A safety monitoring system and method for urban integrated pipe gallery

By using a safety monitoring system combined with ultra-narrow linewidth laser and deep learning BP neural network in urban integrated pipelines, the problem of low safety monitoring efficiency of pipeline corridors is solved, and high accuracy and high efficiency vibration and temperature monitoring are achieved. Reduces costs and improves the ability to identify security incidents.

CN114018334BActive Publication Date: 2024-11-22HC OPTICAL SCI & TECH CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202111325388.9
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2021-11-10
Publication Date
2024-11-22
Estimated Expiration
2041-11-10

AI Technical Summary

Technical Problem

In the prior art, the safety monitoring efficiency of urban comprehensive pipeline corridors is low, making it difficult to achieve real-time monitoring and rapid response. Especially when underground pipeline accidents occur, they face huge risks from personnel and equipment and facilities, and the management is complex, so they are prone to blind inspection areas. .

Method used

A comprehensive urban pipeline safety monitoring system including an ultra-narrow linewidth laser, an optical pulse modulator, a first ring, a first splitter, a balanced light detector and a data acquisition card is adopted, combining a deep learning BP neural network and Coherent modulation technology realizes vibration and temperature monitoring, and uses wavelength division multiplexer to monitor simultaneously on a single device to reduce costs.

Benefits of technology

The safety monitoring accuracy rate has been improved to more than 96%, the deployment efficiency has been improved, and vibration monitoring with a maximum sensing distance of 52 kilometers and a spatial resolution of ±3 meters has been achieved, reducing the system false alarm rate and achieving practical level.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN114018334B_ABST
    Figure CN114018334B_ABST
Patent Text Reader

Abstract

The present invention relates to a safety monitoring system and method for an urban integrated pipe gallery, the safety monitoring system for an urban integrated pipe gallery comprising an ultra-narrow linewidth laser, an optical pulse modulator, a first circulator, a first splitter, a balanced light detector and a data acquisition card, wherein the ultra-narrow linewidth laser, the optical pulse modulator and the first port of the first circulator are connected in sequence, and the second port of the first circulator is connected to an optical cable; the third end of the first splitter, the first splitter, the balanced light detector and the data acquisition card are connected in sequence, and the first splitter is also connected to the ultra-narrow linewidth laser. Compared with the prior art, the present invention improves the accuracy and deployment efficiency of safety monitoring; realizes simultaneous vibration monitoring and temperature monitoring, and can effectively reduce costs.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of pipe gallery safety monitoring, and in particular to a system and method for monitoring the safety of an urban integrated pipe gallery. Background Art

[0002] With the rapid development of economy and society, security issues have become increasingly prominent, and people's awareness and requirements for security are getting higher and higher. In particular, the security of oil and gas pipelines and urban integrated pipeline corridors is particularly important. Once an intrusion or sabotage occurs, it will cause significant losses to the people, enterprises, society and even the country.

[0003] The integrated pipe gallery is an underground urban pipeline integrated corridor, that is, a tunnel space is built underground in the city, which integrates various engineering pipelines such as electricity, communications, gas, heating, water supply and drainage, and is equipped with special inspection ports, lifting ports and monitoring systems. The underground integrated pipe gallery has many advantages, making it a major trend in future urban infrastructure construction. The integrated pipe gallery system faces very complex management and becomes very difficult due to its multi-functional integration. Daily inspections are inefficient and prone to blind inspection areas. In particular, a more intelligent management method is needed for public safety hazards caused by abnormal temperature changes and external force damage in the pipeline.

[0004] The biggest advantage of the existing integrated pipe gallery is that it can construct and maintain underground pipelines without breaking the road, which means that most of the maintenance personnel's work must be done underground. The trunk lines and branches in the underground integrated pipe gallery are distributed vertically and horizontally, with long distances, sparse personnel, dim light, and poor ventilation. Once an accident occurs in the underground pipeline, the safety of personnel and equipment facilities will face huge risks. Therefore, a system is urgently needed to carry out real-time monitoring and safety protection of the operation status of professional pipelines in the underground integrated pipe gallery, improve the rapid response and safety prevention and control capabilities of the operation management of the urban underground integrated pipe gallery, and ensure the safety of the underground integrated pipe gallery. Summary of the invention

[0005] The present invention aims to solve at least one of the above technical problems in the prior art to a certain extent. To this end, one purpose of the present invention is to propose a system and method for urban integrated pipe gallery safety monitoring that improves the accuracy of safety monitoring and deployment efficiency, realizes simultaneous vibration monitoring and temperature monitoring, and can effectively reduce costs.

[0006] The technical solution of the present invention to solve the above technical problems is as follows: a city integrated pipeline corridor safety monitoring system, including an ultra-narrow linewidth laser, an optical pulse modulator, a first circulator, a first splitter, a balanced light detector and a data acquisition card, the ultra-narrow linewidth laser, the optical pulse modulator and the first port of the first circulator are connected in sequence, and the second port of the first circulator is connected to the optical cable; the third port of the first splitter, the first splitter, the balanced light detector and the data acquisition card are connected in sequence, and the first splitter is also connected to the ultra-narrow linewidth laser.

[0007] Based on the above technical solution, the present invention can also be improved as follows.

[0008] Furthermore, it also includes a fiber laser, a second circulator and an avalanche light detector, wherein the fiber laser is connected to the third port of the second circulator; the input end of the avalanche light detector is connected to the first port of the second circulator, and the output end thereof is connected to the data acquisition card; and the second port of the second circulator is connected to the optical cable.

[0009] Furthermore, it also includes a wavelength division multiplexer, and the second port of the first circulator and the second port of the second circulator are both connected to the optical cable through the wavelength division multiplexer.

[0010] Furthermore, it also includes an optical amplifier, a second splitter and a Raman amplifier, which are connected in series between the optical pulse modulator and the first circulator in sequence, the input end of the optical amplifier is connected to the output end of the optical pulse modulator, and the output end of the Raman amplifier is connected to the first port of the first circulator.

[0011] The beneficial effects of the present invention are: using deep learning BP neural network to process signals, the accuracy of safety monitoring is improved to more than 96%, and the efficiency of safety monitoring deployment is improved; using wavelength division multiplexer to realize vibration monitoring and temperature monitoring on a single-core optical cable of a single device, which can effectively reduce the cost of safety monitoring; also using coherent modulation technology to achieve a maximum vibration monitoring sensing distance of 52 kilometers and a spatial resolution of ±3 meters; accurate identification of safety events in complex scenarios is achieved, and the system false alarm rate is reduced from 1 time / day to 1 time / month, reaching a practical level.

[0012] Another technical solution of the present invention to solve the above technical problem is as follows: A method for monitoring the safety of an urban integrated pipe gallery comprises the following steps:

[0013] The ultra-narrow linewidth laser sends a first laser signal to the optical pulse modulator and the first splitter respectively, the optical pulse modulator modulates the first laser signal into a stable pulse laser signal and transmits it to the first circulator, and the third port of the first circulator transmits the stable pulse laser signal to the optical cable;

[0014] The second port of the first circulator receives the back Rayleigh scattered light signal in the optical cable, and the back Rayleigh scattered light signal is transmitted to the first splitter through the second port of the first circulator; the first splitter performs optical mixing on the back Rayleigh scattered light signal and the first laser signal to generate a mixed signal, which is transmitted to the balanced light detector; the balanced light detector performs photoelectric conversion on the mixed signal to generate a coherent mixed electrical signal, which is transmitted to the data acquisition card; the data acquisition card pre-processes the coherent mixed electrical signal through a deep learning BP neural network to obtain a data frame, and uses a data processing algorithm to process the data frame to obtain vibration data.

[0015] Further, the method further comprises the following steps:

[0016] The optical fiber laser sends a second laser signal to the third port of the second circulator, and the second port of the second circulator transmits the second laser signal to the optical cable; the second port of the second circulator receives the Raman scattered light signal in the optical cable, and transmits the Raman scattered light signal to the avalanche light detector, and the avalanche light detector performs photoelectric conversion on the Raman scattered light signal to generate a Raman scattered electrical signal and transmits it to the data acquisition card, and the data acquisition card pre-processes the Raman scattered electrical signal through a deep learning BP neural network to obtain a data frame, and uses a data processing algorithm to process the data frame to obtain temperature data.

[0017] Furthermore, the training method of the deep learning BP neural network comprises the following steps:

[0018] Format the collected data, build a weight matrix to weight the data, and perform cross-validation to obtain the initial model; determine whether the initial model is the optimal model. When the initial model is determined to be the optimal model, complete the deep learning BP neural network model;

[0019] When it is determined that the initial model is not the optimal model, the error of the data weight is calculated, the gradient descent method is used to calculate the gradient descent change of the error, the data weight is updated according to the gradient descent change of the error, and then the initial model is cross-validated.

[0020] Further, the data processing algorithm comprises the following steps:

[0021] Parse the data frame to obtain multiple groups of data, reorganize the data in the corresponding sorting position in each group of data to obtain multiple groups of new data, and calculate the standard deviation σ of each group of new data 1 , the standard deviation σ 1 Difference is performed with the data of the corresponding sorting position of each group of data in the previous data frame to obtain the difference matrix D 1 ; Through the difference matrix D 1 Draw the vibration curve.

[0022] Furthermore, the data acquisition card performs Fourier transform on the vibration data or temperature data:

[0023]

[0024] Among them, F(ω) is the image function of f(t); ω is the frequency, t is the time, e -iωt is a complex function.

[0025] The beneficial effects of the present invention are: using deep learning BP neural network to process signals, the accuracy of safety monitoring is improved to more than 96%, and the efficiency of safety monitoring deployment is improved; using wavelength division multiplexer to realize vibration monitoring and temperature monitoring on a single-core optical cable of a single device, which can effectively reduce the cost of safety monitoring; also using coherent modulation technology to achieve a maximum vibration monitoring sensing distance of 52 kilometers and a spatial resolution of ±3 meters; accurate identification of safety events in complex scenarios is achieved, and the system false alarm rate is reduced from 1 time / day to 1 time / month, reaching a practical level. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] Figure 1 This is a module block diagram of a city comprehensive pipe gallery safety monitoring system of the present invention;

[0027] Figure 2 A flow chart of a method for monitoring safety of an urban integrated pipe gallery according to the present invention;

[0028] Figure 3 Another flow chart of a method for monitoring safety of an urban integrated pipe gallery according to the present invention;

[0029] Figure 4 The flowchart of the training method of the deep learning BP neural network of the present invention.

[0030] In the accompanying drawings, the components represented by the reference numerals are listed as follows:

[0031] 1. Ultra-narrow linewidth laser, 2. Optical pulse modulator, 3. First circulator, 4. First splitter, 5. Balanced light detector, 6. Data acquisition card, 7. Optical cable, 8. Fiber laser, 9. Second circulator, 10. Avalanche light detector, 11. Wavelength division multiplexer, 12. Optical amplifier, 13. Second splitter, 14. Raman amplifier. DETAILED DESCRIPTION

[0032] The principles and features of the present invention are described below in conjunction with the accompanying drawings. The examples given are only used to explain the present invention and are not used to limit the scope of the present invention.

[0033] Embodiment 1:

[0034] like Figure 1As shown, a city integrated pipe gallery safety monitoring system includes an ultra-narrow linewidth laser 1, an optical pulse modulator 2, a first circulator 3, a first splitter 4, a balanced light detector 5 and a data acquisition card 6. The ultra-narrow linewidth laser 1, the optical pulse modulator 2 and the first port of the first circulator 3 are connected in sequence, and the second port of the first circulator 3 is connected to the optical cable 7; the third port of the first splitter 4, the first splitter 4, the balanced light detector 5 and the data acquisition card 6 are connected in sequence, and the first splitter 4 is also connected to the ultra-narrow linewidth laser 1.

[0035] The above embodiment also includes a fiber laser 8, a second circulator 9 and an avalanche light detector 10, wherein the fiber laser 8 is connected to the third port of the second circulator 9; the input end of the avalanche light detector 10 is connected to the first port of the second circulator 9, and the output end thereof is connected to the data acquisition card 6; and the second port of the second circulator 9 is connected to the optical cable 7.

[0036] The above embodiment further includes a wavelength division multiplexer 11 , and the second port of the first circulator 3 and the second port of the second circulator 9 are both connected to the optical cable 7 through the wavelength division multiplexer 11 .

[0037] The above embodiment also includes an optical amplifier 12, a second splitter 13 and a Raman amplifier 14, and the optical amplifier 12, the second splitter 13 and the Raman amplifier 14 are connected in series between the optical pulse modulator 2 and the first circulator 3 in sequence, the input end of the optical amplifier 12 is connected to the output end of the optical pulse modulator 2, and the output end of the Raman amplifier 14 is connected to the first port of the first circulator 3.

[0038] In this embodiment, 90% of the light from the ultra-narrow linewidth laser 1 is used as the initial signal to enter the optical pulse modulator 2, and the optical pulse modulator 2 modulates the first laser signal into a stable pulse laser signal that can be controlled by software. The pulse laser signal enters the second splitter 13 after being amplified by the optical amplifier 12. The second splitter 13 is a 50:50 splitter. The second splitter 13 divides the pulse laser signal into two beams, one for expansion and reservation, and the other enters the first circulator 3 after being amplified by the Raman amplifier 14; the first circulator 3 transmits the pulse laser signal to the optical cable 7 through the second port, and the optical cable 7 is a sensor optical cable;

[0039] When sound waves or vibrations are applied to the optical cable 7, the optical cable 7 will produce a slight deformation. The material density will change at the position where the optical cable 7 is deformed, which will in turn affect the refractive index of the optical cable 7 at that position. The backscattered Rayleigh scattered light is affected by the refractive index and its intensity will change. At the same time, the length of the optical cable 7 is also changed due to the deformation, so that the Rayleigh scattering phase also changes. A part of the backscattered Rayleigh scattered light signal will return along the optical cable 7 to the second port of the first circulator 3, and enter the first splitter 4 through the third port of the first circulator 3, wherein the first splitter 4 is a 2×2 splitter. The first splitter 4 also 10% light of the ultra-narrow linewidth laser 1 is connected from the ultra-narrow linewidth laser 1 as a reference signal. Since the initial signal and the reference signal are signals separated by the ultra-narrow linewidth laser 1, the initial signal and the reference signal are strongly coherent light. The back Rayleigh scattered light signal and the reference signal enter the first splitter 4 for optical mixing, and then are sent to the balanced light detector 5 for photoelectric conversion to generate a coherent mixing electrical signal that is transmitted to the data acquisition card 6. The data acquisition card 6 pre-processes the coherent mixing electrical signal through a deep learning BP neural network to obtain a data frame, and uses a data processing algorithm to process the data frame to obtain vibration data.

[0040] The optical fiber laser 8 sends a second laser signal to the third port of the second circulator 9, and the second port of the second circulator 9 transmits the second laser signal to the optical cable 7; the second port of the second circulator 9 receives the Raman scattered light signal in the optical cable 7, and transmits the Raman scattered light signal to the avalanche light detector 10, and the avalanche light detector 10 performs photoelectric conversion on the Raman scattered light signal to generate a Raman scattered electrical signal and transmits it to the data acquisition card 6, and the data acquisition card 6 pre-processes the Raman scattered electrical signal through the deep learning BP neural network to obtain a data frame, and uses the data processing algorithm to process the data frame to obtain temperature data.

[0041] This embodiment uses a deep learning BP neural network to process signals, thereby improving the accuracy of safety monitoring to more than 96% and improving the efficiency of safety monitoring deployment; uses a wavelength division multiplexer 11 to separate the optical cable into two beams for vibration sensing and temperature sensing respectively, thereby achieving simultaneous vibration monitoring and temperature monitoring on a single-core optical cable of a single device, which can effectively reduce the cost of safety monitoring; and also uses coherent modulation technology to achieve a maximum vibration monitoring sensing distance of 52 kilometers and a spatial resolution of ±3 meters.

[0042] Embodiment 2:

[0043] like Figure 2 and Figure 3 As shown, a method for monitoring the safety of an urban integrated pipe gallery comprises the following steps:

[0044] The ultra-narrow linewidth laser 1 sends a first laser signal to the optical pulse modulator 2 and the first splitter 4 respectively. The optical pulse modulator 2 modulates the first laser signal into a stable pulse laser signal and transmits it to the first circulator 3. The third port of the first circulator 3 transmits the stable pulse laser signal to the optical cable 7.

[0045] The second port of the first circulator 3 receives the back Rayleigh scattered light signal in the optical cable 7, and the back Rayleigh scattered light signal is transmitted to the first splitter 4 through the second port of the first circulator 3; the first splitter 4 performs optical mixing on the back Rayleigh scattered light signal and the first laser signal to generate a mixed signal, which is transmitted to the balanced light detector 5, and the balanced light detector 5 performs photoelectric conversion on the mixed signal to generate a coherent mixed electrical signal, which is transmitted to the data acquisition card 6, and the data acquisition card 6 pre-processes the coherent mixed electrical signal through a deep learning BP neural network to obtain a data frame, and uses a data processing algorithm to process the data frame to obtain vibration data.

[0046] In the above embodiment, the following steps are also included:

[0047] The optical fiber laser 8 sends a second laser signal to the third port of the second circulator 9, and the second port of the second circulator 9 transmits the second laser signal to the optical cable 7; the second port of the second circulator 9 receives the Raman scattered light signal in the optical cable 7, and transmits the Raman scattered light signal to the avalanche light detector 10, and the avalanche light detector 10 performs photoelectric conversion on the Raman scattered light signal to generate a Raman scattered electrical signal and transmits it to the data acquisition card 6, and the data acquisition card 6 pre-processes the Raman scattered electrical signal through the deep learning BP neural network to obtain a data frame, and uses the data processing algorithm to process the data frame to obtain temperature data.

[0048] In the above embodiment, if Figure 4 As shown, the training method of the deep learning BP neural network includes the following steps:

[0049] Format the collected data, build a weight matrix to weight the data, and perform cross-validation to obtain the initial model; determine whether the initial model is the optimal model. When the initial model is determined to be the optimal model, complete the deep learning BP neural network model;

[0050] When it is determined that the initial model is not the optimal model, the error of the data weight is calculated, the gradient descent method is used to calculate the gradient descent change of the error, the data weight is updated according to the gradient descent change of the error, and then the initial model is cross-validated.

[0051] In this embodiment, the deep learning BP neural network is divided into input, hidden and output layers. The sensor data is input from the input layer, analyzed by the hidden layer, and the results are output from the output layer. Among them, the hidden layer may contain multiple layers, and each layer can be sequentially divided into L 0 To L n Sort in order.

[0052] If there are n data, then:

[0053] w 1 s 1 +w 2 s 2 +...+w n s n =M

[0054] Where w is the weight, s is the sample data, and M is the pattern to be recognized.

[0055] In the deep learning BP neural network, each layer corresponds to a matrix A Ln and the previous layer A Ln-1 (n≥1), then:

[0056] A Ln =[ln 1 ln 2 … ln m ]

[0057] A Ln-1 =[ln-1 1 ln-1 2 … ln-1 n ]

[0058] A Ln Each element in represents the neuron of this layer. When n=1, that is, A L0 In , the elements represent the data input into the neural network.

[0059] Based on the above, an m×n weight matrix can be established

[0060]

[0061] L n Layer Matrix A Ln L n-1 Layers and Weight Matrices The dot product is:

[0062]

[0063] The output of the deep learning BP neural network is the probability of each event, so when L 1Finally, we need to perform logistic regression on it. In the algorithm, we use the Sigmoid function to map the output of the neural network to 0 to 1.

[0064] The prototype of the Sigmoid function is:

[0065]

[0066] but:

[0067]

[0068] For an n-layer neural network, we can The pattern of each layer of neurons is obtained in turn, and finally L n Layer Matrix A Ln . A Ln Each element in is the probability of pattern recognition. This process is cross-validation, also known as forward derivation. However, for a newly initialized neural network, due to the error in the initial weights, 0 Layer to L n In the positive direction of the layer, the error gradually accumulates, resulting in L n The output pattern recognition results have deviations, so the error needs to be calculated. n , during training, the data samples determine the pattern matrix A M , so the error

[0069] Since the error comes from the forward propagation of the weight error in the neural network, the error needs to be used as a parameter, and the gradient descent method is used to gradually approach the local optimal solution. Suppose the change in the weight adjustment is By gradient descent method, we can get:

[0070]

[0071] but:

[0072]

[0073] Among them, Sigmoid′(A Ln ) is A Ln Each element in Sigmoid(A Ln ). That is:

[0074]

[0075] Change yes The rate of change of The local optimal solution in the current state is updated as follows

[0076]

[0077] For the non-output layer, since the error is transmitted layer by layer, the error calculation of its neurons needs to pass through the weight of the next layer. and its correction Reverse deduction. Similarly, through the gradient descent method, the previous layer L n-1 Error For L n Layer Weights and its variation The inner product of is:

[0078]

[0079] Then L n-1 Layer weight change for:

[0080]

[0081] And by Updatable

[0082]

[0083] The algorithm is implemented by L n The error update L of the layer n weights, and through the propagation of errors in the neural network, from L n Layer to L n-1 The error of each layer is calculated by gradient descent. This process is called back propagation.

[0084] Backpropagation will update the neural network After the update is completed, the algorithm starts from L 0 Input the data sample determined by the new pattern and deduce it forward again. The new error will be corrected again through the back propagation process. After repeated training, a set of optimal solutions is finally obtained. This optimal solution This is the model of deep learning BP neural network.

[0085] In the above embodiment, the data processing algorithm includes the following steps:

[0086] Parse the data frame to obtain multiple groups of data, reorganize the data in the corresponding sorting position in each group of data to obtain multiple groups of new data, and calculate the standard deviation σ of each group of new data 1 , the standard deviation σ 1Difference is performed with the data of the corresponding sorting position of each group of data in the previous data frame to obtain the difference matrix D 1 ; Through the difference matrix D 1 Draw the vibration curve.

[0087] In this embodiment, for ease of understanding, it can be considered that there are several sensing points evenly distributed in the optical cable. Each sensing point can return the optical signal data of the point; 100 sets of data in the data frame are taken for analysis, and the standard deviation σ of the first sensing point of these 100 sets of data is calculated. 1 That is, take the first element of 100 sets of data to form a new set of data, and calculate its standard deviation.

[0088] For example, the data of the first sensor point at the current time T forms the matrix A 1 , then:

[0089] (Subscript 1-1 is the first data in the first group of data in the data frame, and so on);

[0090] For each of the formulas The lower 16 bits of phase data are taken, and the standard deviation σ corresponding to the first sensing point is 1 for:

[0091]

[0092] Similarly, there is a corresponding data frame at the previous time T-1;

[0093] σ 1 The first sensor point matrix A′ of the previous data frame 1 Difference, and get the difference matrix D 1 :

[0094]

[0095] Difference matrix D 1 That is, it reflects the vibration of the first sensing point within a time slice. In the ideal state without vibration, D 1 The elements are kept between 2 and -2 to maintain dynamic balance, that is:

[0096] D 1 =[-0.6 0.2 0.9 -2 0 0.4 ... -0.2]

[0097] The vibration curve is drawn through the difference matrix. The difference matrix represents the data change of a sensing point, that is, the vibration change of a certain point in the optical cable.

[0098] In the above embodiment, the data acquisition card 6 performs Fourier transform on the vibration data or temperature data:

[0099]

[0100] Among them, F(ω) is the image function of f(t); ω is the frequency, t is the time, e -iωt is a complex function.

[0101] In this embodiment, the data acquisition card 6 cannot directly capture the frequency domain characteristics of the signal, and data analysis requires the frequency domain characteristics of the data. Therefore, the data acquisition card 6 needs to perform Fourier transform on the original data.

[0102] This embodiment uses a deep learning BP neural network to process signals, thereby improving the accuracy of safety monitoring to more than 96% and improving the efficiency of safety monitoring deployment; uses a wavelength division multiplexer 11 to separate the optical cable into two beams for vibration sensing and temperature sensing respectively, thereby achieving simultaneous vibration monitoring and temperature monitoring on a single-core optical cable of a single device, which can effectively reduce the cost of safety monitoring; and also uses coherent modulation technology to achieve a maximum vibration monitoring sensing distance of 52 kilometers and a spatial resolution of ±3 meters.

[0103] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A city integrated pipe gallery safety monitoring system, characterized by: The invention comprises an ultra-narrow linewidth laser (1), an optical pulse modulator (2), a first circulator (3), a first splitter (4), a balanced light detector (5) and a data acquisition card (6), wherein the ultra-narrow linewidth laser (1), the optical pulse modulator (2) and the first port of the first circulator (3) are connected in sequence, and the second port of the first circulator (3) is connected to an optical cable (7); the third port of the first splitter (4), the first splitter (4), the balanced light detector (5) and the data acquisition card (6) are connected in sequence, and the first splitter (4) is also connected to the ultra-narrow linewidth laser (1); The data acquisition card (6) pre-processes the coherent frequency mixing electrical signal through a deep learning BP neural network to obtain a data frame, and processes the data frame using a data processing algorithm to obtain vibration data; The data processing algorithm comprises the following steps: The data frame is parsed to obtain multiple groups of data, and the data at the corresponding sorting position in each group of data are reorganized to obtain multiple groups of new data. The standard deviation σ1 of each group of new data is calculated, and the standard deviation σ1 is differentiated from the data at the corresponding sorting position of each group of data in the previous data frame to obtain the difference matrix D1; the vibration curve is drawn through the difference matrix D1.

2. The urban integrated pipe gallery safety monitoring system according to claim 1 is characterized in that: It also comprises a fiber laser (8), a second circulator (9) and an avalanche light detector (10), wherein the fiber laser (8) is connected to the third port of the second circulator (9); the input end of the avalanche light detector (10) is connected to the first port of the second circulator (9), and the output end thereof is connected to the data acquisition card (6); and the second port of the second circulator (9) is connected to the optical cable (7).

3. The urban integrated pipe gallery safety monitoring system according to claim 2 is characterized in that: It also comprises a wavelength division multiplexer (11), wherein the second port of the first circulator (3) and the second port of the second circulator (9) are both connected to the optical cable (7) via the wavelength division multiplexer (11).

4. The urban integrated pipe gallery safety monitoring system according to claim 1 is characterized in that: The optical amplifier (12), the second splitter (13) and the Raman amplifier (14) are connected in series between the optical pulse modulator (2) and the first circulator (3) in sequence; the input end of the optical amplifier (12) is connected to the output end of the optical pulse modulator (2), and the output end of the Raman amplifier (14) is connected to the first port of the first circulator (3).

5. A method for monitoring the safety of an urban integrated pipe gallery, characterized in that: The following steps are involved: The ultra-narrow linewidth laser (1) sends a first laser signal to the optical pulse modulator (2) and the first splitter (4) respectively; the optical pulse modulator (2) modulates the first laser signal into a stable pulse laser signal and transmits it to the first circulator (3); the third port of the first circulator (3) transmits the stable pulse laser signal to the optical cable (7); The second port of the first circulator (3) receives the back Rayleigh scattered light signal in the optical cable (7), and the back Rayleigh scattered light signal is transmitted to the first splitter (4) through the second port of the first circulator (3); the first splitter (4) performs optical mixing on the back Rayleigh scattered light signal and the first laser signal to generate a mixed signal, which is transmitted to the balanced light detector (5); the balanced light detector (5) performs photoelectric conversion on the mixed signal to generate a coherent mixed electrical signal, which is transmitted to the data acquisition card (6); the data acquisition card (6) pre-processes the coherent mixed electrical signal through a deep learning BP neural network to obtain a data frame, and processes the data frame using a data processing algorithm to obtain vibration data; The data processing algorithm comprises the following steps: The data frame is parsed to obtain multiple groups of data, and the data at the corresponding sorting position in each group of data are reorganized to obtain multiple groups of new data. The standard deviation σ1 of each group of new data is calculated, and the standard deviation σ1 is differentiated from the data at the corresponding sorting position of each group of data in the previous data frame to obtain the difference matrix D1; the vibration curve is drawn through the difference matrix D1.

6. The urban integrated pipe gallery safety monitoring method according to claim 5 is characterized in that: The following steps are also included: The optical fiber laser (8) sends a second laser signal to the third port of the second circulator (9), and the second port of the second circulator (9) transmits the second laser signal to the optical cable (7); the second port of the second circulator (9) receives the Raman scattered light signal in the optical cable (7), and transmits the Raman scattered light signal to the avalanche light detector (10); the avalanche light detector (10) performs photoelectric conversion on the Raman scattered light signal to generate a Raman scattered electric signal, which is transmitted to the data acquisition card (6); the data acquisition card (6) pre-processes the Raman scattered electric signal through a deep learning BP neural network to obtain a data frame, and processes the data frame using a data processing algorithm to obtain temperature data.

7. The urban integrated pipe gallery safety monitoring method according to claim 5 or 6, characterized in that: The training method of the deep learning BP neural network comprises the following steps: Format the collected data, build a weight matrix to weight the data, and perform cross-validation to obtain the initial model; determine whether the initial model is the optimal model. When the initial model is determined to be the optimal model, complete the deep learning BP neural network model; When it is determined that the initial model is not the optimal model, the error of the data weight is calculated, the gradient descent change of the error is calculated using the gradient descent method, the data weight is updated according to the gradient descent change of the error, and then the initial model is cross-validated.

8. The urban integrated pipe gallery safety monitoring method according to claim 5 or 6, characterized in that: The data acquisition card (6) performs Fourier transform on the vibration data or temperature data: Among them, F(ω) is the image function of f(t); ω is the frequency, t is the time, e -iωt is a complex function.