A double train state monitoring system and method based on Φ-OTDR

Abstract

The application discloses a double-train state monitoring system and method based on a Phi-OTDR, and the system comprises a Phi-OTDR distributed optical fiber sensing optical path, an AD acquisition module, an FPGA circuit and a host computer. The Rayleigh scattering light signal is generated through the distributed optical fiber sensing optical path, is digitized through photoelectric conversion and the AD acquisition module, and then the phase data is extracted through the FPGA circuit for digital quadrature demodulation. The phase data is processed through a Hilbert transform algorithm, and a mapping relationship between the phase difference and the vibration amplitude and frequency is established. The host computer receives the data and generates a time-position two-dimensional waterfall diagram, the running directions of the double trains are judged through the isotropy or reverse of the vibration point track, the displacement change amount and the time difference of the two trains are calculated based on a time series analysis algorithm, the speed data is optimized through a sliding average filtering and a Kalman filtering, and the real-time positions and speeds of the double trains are output. The application realizes accurate monitoring of the running states of the double trains, and improves the precision, real-time performance and reliability of railway safety monitoring.

Description

Technical Field

[0001] This invention relates to the field of distributed fiber optic vibration sensing technology, specifically to a dual-train condition monitoring system and method based on Φ-OTDR. Background Technology

[0002] Currently, with the rapid development of railway transportation and the continuous increase in train speeds, railway transportation safety has become a crucial issue. Therefore, real-time monitoring of train operating status, including location, speed, and direction of travel, is of great significance for ensuring railway transportation safety.

[0003] Traditional train monitoring technologies mainly rely on track circuits, video surveillance, or amplitude-based fiber optic sensing systems, but these methods have significant limitations:

[0004] Limitations of single-train monitoring: Existing amplitude-based distributed fiber optic sensing systems (such as traditional OTDRs) can only locate a single train by vibration intensity and cannot distinguish between vibration signals of two trains traveling in the same or opposite directions, leading to data confusion in multi-train scenarios.

[0005] Insufficient direction and speed recognition: Existing technologies lack the ability to accurately determine the direction of train travel, and speed calculation relies on data from a single sensor, which is susceptible to noise interference, resulting in insufficient accuracy.

[0006] Poor environmental adaptability: Harsh environments such as electromagnetic interference and extreme temperatures may affect the stability of traditional sensors, while existing phase demodulation algorithms are highly complex and difficult to process massive amounts of data in real time.

[0007] Φ-OTDR, as an advanced distributed fiber optic sensor, utilizes the phase change of Rayleigh scattered light in optical fibers to detect external disturbances, offering advantages such as high sensitivity, high spatial resolution, and a wide measurement range. However, current Φ-OTDR applications still focus on single-target monitoring, failing to address the issues of separating vibration signals from dual trains, determining their direction, and calculating their velocity.

[0008] Therefore, there is an urgent need for an innovative method based on Φ-OTDR to achieve accurate monitoring of the operating status of two trains, including real-time calculation of position, direction and speed in both same-direction and opposite-direction driving scenarios, while improving anti-interference capability and system robustness. Summary of the Invention

[0009] In view of this, the present invention provides a dual-train status monitoring system and method based on Φ-OTDR, which realizes accurate monitoring of the operating status of dual trains, including real-time calculation of position, direction and speed in the same and opposite directions of travel, while improving anti-interference ability and system robustness.

[0010] To achieve the above objectives, the present invention adopts the following technical solution:

[0011] In a first aspect, embodiments of the present invention provide a dual-train status monitoring system based on Φ-OTDR, comprising: a Φ-OTDR distributed fiber optic sensing optical path, an AD acquisition module, an FPGA circuit, and a host computer.

[0012] The Φ-OTDR distributed optical fiber sensing optical path is buried next to the railway track and is used to generate scattered light signals when a train passes by and convert them into electrical signals.

[0013] The AD acquisition module digitizes the electrical signal;

[0014] The FPGA circuit includes:

[0015] The signal demodulation module is used to perform digital quadrature demodulation on digital signals and extract vibration phase data along the railway track.

[0016] The Rayleigh phase processing module uses the Hilbert transform algorithm to calculate the phase difference and establishes a mapping relationship between the phase difference and the vibration amplitude and frequency.

[0017] The data interaction module transmits the data processed by the Rayleigh phase processing module to the host.

[0018] The host computer generates a time-position two-dimensional waterfall plot based on the received data, determines the direction of travel of the two trains by the same or opposite direction of the vibration point trajectory, calculates the displacement change and time difference based on the time series analysis algorithm, and optimizes the speed data by combining moving average filtering and Kalman filtering to output the real-time position and speed of the two trains.

[0019] Furthermore, the Φ-OTDR distributed fiber optic sensing optical path comprises: an ultra-narrow linewidth laser, an acousto-optic modulator (AOM), an erbium-doped fiber amplifier (EDFA), a beam splitter, a circulator, a 3dB coupler, and a balanced detector (BPD).

[0020] The ultra-narrow linewidth laser is connected to the acousto-optic modulator AOM via a coupler;

[0021] The output of the acousto-optic modulator AOM is connected to the erbium-doped fiber amplifier EDFA.

[0022] The output of the erbium-doped fiber amplifier (EDFA) is connected to the first port of the circulator.

[0023] The second port of the circulator is connected to the optical cable laid beside the railway track, and the third port is connected to the balance detector BPD through a 3dB coupler.

[0024] Furthermore, the signal demodulation module includes:

[0025] The digital quadrature demodulation unit is used to mix the digitized signal with the same frequency quadrature carrier to generate I and Q signals respectively;

[0026] A low-pass filter unit is used to filter out high-frequency components from the I and Q signals by passing them through a low-pass filter.

[0027] The CORDIC demodulation unit is used to obtain phase information from the I and Q signals (after filtering out high-frequency components) through the arctangent function of the CORDIC IP core.

[0028] Furthermore, the Rayleigh phase processing module includes:

[0029] For the input demodulated vibration data sequence along the railway track x(n), a Hilbert transform unit is constructed to generate a Hilbert transform h(n) whose unit impulse response satisfies the following condition: h(0) = 0; the analytic signal z(n) = x(n) + jy(n) is obtained through convolution operation y(n) = x(n) * h(n), and the phase of the signal is extracted.

[0030] Phase difference calculation unit, based on Define the continuous phase after untangling

[0031] Calculate phase difference

[0032] Data combination unit, used to determine phase difference It is linearly related to the vibration amplitude A: k1 is a proportionality coefficient determined through experimental calibration; and based on the phase difference... The rate of change over time, i.e., frequency. Δt is the time interval between two adjacent samples; the phase difference is combined with the vibration amplitude and frequency to form the final data.

[0033] Furthermore, the host includes:

[0034] The data receiving unit is used to receive processed data from the data interaction module of the FPGA circuit, including vibration position, phase difference, amplitude, and frequency information.

[0035] The graphics rendering unit displays the received data in the form of a waterfall chart, with the horizontal axis representing the position along the optical fiber, the vertical axis representing time, and color or grayscale representing the vibration intensity.

[0036] The direction determination unit is used to extract the set of vibration points of the two trains from the waterfall diagram by analyzing the trajectory characteristics of the vibration points in the waterfall diagram, calculate the displacement direction of adjacent time segments, and determine the direction of travel of the trains.

[0037] The speed calculation unit, based on time series analysis, calculates the displacement change and time difference of the train, and applies the speed formula to calculate the preliminary speed;

[0038] The filtering and optimization unit uses moving average filtering and Kalman filtering to smooth and optimize the initial velocity data and eliminate noise interference.

[0039] The result output module is used to output the final status information of the two trains, including position, speed, and direction.

[0040] Secondly, embodiments of the present invention also provide a dual-train condition monitoring method based on Φ-OTDR, applying the dual-train condition monitoring system based on Φ-OTDR as described in any one of the first aspects, the method comprising:

[0041] S1. Rayleigh scattering light signals are generated through a distributed optical fiber sensing optical path, and after being digitized by photoelectric conversion and AD acquisition modules, digital quadrature demodulation is performed by FPGA circuit to extract vibration phase data along the railway track;

[0042] S2. The phase data is processed using the Hilbert transform algorithm to construct an analytical signal and calculate the phase difference, and establish a mapping relationship between the phase difference and the vibration amplitude and frequency;

[0043] S3. The host receives the processed data and generates a time-position two-dimensional waterfall plot. The direction of travel of the two trains is determined by analyzing the same or opposite direction of the vibration point trajectory.

[0044] S4. The host calculates the displacement changes and time difference of the two trains based on the time series analysis algorithm, and optimizes the speed data by combining moving average filtering and Kalman filtering, and outputs the real-time position and speed of the two trains.

[0045] Further, in step S1, the FPGA circuit performs digital quadrature demodulation to extract vibration phase data along the railway track, including:

[0046] The digital signal is mixed with an orthogonal carrier of the same frequency to generate two signals, I and Q; the I and Q signals are filtered out by a low-pass filter to remove high-frequency components.

[0047] The phase information of the I and Q signals, after filtering out high-frequency components, is obtained through the arctangent function of the cordic ip core.

[0048] Further, step S2 includes:

[0049] For the input demodulated vibration data sequence along the railway track x(n), a Hilbert transformer h(n) is constructed, whose unit impulse response satisfies h(0) = 0; the analytic signal z(n) = x(n) + jy(n) is obtained through convolution operation y(n) = x(n) * h(n), and the phase of the signal is extracted.

[0050] based on Define the continuous phase after untangling Calculate phase difference

[0051] Determine the phase difference It is linearly related to the vibration amplitude A: k1 is a proportionality coefficient determined through experimental calibration; and based on the phase difference... The rate of change over time, i.e., frequency. Δt is the time interval between two consecutive samples;

[0052] The phase difference is combined with the vibration amplitude and frequency to form the final data.

[0053] Further, step S3 includes:

[0054] The data processing module of the FPGA circuit receives processed data, including vibration position, phase difference, amplitude, and frequency information.

[0055] The received data is displayed in the form of a waterfall chart, with the horizontal axis representing the position along the optical fiber, the vertical axis representing time, and color or grayscale representing the vibration intensity.

[0056] By analyzing the trajectory characteristics of vibration points in the waterfall diagram, the set of vibration points for the two trains is extracted from the waterfall diagram.

[0057] When two trains are moving in the same direction and at similar speeds, they are traveling in the same direction; when they are moving in opposite directions, they are traveling in opposite directions.

[0058] Further, step S4 includes:

[0059] Based on time series analysis, the displacement changes and time differences of the two trains are calculated, and the initial speeds are calculated using the speed formulas.

[0060] The initial velocity data was smoothed and optimized using moving average filtering and Kalman filtering to eliminate noise interference;

[0061] Output the final status information of the two trains, including position, speed, and direction.

[0062] As can be seen from the above technical solution, compared with the prior art, the present invention has the following technical advantages:

[0063] 1. Breakthrough in multi-train monitoring capabilities

[0064] This invention utilizes the high-sensitivity phase detection of Φ-OTDR, combined with Hilbert transform to extract phase difference features, to separate the vibration signals of two trains. By leveraging the trajectory characteristics of vibration points in a waterfall plot (parallel movement in the same direction or intersecting in opposite directions), the traveling directions of the two trains can be accurately distinguished. It is applicable to scenarios where two trains run parallel or intersect, such as high-speed rail double tracks and busy railway sections.

[0065] 2. High-precision direction and velocity calculation

[0066] This invention utilizes a trajectory analysis algorithm based on a time-position waterfall plot, combined with a dynamic threshold mechanism (such as sliding window counting), to automatically determine the train's direction of travel. It extracts the displacement difference (ΔS) and time difference (Δt) through a time series analysis algorithm, and optimizes the data using moving average filtering and Kalman filtering to eliminate instantaneous noise. This allows for accurate determination of direction and speed.

[0067] 3. Improved real-time performance and system robustness

[0068] This invention utilizes FPGA parallel processing, which implements hardware acceleration in FPGA circuits based on core algorithms such as signal demodulation and phase extraction, resulting in low processing latency. In addition, it has a dynamic adaptive mechanism, where the moving average filter window length and Kalman gain parameters can be dynamically adjusted according to environmental noise, improving system adaptability, real-time performance, and system robustness.

[0069] This invention solves the core problems of insufficient multi-train monitoring capability, reliance on manual direction judgment, and susceptibility to interference in speed calculation in traditional technologies by using phase demodulation algorithms, multi-level filtering optimization, FPGA hardware acceleration, and visual interactive design. It significantly improves the accuracy, real-time performance, and reliability of railway safety monitoring, and has clear technological advancement and practical value. Attached Figure Description

[0070] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on the provided drawings without creative effort.

[0071] Figure 1 The structural diagram of the dual train condition monitoring system based on Φ-OTDR provided by the present invention.

[0072] Figure 2 The flowchart of the speed detection algorithm provided by this invention.

[0073] Figure 3 A flowchart of the train running direction algorithm provided by the present invention.

[0074] Figure 4The flowchart of the dual-train status monitoring method based on Φ-OTDR provided by the present invention is shown.

[0075] Figure 5 Optical path diagram of the distributed optical fiber sensing optical path system provided by the present invention. Detailed Implementation

[0076] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0077] Reference Figure 1 As shown in the figure, this invention discloses a dual-train status monitoring system based on Φ-OTDR, which consists of a Φ-OTDR distributed fiber optic sensing optical path, an AD acquisition module, an FPGA circuit, and a host computer. The FPGA circuit includes a signal demodulation module, a Rayleigh phase processing module, and a data interaction module.

[0078] The Φ-OTDR distributed fiber optic sensing optical path is buried beside the railway track to generate scattered light signals when a train passes by and convert them into electrical signals; the AD acquisition module digitizes the electrical signals.

[0079] The signal demodulation module performs digital quadrature demodulation on the digitized signal to extract vibration phase data along the railway track. The Rayleigh phase processing module uses the Hilbert transform algorithm to calculate the phase difference and establish a mapping relationship between the phase difference and vibration amplitude and frequency. The data interaction module transmits the data processed by the Rayleigh phase processing module to the host computer. Based on the received data, the host computer generates a time-position two-dimensional waterfall plot, determines the direction of travel of the two trains by the same or opposite direction of the vibration point trajectory, and calculates the displacement change and time difference based on the time series analysis algorithm. It also optimizes the speed data by combining moving average filtering and Kalman filtering to output the real-time position and speed of the two trains.

[0080] The optical path of the Φ-OTDR distributed fiber optic sensing optical path system consists of an ultra-narrow linewidth laser, an acousto-optic modulator (AOM), an erbium-doped amplifier (EDFA), a beam splitter, a circulator, a 3dB coupler, and a balanced detector (BPD). The ultra-narrow linewidth laser is connected to the acousto-optic modulator (AOM) via a coupler. The acousto-optic modulator is connected to the erbium-doped amplifier (EDFA). The EDFA is connected to port 1 of the circulator. Port 2 of the circulator is connected to the laid optical cable. Port 3 of the circulator is connected to the 3dB coupler. The coupler is connected to the 3dB coupler. The 3dB coupler is connected to the balanced detector (BPD).

[0081] Working Principle: The components of the Φ-OTDR distributed fiber optic sensing optical path operate according to a specific connection method, generating scattered light signals and converting them into electrical signals. The acquisition module digitizes the electrical signals, separates them through signal processing, and inputs them to the signal demodulation module. Then, a Hilbert transform algorithm is used to construct a Hilbert transform to extract phase features from the demodulated vibration data sequence along the railway line, establishing the conversion relationship between phase difference and vibration amplitude and frequency, and combining them into the final data. The host receives the final data acquired and preliminarily processed by the sensing optical path, quantifies the vibration intensity, and uses a graphic rendering algorithm to display the data in the form of a waterfall plot. The correspondence between color or grayscale and vibration intensity visually presents the differences in vibration intensity at different locations and times. Based on the trajectory of the vibration points in the time-fiber position two-dimensional coordinate system in the waterfall plot, the direction of train travel is determined. When traveling in the same direction, the displacement change direction of the vibration point set is the same and the rate is similar; when traveling in opposite directions, the vibration points move in opposite directions. The host uses a speed detection algorithm based on time series analysis to record the vibration time and position information of the vibration points of two trains, using the distance change as displacement and the time difference as time, and calculates the travel speed according to the speed calculation formula. Simultaneously, a moving average filtering algorithm is used to process displacement and time data to ensure the accuracy of speed calculation. A Kalman filter algorithm is employed to predict and correct the speed data. By continuously repeating the prediction and update process, the final output train speed data accurately reflects the actual train speed, thereby enabling the monitoring of the operating status, position, and speed of two trains.

[0082] In one embodiment, the signal demodulation module includes:

[0083] The digital quadrature demodulation unit is used to mix the digitized signal with the same frequency quadrature carrier to generate I and Q signals respectively;

[0084] A low-pass filter unit is used to filter out high-frequency components from the I and Q signals by passing them through a low-pass filter.

[0085] The CORDIC demodulation unit is used to obtain phase information from the I and Q signals (after filtering out high-frequency components) through the arctangent function of the CORDIC IP core.

[0086] In this embodiment, digital quadrature demodulation mainly includes three parts: mixing, filtering, and Cordic demodulation. The signal demodulation module performs digital domain demodulation on the digitized signal; the digitized signal is multiplied by orthogonal signals of the same frequency, and the multiplication results in two mixed signals. After low-pass filtering, the resulting quadrature I and Q signals are obtained. The amplitude information of the I and Q signals is obtained by taking the square root of the Cordic IP core; the phase information of the two signals is obtained by the arctangent function of the Cordic IP core.

[0087] In one embodiment, the Rayleigh phase processing module uses the Hilbert transform algorithm for phase feature extraction. For the input demodulated rail vibration data sequence x(n), a Hilbert transform h(n) is constructed, whose unit impulse response satisfies... h(0) = 0. The analytic signal z(n) = x(n) + jy(n) is obtained through the convolution operation y(n) = x(n) * h(n), where j represents the imaginary unit. The phase of the signal...

[0088] Based on theoretical derivation and experimental calibration, the conversion relationship between phase difference and vibration amplitude and frequency is established; phase difference It has a linear relationship with the vibration amplitude A. Where k1 is the proportionality coefficient determined through experimental calibration; when the laser center wavelength is 1550.12 nm and the fiber core refractive index is 1.46, k1 = 110.37 nε·m / rad, and ε is the strain. For the vibration frequency f, it is calculated based on the rate of change of the phase difference with time, i.e. Where Δt is the time interval between two adjacent samples. The phase difference is combined with the vibration amplitude and frequency to form the final data: {position pos(n), time t(n), amplitude A(n), frequency f(n)}.

[0089] In one embodiment, the host includes:

[0090] The data receiving unit is used to receive processed data from the data interaction module of the FPGA circuit, including vibration position, phase difference, amplitude, and frequency information.

[0091] The graphics rendering unit displays the received data in the form of a waterfall chart, with the horizontal axis representing the position along the optical fiber, the vertical axis representing time, and color or grayscale representing the vibration intensity.

[0092] The direction determination unit is used to extract the set of vibration points of the two trains from the waterfall diagram by analyzing the trajectory characteristics of the vibration points, calculate the displacement direction of adjacent time segments, and determine the direction of travel of the trains. Taking the two trains traveling in the same direction as an example, their vibration point trajectories are represented as two parallel moving curves in the waterfall diagram (for example, both with a displacement rate of 150m / s). If they are traveling in opposite directions, the trajectories will cross in opposite directions (for example, one moves eastward at +120m / s, and the other moves westward at -130m / s).

[0093] The speed calculation unit, based on time series analysis, calculates the displacement change and time difference of the train, and applies the speed formula to calculate the preliminary speed;

[0094] The filtering and optimization unit uses moving average filtering and Kalman filtering to smooth and optimize the initial velocity data and eliminate noise interference.

[0095] The result output module is used to output the final status information of the two trains, including position, speed, and direction.

[0096] In this embodiment, the host receives the final data acquired and preliminarily processed from the sensing optical path. This data contains relevant information from different times and locations along the optical fiber. When acquiring and processing vibration data, the host quantifies the vibration intensity and sets a range of vibration intensity values, starting from a minimum value of I... min to the maximum value I max To cover all possible vibration intensity scenarios, the host computer uses OpenGL or CUDA-accelerated graphics rendering algorithms to generate waterfall plots. The collected data is then displayed in waterfall plot format. The waterfall plot constructs a two-dimensional coordinate system, with the vertical axis representing time and the horizontal axis representing the position along the fiber optic cable.

[0097] For the correspondence between color or grayscale and vibration intensity, a matching range is predefined. For color, a gradient color scheme is used, transitioning from colors representing weaker vibration intensity to those representing stronger vibration intensity; for grayscale, a transition is used, moving from grayscale corresponding to weaker vibration intensity to grayscale corresponding to stronger vibration intensity. Once the vibration intensity I at a certain location and time is acquired, its range within the vibration intensity range [I] is calculated. min ,I max The relative positions within [] are determined by the formula. A scale value between 0 and 1 is obtained. This scale value corresponds to a specific position within a pre-defined color or grayscale range, thereby displaying the differences in vibration intensity at different locations and times in the waterfall plot, providing an intuitive data display basis for subsequent analysis.

[0098] The vibration points captured and processed by Φ-OTDR, as reflected in the waterfall plot, exhibit highly similar trajectories in a two-dimensional time-fiber position coordinate system. Because the two trains travel in the same direction, their vibration excitation on the fiber is unidirectional in the time series. Specifically, in continuous time slices, the displacement changes of the vibration point sets corresponding to the two trains on the horizontal axis (position along the fiber) are in the same direction, and the displacement change rates are similar in magnitude within the same time interval. Conversely, when the two trains travel in opposite directions, their vibration excitation on the fiber exhibits opposite characteristics in the time series. Reflected in the waterfall plot, the movement directions of the vibration points corresponding to the two trains in the time-fiber position two-dimensional coordinate system are completely opposite. At the beginning of the waterfall plot, the vibration points of the two trains are located in different regions along the fiber. As time progresses, these two sets of vibration points gradually approach each other. During this approach, their displacement changes on the horizontal axis are in opposite directions, forming two opposing curves in the waterfall plot that gradually converge.

[0099] The real-time speed of the train can be calculated by using the ratio of the change in distance from the train's center position to time. For example... Figure 2The recorded current vibration position center pos_start_A of train A is shown, and the current system time time_start_A is recorded using the timestamp function. Traverse the next train phase curve. Position information can be obtained from the phase information. Confirm the amplitude center position pos_end_A of the next vibration position of train A, and record the current time time_end_A. Confirm the train running direction based on the relationship between the center positions of the train at two time points. If pos_end_A > pos_start_A, calculate the forward running speed of the train as (pos_end_A - pos_start_A) / (time_end_A - time_start_A); if pos_end_A < pos_start_A, calculate the reverse running speed of the train as (pos_start_A - pos_end_A) / (time_end_A - time_start_A).

[0100] Through Figure 3 The algorithm can determine the movement directions of two trains. Set the number of comparisons N. The number of times pos_end > pos_start is recorded as time_1 and time_2 respectively. On the premise of confirming that it is the vibration of train A, obtain the position information from the phase information, and record the amplitude center point of the current train A, denoted as pos_start_A = (pos + i) / 2. Traverse the next phase curve of train A, and confirm the amplitude center position of the next vibration position of train A from the phase information, denoted as pos_end_A = (pos + i) / 2. Then compare pos_end_A and pos_start_A: if pos_end_A > pos_start_A, let time_1 = time_1 + 1; if pos_end_A < pos_start_A, let time_1 = time_1 - 1. Continuously update pos_start_A, pos_end_A and compare according to the above steps, and continuously update time_1. When time_1 > N is satisfied, it means that train A continuously moves in the direction of increasing position within a period of time, and it can be judged that train A is moving forward; when time_1 < -N is satisfied, it means that train A continuously moves in the direction of decreasing position within a period of time, and it can be judged that train A is moving in reverse. Repeat the above steps to obtain the running direction of train B. When time_2 > N is satisfied, it means that train B is moving forward; when time_2 < -N is satisfied, it means that train B is moving in reverse. After judging the respective movement directions of train A and train B, if (time_1 > N and time_2 > N) or (time_1 < -N and time_2 < -N) is satisfied, it can be judged that the two trains are moving in the same direction; in other cases, it is judged that the two trains are moving in opposite directions.

[0101] Here, N is the threshold for the number of comparisons, which is related to data stability and system response speed requirements. Different types of trains have different vibration characteristics during operation. For example, high-speed trains run relatively smoothly, and their vibration data is highly regular, so the value of N can be relatively small; freight trains may experience more complex vibrations due to factors such as load and road conditions, so the value of N may need to be set larger. Assuming that a value of 10-12 comparisons is sufficient for high-speed trains to determine direction, freight trains may require 15-20 comparisons.

[0102] After determining the train's direction of travel, the main unit uses a speed detection algorithm based on time series analysis to calculate the speed. This algorithm records the vibration time and position information of the two train's vibration points in real time. Using the change in distance between the two train's vibration points at different times as displacement S, and the time difference between the first detection and the next moment as time t, the speeds of the two trains are calculated according to the speed calculation formula v = S / t.

[0103] The host computer processes the collected displacement and time data using a moving average filtering algorithm, performing multiple verifications and corrections. The moving average filtering algorithm processes the displacement and time data within a fixed-length time window. This time window contains n consecutive data points, for a displacement data sequence {S1, S2, ..., S...}. n Displacement data after moving average filtering The average of these n data points is... As time progresses, the time window slides forward, new data points enter the window, and old data points move out. This process is repeated to calculate a new average value, thus smoothing the shifted data and eliminating instantaneous fluctuations caused by noise and interference. For the time data sequence {t1, t2, ..., t...} n Similarly, using a similar method, the average value of n time points within the time window is calculated. Smoothed time data.

[0104] The Kalman filter algorithm predicts and corrects velocity data; it uses the velocity estimate from the previous moment. And a predictive model of velocity change, predicting the velocity at the current moment. When the new velocity measurement value v k After obtaining the results from the preceding calculations, the Kalman filter algorithm will calculate the Kalman gain K based on the difference between the measured and predicted values, as well as the statistical characteristics of the measurement noise and system noise. k Using Kalman gain, the predicted value... and measured value v k By performing weighted fusion, a more accurate velocity estimate for the current moment can be obtained. Right now By continuously repeating the prediction and update process, the speed data is continuously optimized, and the final output train speed data accurately reflects the actual operating speed of the train.

[0105] For example: The host records the initial position pos of the vibration point. start and time t start After 10 consecutive samplings, the displacement difference ΔS and time difference Δt are calculated, and the initial velocity v = ΔS / Δt. A moving average filter (window length 8) is used to smooth the data, and then a Kalman filter (process noise variance 0.1, observation noise variance 0.01) is used to correct the velocity values. Finally, the real-time speeds of the two trains are output (e.g., Train 1: 153.2 ± 0.5 km / h, Train 2: 148.7 ± 0.5 km / h).

[0106] Based on the same inventive concept, this invention also provides a dual-train condition monitoring method based on Φ-OTDR, applying the dual-train condition monitoring system based on Φ-OTDR as described in the above embodiments. Since the principle by which this method solves the problem is similar to that of the aforementioned dual-train condition monitoring system based on Φ-OTDR, the implementation of this method can refer to the implementation of the aforementioned system; repeated details will not be elaborated further. (Refer to...) Figure 4 As shown, the method includes:

[0107] S1. Rayleigh scattering light signals are generated through a distributed optical fiber sensing optical path, and after being digitized by photoelectric conversion and AD acquisition modules, digital quadrature demodulation is performed by FPGA circuit to extract vibration phase data along the railway track;

[0108] S2. The phase data is processed using the Hilbert transform algorithm to construct an analytical signal and calculate the phase difference, and establish a mapping relationship between the phase difference and the vibration amplitude and frequency;

[0109] S3. The host receives the processed data and generates a time-position two-dimensional waterfall plot. The direction of travel of the two trains is determined by analyzing the same or opposite direction of the vibration point trajectory.

[0110] S4. The host calculates the displacement changes and time difference of the two trains based on the time series analysis algorithm, and optimizes the speed data by combining moving average filtering and Kalman filtering, and outputs the real-time position and speed of the two trains.

[0111] This method relies on a monitoring system comprised of a Φ-OTDR distributed fiber optic sensing optical path, an AD acquisition module, an FPGA circuit, and a host computer. The FPGA circuit integrates signal demodulation, Rayleigh phase processing, and data interaction modules. Scattered light signals are generated through the Φ-OTDR distributed fiber optic sensing optical path, undergoing photoelectric conversion, digitization, and a series of signal processing and demodulation steps to acquire phase data along the railway track. The Rayleigh phase processing module extracts phase features based on the phase characteristics of Rayleigh scattered light using the Hilbert transform algorithm. After receiving the final data, the host computer visually presents the vibration status using a waterfall plot. The direction of travel is determined based on the different trajectory characteristics of vibration points in the waterfall plot when two trains are traveling in the same direction and opposite directions, and the train speed is calculated using a time series analysis algorithm. Through the above algorithms and system architecture, accurate monitoring of the operating status of two trains is achieved.

[0112] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on its differences from other embodiments. Similar or identical parts between embodiments can be referred to interchangeably. For the apparatus disclosed in the embodiments, since they correspond to the methods disclosed in the embodiments, the description is relatively simple; relevant parts can be referred to the method section.

[0113] The above description of the disclosed embodiments enables those skilled in the art to make or use the invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the invention. Therefore, the invention is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A dual-train condition monitoring system based on Φ-OTDR, characterized in that, include: Φ-OTDR distributed fiber optic sensing optical path, AD acquisition module, FPGA circuit and host: The Φ-OTDR distributed optical fiber sensing optical path is buried next to the railway track and is used to generate scattered light signals when a train passes by and convert them into electrical signals. The AD acquisition module digitizes the electrical signal; The FPGA circuit includes: The signal demodulation module is used to perform digital quadrature demodulation on digital signals and extract vibration phase data along the railway track. The Rayleigh phase processing module uses the Hilbert transform algorithm to calculate the phase difference and establishes a mapping relationship between the phase difference and the vibration amplitude and frequency. The data interaction module transmits the data processed by the Rayleigh phase processing module to the host. The host computer generates a time-position two-dimensional waterfall plot based on the received data, determines the direction of travel of the two trains by the same or opposite direction of the vibration point trajectory, calculates the displacement change and time difference based on the time series analysis algorithm, and optimizes the speed data by combining moving average filtering and Kalman filtering to output the real-time position and speed of the two trains.

2. The dual-train condition monitoring system based on Φ-OTDR according to claim 1, characterized in that, The Φ-OTDR distributed fiber optic sensing optical path comprises: an ultra-narrow linewidth laser, an acousto-optic modulator (AOM), an erbium-doped fiber amplifier (EDFA), a beam splitter, a circulator, a 3dB coupler, and a balanced detector (BPD). The ultra-narrow linewidth laser is connected to the acousto-optic modulator AOM via a coupler; The output of the acousto-optic modulator AOM is connected to the erbium-doped fiber amplifier EDFA. The output of the erbium-doped fiber amplifier (EDFA) is connected to the first port of the circulator. The second port of the circulator is connected to the optical cable laid beside the railway track, and the third port is connected to the balance detector BPD through a 3dB coupler.

3. The dual-train condition monitoring system based on Φ-OTDR according to claim 1, characterized in that, The signal demodulation module includes: The digital quadrature demodulation unit is used to mix the digitized signal with the same frequency quadrature carrier to generate I and Q signals respectively; A low-pass filter unit is used to filter out high-frequency components from the I and Q signals by passing them through a low-pass filter. The CORDIC demodulation unit is used to obtain phase information from the I and Q signals (after filtering out high-frequency components) through the arctangent function of the CORDIC IP core.

4. A dual-train condition monitoring system based on Φ-OTDR according to claim 1, characterized in that, The Rayleigh phase processing module includes: For the input demodulated vibration data sequence along the railway track x(n), a Hilbert transform unit is constructed to generate a Hilbert transform h(n) whose unit impulse response satisfies the following condition: The analytic signal z(n) = x(n) + jy(n) is obtained by convolution operation y(n) = x(n) * h(n), and the phase of the signal is extracted. Phase difference calculation unit, based on Define the continuous phase after untangling Calculate phase difference Data combination unit, used to determine phase difference It is linearly related to the vibration amplitude A: k1 is a proportionality coefficient determined through experimental calibration; and based on the phase difference... The rate of change over time, i.e., frequency. Δt is the time interval between two adjacent samples; the phase difference is combined with the vibration amplitude and frequency to form the final data.

5. A dual-train condition monitoring system based on Φ-OTDR according to claim 1, characterized in that, The host includes: The data receiving unit is used to receive processed data from the data interaction module of the FPGA circuit, including vibration position, phase difference, amplitude, and frequency information. The graphics rendering unit displays the received data in the form of a waterfall chart, with the horizontal axis representing the position along the optical fiber, the vertical axis representing time, and color or grayscale representing the vibration intensity. The direction determination unit is used to extract the set of vibration points of the two trains from the waterfall diagram by analyzing the trajectory characteristics of the vibration points in the waterfall diagram, calculate the displacement direction of adjacent time segments, and determine the direction of travel of the trains. The speed calculation unit, based on time series analysis, calculates the displacement change and time difference of the train, and applies the speed formula to calculate the preliminary speed; The filtering and optimization unit uses moving average filtering and Kalman filtering to smooth and optimize the initial velocity data and eliminate noise interference. The result output module is used to output the final status information of the two trains, including position, speed, and direction.

6. A dual-train condition monitoring method based on Φ-OTDR, characterized in that, The method of using the dual-train condition monitoring system based on Φ-OTDR as described in any one of claims 1-5 includes: S1. Rayleigh scattering light signals are generated through a distributed optical fiber sensing optical path, and after being digitized by photoelectric conversion and AD acquisition modules, digital quadrature demodulation is performed by FPGA circuit to extract vibration phase data along the railway track; S2. The phase data is processed using the Hilbert transform algorithm to construct an analytical signal and calculate the phase difference, and establish a mapping relationship between the phase difference and the vibration amplitude and frequency; S3. The host receives the processed data and generates a time-position two-dimensional waterfall plot. The direction of travel of the two trains is determined by analyzing the same or opposite direction of the vibration point trajectory. S4. The host calculates the displacement changes and time difference of the two trains based on the time series analysis algorithm, and optimizes the speed data by combining moving average filtering and Kalman filtering, and outputs the real-time position and speed of the two trains.

7. A dual-train condition monitoring method based on Φ-OTDR according to claim 6, characterized in that, In step S1, the FPGA circuit performs digital quadrature demodulation to extract vibration phase data along the railway track, including: The digital signal is mixed with an orthogonal carrier of the same frequency to generate two signals, I and Q; the I and Q signals are filtered out by a low-pass filter to remove high-frequency components. The phase information of the I and Q signals, after filtering out high-frequency components, is obtained through the arctangent function of the cordic ip core.

8. A dual-train condition monitoring method based on Φ-OTDR according to claim 6, characterized in that, Step S2 includes: For the input demodulated vibration data sequence along the railway track x(n), a Hilbert transformer h(n) is constructed, whose unit impulse response satisfies The analytic signal z(n) = x(n) + jy(n) is obtained by convolution operation y(n) = x(n) * h(n), and the phase of the signal is extracted. based on Define the continuous phase after untangling Calculate phase difference Determine the phase difference It has a linear relationship with the vibration amplitude A: k1 is a proportionality coefficient determined through experimental calibration; and based on the phase difference... The rate of change over time, i.e., frequency. Δt is the time interval between two adjacent samples; The phase difference is combined with the vibration amplitude and frequency to form the final data.

9. A dual-train condition monitoring method based on Φ-OTDR according to claim 6, characterized in that, Step S3 includes: The data processing module of the FPGA circuit receives processed data, including vibration position, phase difference, amplitude, and frequency information. The received data is displayed in the form of a waterfall chart, with the horizontal axis representing the position along the optical fiber, the vertical axis representing time, and color or grayscale representing the vibration intensity. By analyzing the trajectory characteristics of vibration points in the waterfall diagram, the set of vibration points for the two trains is extracted from the waterfall diagram. When two trains are moving in the same direction and at similar speeds, they are traveling in the same direction; when they are moving in opposite directions, they are traveling in opposite directions.

10. A dual-train condition monitoring method based on Φ-OTDR according to claim 6, characterized in that, Step S4 includes: Based on time series analysis, the displacement changes and time differences of the two trains are calculated, and the initial speeds are calculated using the speed formulas. The initial velocity data was smoothed and optimized using moving average filtering and Kalman filtering to eliminate noise interference; Output the final status information of the two trains, including position, speed, and direction.

Images