codes.

[0148] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0149] Moreover, these computer program instructions can also be stored in a computer-readable memory that can guide a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer-readable memory produce a product including an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 The function specified in one or more boxes.

[0150] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for executing on the computer or other programmable device to implement the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0151] The above is only an embodiment of the present application and is not intended to limit the present application. For those skilled in the art, the present application may have various changes and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application should be included in the scope of the claims of the present application.

Claims

1. A method for measuring temperature using optical fiber. It is characterized in that The method includes: The multidimensional convolution kernel parameters are determined based on the light intensity ratios at different fiber winding lengths and different test temperatures; if any fiber winding length is less than the set minimum fiber physical distance D_limit, any light intensity ratio It refers to the ratio of the intensity of the anti-Stokes AST signal to the intensity of the Stokes ST signal collected after the laser signal is input into the optical fiber under the premise that the optical fiber winding length is d and the test temperature is t; Based on the multidimensional convolution kernel parameters, a convolution result representation is obtained; the convolution result representation includes: a convolution result obtained by performing a convolution operation on the label data representation in each pulse interval using the multidimensional convolution kernel parameters; the label data representation is obtained by performing a convolution operation on the label data representation in each pulse interval using the multidimensional convolution kernel parameters; Insert multiple pulse signals on the line to obtain; each pulse interval refers to the line segment The interval corresponding to each pulse signal inserted on the line segment It is obtained by simulating the ratio of the light intensity value of the AST signal to the light intensity value of the ST signal in a specified coordinate system to decay with distance, wherein the ordinate of the specified coordinate system indicates the ratio of the light intensity value of the AST signal to the light intensity value of the ST signal, and the abscissa indicates the optical fiber winding length or the number of sampling points mapped by the optical fiber winding length; By adding noise to the convolution result representation, a training data representation in the specified coordinate system is obtained; Based on the training data representation and the label data representation, training data and label data corresponding to the training data are generated to train a deep learning model based on the training data and the label data; the deep learning model is used to compensate for the ratio of the light intensity value of the AST signal to the light intensity value of the ST signal during the optical fiber temperature measurement process to suppress the influence of the optical fiber winding length on the optical fiber temperature measurement.

2. The method according to claim 1, It is characterized in that Determining the multidimensional convolution kernel parameters according to the light intensity ratios at different optical fiber winding lengths and different test temperatures includes: For each optical fiber winding length, based on the obtained light intensity ratios at different test temperatures under the optical fiber winding length, determine the weight of the optical fiber winding length; Based on the weight of each optical fiber winding length and the obtained convolution kernel related parameter vector, the multidimensional convolution kernel parameters corresponding to each optical fiber winding length are determined; the convolution kernel related parameter vector is generated using the light intensity value at each sampling point involved in the falling edge of the light intensity value curve; the light intensity value curve is a curve formed by the light intensity value of the laser signal measured at the tail end of the optical fiber after the laser signal is input at the head end of the optical fiber.

3. The method according to claim 2, It is characterized in that The convolution kernel related parameter vector is generated by the following steps: Normalizing the light intensity values ​​at each sampling point involved in the falling edge of the light intensity value curve to obtain a k-dimensional one-dimensional vector H, where k represents the number of sampling points; The convolution kernel parameters corresponding to each sampling point are calculated according to the following formula: ; Where i represents the i-th sampling point, i=0 represents the last sampling point, express The normalized result of the light intensity value corresponding to the i-th sampling point in is: Represents the convolution kernel related parameters corresponding to the i-th sampling point; Based on the convolution kernel related parameters corresponding to each sampling point, the convolution kernel related parameter vector is determined.

4. The method according to claim 1, It is characterized in that The pulse interval corresponding to each pulse signal is determined based on the starting position and pulse width of the pulse signal; The interval start position of the pulse interval corresponding to each pulse signal is the difference between the start position and the set value, and the interval end position is the sum of the start position, the pulse width and the set value, so as to ensure the continuity of the convolution operation.

5. The method according to claim 1, It is characterized in that Based on the training data representation and the label data representation, generating training data and label data corresponding to the training data includes: Normalizing the training data representation and the label data representation to eliminate the impact of data differences on deep learning model training; The light intensity ratio corresponding to each designated sampling point in the normalized training data representation is used as training data; The light intensity ratios corresponding to each designated sampling point in the normalized label data representation are used as label data; the label data corresponding to the same sampling point as the training data is the label data corresponding to the training data.

6. The method according to claim 5, It is characterized in that The normalizing the training data representation and the label data representation comprises: Determine a maximum light intensity ratio value having a maximum value from the training data representation and the label data representation The ratio of the minimum light intensity to the minimum value ; Based on the light intensity ratio corresponding to each designated sampling point in the training data representation, and the and stated , determine the training data corresponding to the specified sampling point; Based on the light intensity ratio of each designated sampling point in the label data representation, and the and stated , determine the label data corresponding to the specified sampling point.

7. The method according to claim 1, It is characterized in that The method further comprises: In the process of optical fiber temperature measurement, an effective light intensity ratio interval that has been collected is obtained; the effective light intensity ratio interval includes a plurality of different light intensity ratios, and any light intensity ratio is the ratio of the light intensity value of the AST signal collected within the effective interval of the optical fiber to the light intensity value of the ST signal; Based on each light intensity ratio in the effective light intensity ratio interval, and in accordance with the method of determining the training data or the label data, determine the input light intensity ratio; Inputting the input light intensity ratio into the deep learning model to obtain an output light intensity ratio; The temperature is determined based on the output light intensity ratio.

8. The method according to claim 7, It is characterized in that The step of determining the input light intensity ratio based on each light intensity ratio in the effective light intensity ratio interval and in accordance with the method of determining the training data or the label data includes: The effective light intensity ratio interval is normalized according to the following steps: selecting a light intensity ratio with a maximum value and a light intensity ratio with a minimum value from the effective light intensity ratio interval; based on each light intensity ratio in the effective light intensity ratio interval, based on the light intensity ratio, the light intensity ratio with a maximum value and the light intensity ratio with a minimum value, determining an input light intensity ratio; Determining the temperature based on the output light intensity ratio comprises: The inverse process of the normalization process is performed on each of the output light intensity ratios to obtain a target light intensity ratio, and the temperature is determined based on the target light intensity ratio.

9. An optical fiber temperature measuring device, It is characterized in that The device includes: The determination unit is used to determine the multidimensional convolution kernel parameters according to the light intensity ratios of different fiber winding lengths and different test temperatures; if any fiber winding length is less than the set minimum fiber physical distance D_limit, any light intensity ratio It refers to the ratio of the intensity of the anti-Stokes AST signal to the intensity of the Stokes ST signal collected after the laser signal is input into the optical fiber under the premise that the optical fiber winding length is d and the test temperature is t; A processing unit is used to obtain a convolution result representation based on the multidimensional convolution kernel parameters; the convolution result representation includes: a convolution result obtained by performing a convolution operation on the label data representation in each pulse interval using the multidimensional convolution kernel parameters; the label data representation is ... Insert multiple pulse signals on the line to obtain; each pulse interval refers to the line segment The interval corresponding to each pulse signal inserted on the line segment The ratio of the light intensity value of the AST signal to the light intensity value of the ST signal is simulated to attenuate with distance in a specified coordinate system, wherein the ordinate of the specified coordinate system indicates the ratio of the light intensity value of the AST signal to the light intensity value of the ST signal, and the abscissa indicates the winding length of the optical fiber or the number of sampling points mapped by the winding length of the optical fiber; and By adding noise to the convolution result representation, a training data representation in the specified coordinate system is obtained; and, Based on the training data representation and the label data representation, training data and label data corresponding to the training data are generated to train a deep learning model based on the training data and the label data; the deep learning model is used to compensate for the ratio of the light intensity value of the AST signal to the light intensity value of the ST signal during the optical fiber temperature measurement process to suppress the influence of the optical fiber winding length on the optical fiber temperature measurement.

10. An electronic device, It is characterized in that The electronic device includes: processor and machine-readable storage medium; The machine-readable storage medium stores computer instructions, and when the computer instructions are executed by a processor, the steps in any one of the methods of claims 1 to 8 are implemented.

Citation Information

Patent Citations

  • Distributed fiber temperature measurement method under complex environment

    CN106248245A

  • Raman scattering light double-path demodulation method for optical fiber distributed temperature measurement

    CN112683418A