A system and method for identifying vibration signal types based on POTDR

Abstract

The present invention discloses a system and method for identifying vibration signal types based on POTDR, and relates to the field of optical fiber sensing technology. The system includes a laser, a circulator, an optical fiber to be tested, a polarization beam splitter, a photoelectric converter assembly, a signal collector, and a signal processing unit. The method includes obtaining measurement results of vibration signal characteristics based on POTDR demodulation technology when a vibration signal generated by the surrounding environment affects the optical fiber to be tested; collecting vibration signals at any position, establishing a data relationship library of soil environment, soil depth, and vibration signals; based on the established data relationship library, identifying vibration signal characteristics in different soil environments and at different depths, and then identifying the vibration signal type. The present invention is based on POTDR's IQ demodulation technology and integrates soil medium characteristics and burial depth to improve the accuracy of vibration signal type identification.

Description

Technical Field

[0001] The present invention relates to the field of optical fiber sensing technology, and in particular to a system and method for identifying vibration signal types based on POTDR. Background Art

[0002] Fiber optic sensing technology, due to its advantages such as resistance to electromagnetic interference, corrosion resistance, and long transmission distance, has found widespread application in fields such as vibration measurement and structural health monitoring. Existing methods for identifying vibration types using fiber optic sensing technology include support vector machines, deep learning models, and unsupervised learning methods. These methods typically directly build a model library based on the collected data, train the model, and obtain the identification type of the vibration signal. However, establishing a model between the vibration signal and the identification type lacks effective utilization of environmental information. Therefore, it is necessary to design a system and method for identifying vibration signal types based on POTDR that can fully utilize environmental information and improve the accuracy of vibration identification. Summary of the Invention

[0003] The present invention aims to solve the problems in the prior art and provides a system and method for identifying the type of vibration signals based on POTDR.

[0004] To achieve the above object, the technical solution adopted by the present invention is as follows:

[0005] A system for identifying vibration signal types based on POTDR includes a laser, a circulator, an optical fiber to be tested, a polarization beam splitter, a photoelectric converter assembly, a signal collector, and a signal processing unit. Pulsed light emitted by the laser enters the circulator and is modulated by the optical fiber to be tested. When the optical signal scattered back by the optical fiber to be tested is output by the circulator, it reaches the polarization beam splitter and is transmitted to the photoelectric converter assembly. The photoelectric converter assembly converts the optical signal into an electrical signal, which is then sent to the signal collector and finally to the signal processing unit.

[0006] Based on the above technical solution, further, the photoelectric converter assembly includes a first photoelectric converter and a second photoelectric converter, wherein the first photoelectric converter converts the received optical signal of the X vector into an electrical signal in the X-axis direction, and the second photoelectric converter converts the received optical signal of the Y vector into an electrical signal in the Y-axis direction.

[0007] Based on the above technical solution, further, the polarization beam splitter decomposes the received signal into an X-vector optical signal and a Y-vector optical signal, and transmits the X-vector optical signal to the first photoelectric converter and transmits the Y-vector optical signal to the second photoelectric converter.

[0008] Based on the above technical solution, further, the signal collector uses a digital coherent IQ demodulation algorithm to calculate the angle of polarized light from the polarization signal collected at each position of the sensing optical fiber, and sends the calculated angle to the signal processing unit.

[0009] A method for identifying vibration signal types based on POTDR comprises the following steps: step S1, when a vibration signal generated by the surrounding environment affects an optical fiber to be tested, obtaining a measurement result of the vibration signal characteristics based on the POTDR demodulation technology; step S2, collecting vibration signals at any position, and establishing a data relationship library of soil environment, soil depth and vibration signals; step S3, based on the established data relationship library, identifying the vibration signal characteristics in different soil environments and different depths, and then identifying the vibration signal type.

[0010] Based on the above technical solution, further, in step S1, the acquisition process is: the parameter signal of the polarized light carried by the received backscattered Rayleigh scattering echo of each section of optical fiber is solved by the signal processing unit, and finally the measurement result of the vibration signal characteristic is obtained.

[0011] Based on the above technical solution, further, in step S2, the soil environment is set to a general soil environment and a high-water content environment, wherein the general soil environment is set to low-water content soil with a water content between 11% and 13%, and the high-water content environment is set to high-water content soil with a water content between 18% and 20%; the soil depth is set to 2 meters, 5 meters, 10 meters, 15 meters, and 20 meters; and the parameters of the vibration signal include amplitude, frequency, and phase.

[0012] Based on the above technical solution, further, in step S3, the recognition process includes: step S31, first use the support vector machine to train the relationship model A between the vibration signal characteristics and the soil environment and soil depth; step S32, then use the LSTM model in deep learning to establish the category model B of the time domain characteristics of the vibration signal and the signal category; step S33, when a new vibration signal is collected, the trained relationship model A is used to determine which environmental category the vibration signal belongs to; the category model B determines which environmental condition to select the category model B according to the environmental category, thereby identifying the signal type.

[0013] Based on the above technical solution, further, in step S31, the process is: step A1, setting the vibration signal characteristics to [amplitude, frequency, phase]; step A2, combining the soil environment and soil depth; step A3, cleaning and filtering the collected vibration signal data, and using a support vector machine to establish a relationship model A between the vibration signal characteristic data and the environmental category.

[0014] Based on the above technical solution, further, in step S32, the process is: step B1, determine the environmental conditions based on the combination of soil environment and soil depth; step B2, use the LSTM model to train the category model B of time domain features and signal types based on the environmental conditions.

[0015] Compared with the prior art, the present invention has the following beneficial effects:

[0016] This invention uses POTDR technology to accurately identify vibration signal types by matching soil medium properties. It primarily builds a POTDR-based IQ demodulation technique that incorporates soil medium properties and burial depth to improve the accuracy of vibration signal type identification. Furthermore, by fully utilizing environmental information and accurately selecting a model that corresponds to this environmental information, the identification of vibration signal types can be improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 Schematic diagram of the structure of the system of the present invention;

[0018] Figure 2 This is a demodulation flow chart of the digital coherent IQ demodulator in the present invention;

[0019] Figure numerals: 111, laser; 112, circulator; 113, optical fiber to be tested; 114, polarization beam splitter; 115, first photoelectric converter; 116, second photoelectric converter; 117, signal collector; 118, signal processing unit; 119, multiplier; 120, low-pass filter; 121, divider. DETAILED DESCRIPTION

[0020] The present invention will be further described and illustrated below with reference to the accompanying drawings and specific embodiments. The technical features of each embodiment of the present invention may be combined accordingly, provided that there is no conflict between them.

[0021] In order to make the above-mentioned objects, features and advantages of the present invention more clearly understood, the specific embodiments of the present invention are described in detail below with reference to the accompanying drawings. In the following description, many specific details are set forth to facilitate a full understanding of the present invention. However, the present invention can be implemented in many other ways than those described herein, and those skilled in the art can make similar improvements without violating the connotation of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below. The technical features in the various embodiments of the present invention can be combined accordingly without conflicting with each other.

[0022] In the description of the present invention, it should be understood that when an element is considered to be "connected" to another element, it can be directly connected to the other element or indirectly connected, that is, there are intermediate elements. On the contrary, when an element is said to be "directly" connected to another element, there are no intermediate elements.

[0023] Example 1

[0024] Combine Figure 1 As shown, this embodiment provides a system for identifying the type of vibration signal based on POTDR, including a laser 111, a circulator 112, an optical fiber to be tested 113, a polarization beam splitter 114, a photoelectric converter assembly, a signal collector 117, and a signal processing unit 118. The photoelectric converter assembly includes a first photoelectric converter 115 and a second photoelectric converter 116.

[0025] In this embodiment, the system's overall workflow is as follows: pulsed light emitted by laser 111 enters circulator 112 and is modulated by optical fiber under test 113. When the surrounding environment changes, such as during an earthquake, optical fiber under test 113 is modulated by external physical quantities, causing the polarization state of the light in the fiber to change. Because scattered light does not change the polarization properties of the incident light at that point, the optical signal scattered back from optical fiber under test 113 carries polarization state information. After being output from circulator 112, it reaches polarization beam splitter 114, which decomposes it into two orthogonal optical signals, one for the X vector and the other for the Y vector. These signals are then transmitted to first photoelectric converter 115 and second photoelectric converter 116, respectively. These converters convert the optical signals into electrical signals, which are then fed into signal acquisition unit 117. The polarization angle of the polarization signal collected at each position of the sensing fiber is calculated using a digital coherent IQ demodulation algorithm and sent to signal processing unit 118. This signal processing unit 118 calculates the relevant parameters of the polarization light carried by the Rayleigh backscattered echo from each section of the fiber.

[0026] Reference Figure 2 The schematic diagram of digital coherent IQ demodulation is shown in the figure. IQ demodulation constructs two orthogonal signals of the same frequency to simulate local oscillator signals LI and LQ based on the carrier frequency of the intermediate frequency signal: Figure 2 The input intermediate frequency IF is multiplied by LI and LQ respectively through multiplier 119, and then the high-frequency components are filtered out by low-pass filter 120 to output two orthogonal components: I component and Q component. Then, they pass through divider 121 and the inverse tangent is calculated based on the inverse tangent function to obtain the output Angle angle.

[0027] Example 2

[0028] Based on the identification system steps disclosed in Example 1, this embodiment provides a method for identifying the type of vibration signal based on POTDR, including the following steps:

[0029] Step S1, obtaining a vibration signal, specifically, when the vibration signal generated by the surrounding environment affects the optical fiber 113 to be tested, obtaining the measurement result of the vibration signal characteristic based on the POTDR demodulation technology.

[0030] In this embodiment, the acquisition process is as follows: when the vibration signal affects the optical fiber 113 under test through physical contact or transmission through a medium, the light wave in the optical fiber 113 under test is modulated, affecting the relevant parameters of the polarized light. The signal processing unit 118 calculates the relevant parameters of the polarized light carried by the backscattered Rayleigh scattered echo of each section of the optical fiber under test, and finally obtains the measurement results of the vibration signal characteristics. It should be noted that POTDR is a fully distributed optical fiber sensor based on optical time domain reflectometry technology, which performs sensing by detecting changes in the polarization state of scattered light waves in the optical fiber.

[0031] Furthermore, the demodulation technology process based on POTDR is as follows: the optical signal scattered back by the optical fiber 113 to be tested is output through the circulator 112; the polarization beam splitter 114 decomposes the signal output by the circulator 112 into mutually orthogonal X vector optical signals and Y vector optical signals, and inputs them to the first photoelectric converter 115 and the second photoelectric converter 116; the first photoelectric converter 115 and the second photoelectric converter 116 convert the received optical signal into mutually orthogonal X electrical signals and Y electrical signals, and send them to the signal collector 117. It should be noted that the first photoelectric converter 115 is used to convert the X vector optical signal into a corresponding electrical signal, and the second photoelectric converter 116 is used to convert the Y vector optical signal into a corresponding electrical signal; the signal collector 117 inputs the obtained mutually orthogonal X electrical signals and Y electrical signals into a bandpass filter with a frequency of 20Hz-20kHz to obtain the I component and Q component, and then based on Figure 2 The demodulation process of the digital coherent IQ demodulator obtains the angle and outputs it.

[0032] It should be noted that the demodulation technology process based on POTDR is part of the acquisition process. The acquisition process can obtain the characteristics of the vibration signal through POTDR demodulation and some processing processes.

[0033] Step S2: collecting vibration signals at any position and establishing a data relationship database of soil environment, soil depth and vibration signals;

[0034] In this embodiment, when the soil environment is either normal or high-water content, and the soil depth is 2 meters, 5 meters, 10 meters, 15 meters, or 20 meters, vibration signals are collected at specific locations to establish a database of relationships between soil, depth, and vibration signals. This is shown in Table 1 below. It should be noted that the normal soil environment refers to low-water content soil, with a moisture content of approximately 12%, preferably 11%-13%, while the high-water content environment refers to high-water content soil, with a moisture content between 18% and 20%.

[0035] Table 1

[0036]

[0037] Step S3: Based on the established data relationship database, the vibration signal characteristics in different soil environments and at different depths are identified, and then the vibration signal type is identified.

[0038] Specifically, the entire recognition process is as follows: Step S31, first use the support vector machine to train the relationship model A between the vibration signal characteristics and the soil environment and soil depth. The specific method is:

[0039] Step A1: Set the vibration signal characteristics to [amplitude, frequency, phase];

[0040] Step A2: combining soil environment and soil depth, and setting the corresponding environment categories to 10 categories, specifically including: -2 meters under general environment, -5 meters under general environment, -10 meters under general environment, -15 meters under general environment, -20 meters under general environment, -2 meters under high water content environment, -5 meters under high water content environment, -10 meters under high water content environment, -15 meters under high water content environment, and -20 meters under high water content environment;

[0041] Step A3: Clean and filter the collected vibration signal data, and then use a support vector machine to establish a relationship model A between the vibration signal feature data and the environment category.

[0042] Step A4: When new data is collected, the model in step A3 is input to determine the corresponding soil environment and soil depth.

[0043] Step S32: Use the LSTM model in deep learning to establish a category model B of the time domain features of the vibration signal and the signal category. The specific method is as follows:

[0044] Step B1, determining the environmental condition according to the combination of soil environment and soil depth = [-2 meters under normal environment, -5 meters under normal environment, -10 meters under normal environment, -15 meters under normal environment, -20 meters under normal environment, -2 meters under high water content environment, -5 meters under high water content environment, -10 meters under high water content environment, -15 meters under high water content environment, -20 meters under high water content environment];

[0045] Step B2: Based on the environmental conditions, use the LSTM model to train the category model B of the time domain features and signal types, that is, train 10 models.

[0046] Step S33: When a new vibration signal is collected, the trained relationship model A can determine which environmental category the vibration signal belongs to; the category model B determines which environmental condition model to select based on the environmental category, thereby identifying the signal type and improving the recognition accuracy.

[0047] Finally, it should be noted that the above content is only used to illustrate the technical solution of the present invention, rather than to limit the scope of protection of the present invention. Simple modifications or equivalent substitutions of the technical solution of the present invention by ordinary technicians in this field do not deviate from the essence and scope of the technical solution of the present invention.

Claims

1. A method for identifying vibration signal types based on POTDR, characterized in that, A system for identifying vibration signal types based on POTDR is used, including the following steps: Step S1: When the vibration signal generated by the surrounding environment affects the optical fiber to be tested, the measurement result of the vibration signal characteristic is obtained based on the POTDR demodulation technology; Step S2: collecting vibration signals at any position and establishing a data relationship database of soil environment, soil depth and vibration signals; Step S3: Based on the established data relationship database, identifying the vibration signal characteristics in different soil environments and at different depths, and then identifying the vibration signal type; In step S3, the identification process includes: Step S31: First, use a support vector machine to train a relationship model A between vibration signal characteristics and soil environment and soil depth; Step S32: Use the LSTM model in deep learning to establish a category model B of the time domain features of the vibration signal and the signal category; Step S33: When a new vibration signal is collected, the trained relationship model A is used to determine which environmental category the vibration signal belongs to; the category model B determines which environmental condition category model B is selected based on the environmental category, thereby identifying the signal type; The system includes a laser, a circulator, an optical fiber to be tested, a polarization beam splitter, an optoelectronic converter assembly, a signal collector, and a signal processing unit. The pulsed light emitted by the laser enters the circulator and is modulated by the optical fiber to be tested. When the optical signal scattered back by the optical fiber to be tested is output by the circulator, it reaches the polarization beam splitter and is transmitted to the photoelectric converter component; the photoelectric converter component converts the optical signal into an electrical signal, sends it to the signal collector, and finally sends it to the signal processing unit.

2. a kind of method based on POTDR identification vibration signal type according to claim 1, is characterized in that, The photoelectric converter assembly includes a first photoelectric converter and a second photoelectric converter, wherein the first photoelectric converter converts the received X-vector optical signal into an electrical signal in the X-axis direction, and the second photoelectric converter converts the received Y-vector optical signal into an electrical signal in the Y-axis direction.

3. a kind of method based on POTDR identification vibration signal type according to claim 2, is characterized in that, The polarization beam splitter decomposes the received signal into an X-vector optical signal and a Y-vector optical signal, and transmits the X-vector optical signal to the first photoelectric converter and transmits the Y-vector optical signal to the second photoelectric converter.

4. a kind of method based on POTDR identification vibration signal type according to claim 1, is characterized in that, The signal collector uses a digital coherent IQ demodulation algorithm to calculate the angle of polarized light from the polarization signal collected at each position, and sends the calculated angle to the signal processing unit.

5. a kind of method based on POTDR identification vibration signal type according to claim 1, is characterized in that, In step S1, the acquisition process is: The signal processing unit calculates the parameter signal of the polarized light carried by the backscattered Rayleigh scattering echo of each section of the optical fiber to be tested, and finally obtains the measurement result of the vibration signal characteristics.

6. a kind of method based on POTDR identification vibration signal type according to claim 1, is characterized in that, In step S2, The soil environment is set as a general soil environment and a high water content environment, wherein the general soil environment is set as a low water content soil with a water content between 11% and 13%, and the high water content environment is set as a high water content soil with a water content between 18% and 20%; Set the soil depth to 2m, 5m, 10m, 15m, and 20m; The parameters of vibration signals include amplitude, frequency and phase.

7. a kind of method based on POTDR identification vibration signal type according to claim 1, is characterized in that, In step S31, the process is: Step A1: Set the vibration signal characteristics to [amplitude, frequency, phase]; Step A2, combining soil environment and soil depth; Step A3: Clean and filter the collected vibration signal data, and then use a support vector machine to establish a relationship model A between the vibration signal feature data and the environment category.

8. a kind of method based on POTDR identification vibration signal type according to claim 7, is characterized in that, In step S32, the process is: Step B1, determining environmental conditions based on a combination of soil environment and soil depth; Step B2: Based on environmental conditions, use the LSTM model to train the classification model B of time domain features and signal types.

Images