Traffic road optical cable measurement area consistency detection system and method

Abstract

The present invention discloses a traffic road optical cable measurement area consistency detection system, which includes a driving vibration signal acquisition module, a signal length alignment module, a signal similarity calculation module, a histogram drawing module and a consistency comparison module; the present invention is based on a time domain alignment method of driving vibration signals between measurement areas based on dynamic time regularization, which makes it possible to calculate the correlation coefficient between signals. After calculating the correlation coefficient and distance between the driving vibration signal of each measurement area and the driving vibration signal of the reference measurement area, a correlation coefficient histogram is derived from the correlation coefficient, which solves the problem that the consistency between the measurement area data is not intuitive enough when the correlation coefficient is directly compared. At the same time, the correlation coefficient histogram can also reflect the distribution of the correlation coefficient between the measurement areas to a certain extent. Finally, the probability density function is fitted to the correlation coefficient histogram of the different segmented optical cables respectively, which realizes the parameterization of the probability density function of the histogram and avoids the influence of the randomness of the data on the histogram.

Description

Technical Field

[0001] The present invention relates to the field of optical cable technology, and in particular to a traffic road optical cable measurement area consistency detection system and method. Background Art

[0002] The signal used for consistency detection is the driving vibration signal generated by the same vehicle (car, subway, train, etc.) when passing through different optical cable measurement areas. The optical cable measurement area refers to the distributed vibration sensors composed of ultra-weak fiber grating arrays (UWFBGs) based on optical fiber sensing technology laid on traffic roads (including rail transportation and road traffic).

[0003] To analyze the consistency of optical cable measurement data on traffic roads, it is necessary to calculate parameters such as the correlation coefficient and distance of the data from the two measurement areas. To calculate these parameters, it is necessary to align the lengths of the vehicle driving vibration signals. Currently, there is no mature solution for this.

[0004] In addition, after obtaining the correlation coefficient or distance between the measurement area signals, although the two can reflect the consistency between the signals to a certain extent, directly observing the correlation coefficient or distance is not intuitive enough, especially when it is necessary to detect the consistency distribution of the measurement areas in different sections of the optical cable and compare the consistency distribution of the measurement areas between different sections of the optical cable, only using the correlation coefficient or analyzing the distance between the signals cannot obtain useful information. Summary of the Invention

[0005] The purpose of the present invention is to provide a traffic road optical cable measurement area consistency detection system and method. The present invention aligns the driving signals of all measurement areas with the driving vibration signals of the reference measurement area in time, then calculates the correlation coefficient between the driving vibration signals of all measurement areas and the driving vibration signals of the reference measurement area, derives the correlation coefficient histogram from the correlation coefficient, and then fits the corresponding probability density function from the correlation coefficient histogram. This can detect the consistency distribution of the measurement areas within the segment and compare the consistency of the measurement areas between segments.

[0006] To achieve this purpose, the present invention provides a traffic road optical cable measurement area consistency detection system, which includes a vehicle vibration signal acquisition module, a signal length alignment module, a signal similarity calculation module, a histogram drawing module, and a consistency comparison module; each vehicle vibration signal measurement area in the traffic road distributed acoustic wave sensing system can sense the vehicle vibration signal sequence in the corresponding measurement area;

[0007] The driving vibration signal acquisition module is used to obtain driving vibration signal sequences of each measurement area along the traffic road to be measured, calculate the spectral entropy of the driving vibration signal sequence of each measurement area, and randomly select a driving vibration signal sequence of the measurement area from the driving vibration signal sequences of the measurement area whose spectral entropy is within a preset spectral entropy interval as a reference driving vibration signal sequence, and the remaining driving vibration signal sequences of the measurement area are used as test driving vibration signal sequences of the measurement area;

[0008] The signal length alignment module is used to align the length of the test driving vibration signal sequence in each measurement area with the length of the reference driving vibration signal sequence;

[0009] The signal similarity calculation module is used to calculate the correlation coefficient between the test driving vibration signal sequence of each measurement area and the reference driving vibration signal sequence after length alignment;

[0010] The histogram drawing module is used to draw the correlation coefficients between the test driving vibration signal sequences and the reference driving vibration signal sequences in all measurement areas and the frequencies of the corresponding correlation coefficients into a correlation coefficient histogram;

[0011] The consistency comparison module is used to obtain a corresponding probability density function fitting graph according to each correlation coefficient histogram, and to obtain a corresponding probability density function comparison graph according to each probability density function fitting graph.

[0012] Beneficial effects of the present invention:

[0013] The present invention is based on a time domain alignment method for driving vibration signals between measurement areas based on dynamic time regularization, which makes it possible to calculate the correlation coefficient between signals, laying the foundation for further investigating the consistency between driving vibration signals in the measurement areas. After calculating the correlation coefficient and distance between the driving vibration signals in each measurement area and the driving vibration signals in the reference measurement area, a correlation coefficient histogram can be derived from the correlation coefficient. This solves the problem that the consistency between the measurement area data is not intuitive enough when the correlation coefficient is directly compared. At the same time, the correlation coefficient histogram can also reflect the distribution of the correlation coefficient between the measurement areas to a certain extent. Finally, the probability density function is fitted to the correlation coefficient histogram of different segmented optical cables, which realizes the parameterization of the probability density function of the histogram and avoids the influence of the randomness of the data on the histogram. By comparing the probability density functions fitted by different segmented optical cables, it is possible to determine the distribution of the consistency of the measurement areas within the segment and to compare the consistency of the measurement areas between the segments.

[0014] The present invention solves the problem of unclear analysis process of data consistency between measurement areas. At the same time, it also solves the problem that when analyzing the consistency of measurement area data, only using correlation coefficient or distance between signals, no useful information can be obtained and the analysis is not intuitive enough. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1It is a structural schematic diagram of the present invention;

[0016] Figure 2 This is a schematic diagram of the laying of three types of optical cables in subway tunnels;

[0017] Figure 3 This is a comparison diagram of the signal in the benchmark measurement area and the signal in a typical measurement area before and after dynamic time warping;

[0018] Figure 4 This is a comparison chart of the baseline measurement area signal and its own dynamic time regularization before and after;

[0019] Figure 5 This is a comparison diagram of the signal in the benchmark measurement area and the signal in a typical measurement area before and after dynamic time warping;

[0020] Figure 6 The distance distribution diagram between each measurement area of ​​the optical cable and the reference measurement area signal;

[0021] Figure 7 It is the correlation coefficient distribution diagram between the signals of each measurement area of ​​the optical cable and the reference measurement area;

[0022] Figure 8 is the correlation coefficient histogram of all survey areas;

[0023] Figure 9 Fit probability density functions to the correlation coefficient histograms of the three optical cables;

[0024] Figure 10 Comparison diagram of the probability density function fitted for the correlation coefficient distribution of the three cable measurement areas;

[0025] Figure 11 This is the schematic diagram of the distributed acoustic wave sensing system. DETAILED DESCRIPTION

[0026] The present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments:

[0027] This embodiment is explained using rail transportation in a traffic road. In a certain traffic line, in order to sense the vibration signal of the train in the subway tunnel, it is necessary to lay optical cables. In order to analyze the specific performance of different types of optical cables, three types of optical cables are tested as follows during construction. Figure 1 As shown in the figure, they are connected and laid in parallel in the subway tunnel. In this way, for the trains running between Station A and Station B, each measuring area of ​​the three cables can be used to sense the train vibration signal, such as Figure 2 As shown;

[0028] The sensing system used in this traffic road is a distributed acoustic sensing system (DAS), such as Figure 11As shown in the figure, in this system, continuous light emitted by a narrow-linewidth laser is modulated into a train of optical pulses by an electro-optical modulator. This pulse train is then amplified by an erbium-doped fiber amplifier and ultimately incident on an ultra-weak fiber Bragg grating (UWFBG) array. After reflection from the UWFBG array, the amplified pulsed light enters an unbalanced Michelson interferometer. The interferometer recovers the amplitude of the time-domain vehicle vibration signal by demodulating the phase shift caused by the optical length variation in the optical fiber between two adjacent UWFBGs. The resulting interference signal is then output to three photodetectors, resulting in the generation of a time-domain vehicle vibration signal after photoelectric conversion. Because the length of the delay fiber in the Michelson interferometer matches the distance between adjacent UWFBGs in the optical cable, each pair of adjacent UWFBGs and the optical fiber between them constitutes a sensor, or a measurement zone. In rail transit, vibration signals occurring within this zone are sensed. Vibration signals from passing trains can be intercepted as needed, representing the vehicle vibration signal within the measurement zone.

[0029] Figure 11 The grating array into which the light of the DAS system is incident is Figure 2 The three optical cables shown are fused end-to-end and laid parallel to the track. When a train runs on the track, its vibration signals are simultaneously sensed by all three cables. For the DAS system, incident light enters Type 1 cable at Station A and reaches Station B. Next, incident light enters Type 2 cable at Station B and reaches Station A. Finally, incident light enters Type 3 cable at Station A and reaches Station B. This allows the system to capture vibration signals caused by the train as it passes through each measurement zone.

[0030] like Figure 1 The traffic road optical cable measurement area consistency detection system shown in the figure includes a driving vibration signal acquisition module, a signal length alignment module, a signal similarity calculation module, a histogram drawing module and a consistency comparison module;

[0031] Two adjacent ultra-weak fiber Bragg gratings and the optical fiber between them in the traffic road distributed acoustic wave sensing system form a vehicle vibration signal measurement area. Each vehicle vibration signal measurement area can sense the corresponding vehicle vibration signal sequence in the measurement area.

[0032] The vehicle vibration signal acquisition module is used to obtain the vehicle vibration signal sequences of each measurement area along the traffic road to be tested, and calculate the spectral entropy of the vehicle vibration signal sequences of each measurement area. Then, a vehicle vibration signal sequence of the measurement area with a spectral entropy within a preset spectral entropy interval is randomly selected as the reference vehicle vibration signal sequence, and the remaining vehicle vibration signal sequences are used as the test vehicle vibration signal sequences of the measurement area. The spectral entropy of the signal describes the relationship between the power spectrum and the entropy rate. The larger the spectral entropy, the greater the uncertainty and confusion of the signal.

[0033] The signal length alignment module is used to perform dynamic time warping operations on the test driving vibration signal sequences of each measurement area and the reference driving vibration signal sequence, so that the lengths of the test driving vibration signal sequences of each measurement area are aligned with the length of the reference driving vibration signal sequence. After dynamic time warping, the originally misaligned signals can be compared for consistency. The comparison diagram before and after dynamic time warping of the reference measurement area signal and the typical measurement area signal is shown in the figure below. Figure 3 Similarly, the dynamic time warping effect diagram of the reference measurement area signal and itself, and the reference measurement area signal and a typical measurement area signal is shown in Figure 4 、 5 As shown;

[0034] The signal similarity calculation module is used to perform similarity calculation on each test driving vibration signal sequence of each measurement area after length alignment and the reference driving vibration signal sequence, and calculate the Euclidean distance and correlation coefficient between each test driving vibration signal sequence of each measurement area after length alignment and the reference driving vibration signal sequence.

[0035] The histogram drawing module is used to draw the correlation coefficients between the test driving vibration signal sequences of all measurement areas and the reference driving vibration signal sequences and the frequencies of the corresponding correlation coefficients (the number of times a certain value of the correlation coefficient appears divided by the total number of measurement areas is the frequency of the correlation coefficient) into a correlation coefficient histogram, such as Figure 8 As shown in the figure, the correlation coefficient between the signal of each measurement area and the signal of the reference measurement area is the horizontal axis of the frequency distribution histogram, so the distribution of the correlation coefficient can be analyzed. Figure 8 Taking the case in as an example, the correlation coefficients are clustered and distributed, but there are also some long-tail distributions;

[0036] The consistency comparison module is used to fit each correlation coefficient histogram with a probability density function to obtain a corresponding probability density function fitting diagram, such as Figure 9 As shown in , the probability density function fitting graphs are combined to obtain the corresponding probability density function comparison graph, as shown in Figure 10 As shown, the consistency of the test vehicle vibration signal in each measurement area and the reference vehicle vibration signal sequence is compared according to the probability density function comparison diagram.

[0037] In the above technical solution, the spectral entropy H(f) of the vehicle vibration signal sequence in the measurement area is calculated as follows:

[0038]

[0039] Among them, f is the frequency of the driving vibration signal sequence in the measurement area, f s is the sampling rate of the vehicle vibration signal sequence in the measurement area, and P(f) is the normalized power spectral density of the vehicle vibration signal sequence in the measurement area.

[0040] In the above technical solution, the preset spectral entropy range is 0≤H(f)≤0.3.

[0041] In the above technical solution, the dynamic time warping operation can respectively lengthen or shorten (compress and expand) the test vehicle vibration signal sequence in the measurement area until the length of the test vehicle vibration signal sequence in each measurement area is consistent with the reference vehicle vibration signal sequence. In this process, the time axis of the test vehicle vibration signal sequence in each measurement area will be distorted or bent so that the sum of the Euclidean distances of all corresponding points in the two signals after alignment is minimized.

[0042] In the above technical solution, the specific method for performing the dynamic time warping operation is:

[0043] For the reference vehicle vibration signal sequence (q n0 ), and the test driving vibration signal sequence of the measuring area with a length of m0 (c m0 ), when m0 is not equal to n0, construct an n0×m0 matrix and calculate the element q at (i,j) of the n0×m0 matrix i and c j The Euclidean distance d(q i ,c j ), The reference driving vibration signal sequence (q n0 ) and the test driving vibration signal sequence in the test area (c m0 ) is aligned, that is, to find a monotonically continuous path starting from the (1,1) element in the matrix and ending at the (n0,m0) element, and the sum of the elements on the path is the smallest among all possible paths, so that the reference vehicle vibration signal sequence (q n0 ) and the test driving vibration signal sequence in the test area (c m0 ) should appear at the same position in time, where the monotonicity and continuity of the path mean that when an element (a′, b′) in the n0×m0 matrix is ​​selected as a point on the path, the next point (a, b) on the path should satisfy 0≤(aa′)≤1 and 0≤(bb′)≤1. The elements and the minimum path found by dynamic time warping (DTW) in the distance matrix indicate that the points of the corresponding two sequences should appear at the same position in time. After knowing the correspondence between the points in the sequence, the time axis of the test signal can be aligned with the time axis of the reference signal by compression and expansion. Specifically, after dynamic time warping finds a path, the number n of points in the matrix covered by the path will be greater than or equal to the length of the longer sequence to be aligned, that is, n≥max(n0,m0), and n is also the length of the two aligned signals. For the reference vehicle vibration signal sequence (q n0 ), starting from its first point, the path has several points in this row in the matrix, the original sequence (qn0 ) appears several times in the aligned sequence, and traverses the points in the original signal in turn, and finally obtains the aligned sequence (q n ). For the test vehicle vibration signal sequence (c m0 ), starting from its first point, the path has several points in this column in the matrix, the original sequence (c m0 ) appears several times in the aligned sequence, and traverses the points in the original signal in turn, and finally obtains the aligned sequence (c n ).

[0044] In the above technical solution, the specific method for calculating the Euclidean distance between the test vehicle vibration signal sequence and the reference vehicle vibration signal sequence in each of the remaining measurement areas after length alignment is:

[0045] For the reference vehicle vibration signal sequence (q n ) and the length of n test area test vehicle vibration signal sequence (c n ), the Euclidean distance d(q,c) between them is calculated as:

[0046]

[0047] Where i represents the i-th point in the sequence, n represents the reference vehicle vibration signal sequence (q n ) and the test driving vibration signal sequence in the test area (c n )’s length n.

[0048] In the above technical solution, the specific calculation method for calculating the correlation coefficient r(q,c) between the test vehicle vibration signal sequence and the reference vehicle vibration signal sequence in each measurement area after length alignment is:

[0049]

[0050] Among them, Cov(q,c) represents the reference driving vibration signal sequence (q n ) and the test driving vibration signal sequence in the test area (c n ), Var[q] is the covariance between the reference vehicle vibration signal sequence (q n ), Var[c] is the variance of the driving vibration signal sequence in the test area (c n )’s variance.

[0051] In the above technical solution, the probability density function is determined as follows:

[0052] Determine the corresponding probability density function type according to the shape distribution of the correlation coefficient histogram, and use the maximum likelihood estimation method to determine the parameters of the probability density function when the probability distribution D of the correlation coefficient is determined. D Parameter θ, for the probability distribution of the correlation coefficient population Z, its distribution law P{Z=z}=p(z;θ), where P{Z=z} represents the probability that the correlation coefficient takes the value z. Here, in fact, the data of the histogram has been processed, that is, the value and probability (frequency) of the correlation coefficient are processed. P(z;θ) represents the probability that the correlation coefficient value is equal to z. For n1 samplings (taking the value of the correlation coefficient once), the event {Z1=z1,Z2=z2,...,Z n1 =z n1 The probability of occurrence is The event indicates that the correlation coefficient value z1 is taken for the first time, the value taken for the second time is z2, etc., where Z1 indicates the first correlation coefficient value, z1 indicates the value taken for the second time is z1, and Z n1 Indicates the correlation coefficient value taken for the n1th time, z n1 Indicates the value is z n1 , i1 represents the i1th value, p(z i1 ; θ) indicates that the value of i1 is z i1 The probability of L(θ) is the likelihood function. θ when the likelihood function reaches its maximum value is the parameter of the probability density function. Then substitute the parameters into the probability density function. For example, the probability density function of the t-local scale distribution has a specific formula, but there are only three parameters that need to be calculated. After they are calculated, these three parameters are sufficient.

[0053] The probability density function is fitted to the histogram, and the parameters obtained by combining the given probability density function type with the maximum likelihood estimation are obtained, so that the histogram is parameterized by the probability density function and the influence of the randomness of the data on the histogram can be avoided. Figure 9 In the figure, the t-local scale distribution is selected as the fitting distribution function for the three cables, and the maximum likelihood estimation method is used to calculate the parameters of the probability density function of the t-local scale distribution.

[0054] Comparing the probability density functions together is more intuitive than comparing them with histograms. The closer the peak of the probability density function is to the right, the more the correlation coefficient tends to converge toward 1, which results in better consistency across the various measurement areas of the optical cable. When comparing the consistency of measurement areas between optical cables, based on the selected data, Type 1 cable has the best consistency, while Type 3 cable has the worst.

[0055] In the above technical solution, in the probability density function comparison diagram, the closer the peak point of the probability density function corresponding to the test vehicle vibration signal in the measurement area is to the right, the more the correlation coefficient of the test vehicle vibration signal in the measurement area tends to be concentrated in the direction of 1, and the corresponding test vehicle vibration signal in the measurement area is more consistent with the benchmark vehicle vibration signal sequence.

[0056] The probability density functions of the t-local scale distributions obtained for the three cables are:

[0057]

[0058] where θ = [μ, σ, ν] T , Γ(·) represents the gamma function, μ is the location parameter (mean), σ is the scale parameter, ν is the shape parameter, and x represents an arbitrary real number. Furthermore, the location parameter for Type 1 cable is 0.940993, the scale parameter is 0.020108, and the shape parameter is 1.34314. The location parameter for Type 2 cable is 0.938807, the scale parameter is 0.020409, and the shape parameter is 1.56289. The location parameter for Type 3 cable is 0.93587, the scale parameter is 0.0267103, and the shape parameter is 18.8716. When the probability density functions of the three cables are of the same type, comparing the peak locations of the probability density functions is ultimately a comparison of how close the mean of the probability density functions is to 1. The closer the mean (location parameter) is to 1, the better the consistency of the corresponding cable.

[0059] The above technical solution also includes a distribution map drawing module, which is used to use the measurement area sequence number of the measurement area test vehicle vibration signal sequence as the horizontal coordinate, and then the Euclidean distance and correlation coefficient between each measurement area test vehicle vibration signal sequence and the reference vehicle vibration signal sequence as the vertical coordinate, to obtain the distance distribution map and correlation coefficient distribution map of all measurement area test vehicle vibration signal sequences and the reference vehicle vibration signal sequence, and use the Euclidean distance and correlation coefficient formula to calculate the distance and correlation coefficient between each test sequence and the reference sequence, and use the measurement area number of the test sequence as the horizontal coordinate, and the distance and correlation coefficient as the vertical coordinate to form a distance distribution map between the signals of each measurement area of ​​the optical cable and the reference measurement area. Figure 6 、 Figure 7 As shown in FIG, the distance distribution diagram between the signals of each measurement area of ​​the optical cable and the reference measurement area is used to preliminarily reflect the consistency of the test vehicle vibration signal sequence in each measurement area and the reference vehicle vibration signal sequence.

[0060] A method for detecting consistency of a traffic road optical cable measurement area comprises the following steps:

[0061] Step 1: Obtain the driving vibration signal sequences of each measurement area along the traffic road to be measured, calculate the spectral entropy of the driving vibration signal sequences of each measurement area, and randomly select a driving vibration signal sequence of the measurement area from the driving vibration signal sequences of the measurement area whose spectral entropy is within a preset spectral entropy interval as the reference driving vibration signal sequence, and the remaining driving vibration signal sequences of the measurement area as the test driving vibration signal sequences of the measurement area;

[0062] Among them, two adjacent ultra-weak fiber Bragg gratings and the optical fiber between them in the traffic road distributed acoustic wave sensing system constitute a vehicle vibration signal measurement area. Each vehicle vibration signal measurement area can sense the corresponding vehicle vibration signal sequence in the measurement area.

[0063] Step 2: Perform dynamic time warping on the test train vibration signal sequences in each measurement area and the reference train vibration signal sequence, so that the length of each test train vibration signal sequence in each measurement area is aligned with the length of the reference train vibration signal sequence (because the time taken by the train to pass through each sensor is inconsistent, the length of the vibration signal caused by the train vibration when the train passes through the measurement area is inconsistent. To compare the consistency, it is necessary to align the signal lengths first);

[0064] Step 3: Calculate the similarity between the length-aligned test driving vibration signal sequences of each measurement area and the reference driving vibration signal sequence, and calculate the correlation coefficient between the length-aligned test driving vibration signal sequences of each measurement area and the reference driving vibration signal sequence;

[0065] Step 4: Plot the correlation coefficients between the test vehicle vibration signal sequences and the reference vehicle vibration signal sequences in all measurement areas and the frequencies of the corresponding correlation coefficients into a correlation coefficient histogram;

[0066] Step 5: Fit each correlation coefficient histogram with a probability density function to obtain a corresponding probability density function fitting graph. Combine each probability density function fitting graph to obtain a corresponding probability density function comparison graph. Compare the consistency of the test vehicle vibration signal in each measurement area with the reference vehicle vibration signal sequence based on the probability density function comparison graph.

[0067] The contents not described in detail in this specification belong to the prior art known to professional and technical personnel in this field.

Claims

1. A traffic road optical cable measurement area consistency detection system, characterized in that: It includes a vehicle vibration signal acquisition module, a signal length alignment module, a signal similarity calculation module, a histogram drawing module and a consistency comparison module; each vehicle vibration signal measurement area in the traffic road distributed acoustic wave sensing system can sense the vehicle vibration signal sequence in the corresponding measurement area; The driving vibration signal acquisition module is used to obtain driving vibration signal sequences of each measurement area along the traffic road to be measured, calculate the spectral entropy of the driving vibration signal sequence of each measurement area, and randomly select a driving vibration signal sequence of the measurement area from the driving vibration signal sequences of the measurement area whose spectral entropy is within a preset spectral entropy interval as a reference driving vibration signal sequence, and the remaining driving vibration signal sequences of the measurement area are used as test driving vibration signal sequences of the measurement area; The signal length alignment module is used to align the length of the test driving vibration signal sequence in each measurement area with the length of the reference driving vibration signal sequence; The signal similarity calculation module is used to calculate the correlation coefficient between the test driving vibration signal sequence of each measurement area and the reference driving vibration signal sequence after length alignment; The histogram drawing module is used to draw the correlation coefficients between the test driving vibration signal sequences and the reference driving vibration signal sequences in all measurement areas and the frequencies of the corresponding correlation coefficients into a correlation coefficient histogram; The consistency comparison module is used to obtain a corresponding probability density function fitting graph according to each correlation coefficient histogram, and to obtain a corresponding probability density function comparison graph according to each probability density function fitting graph.

2. The traffic road optical cable measurement area consistency detection system according to claim 1 is characterized by: The calculation method of the spectral entropy H(f) of the driving vibration signal sequence in the measurement area is: Among them, f is the frequency of the driving vibration signal sequence in the measurement area, f s is the sampling rate of the vehicle vibration signal sequence in the measurement area, and P(f) is the normalized power spectral density of the vehicle vibration signal sequence in the measurement area.

3. The traffic road optical cable measurement area consistency detection system according to claim 1 or 2, characterized in that: The signal length alignment module is used to perform dynamic time warping operations on the test driving vibration signal sequences of each measurement area and the reference driving vibration signal sequence, so that the lengths of the test driving vibration signal sequences of each measurement area are aligned with the length of the reference driving vibration signal sequence; The signal similarity calculation module is used to perform similarity calculation on the test driving vibration signal sequence of each measurement area after length alignment and the reference driving vibration signal sequence, and calculate the correlation coefficient between the test driving vibration signal sequence of each measurement area after length alignment and the reference driving vibration signal sequence. The consistency comparison module is used to fit each correlation coefficient histogram with a probability density function to obtain a corresponding probability density function fitting graph, combine each probability density function fitting graph to obtain a corresponding probability density function comparison graph, and compare the consistency of the test driving vibration signal in each measurement area with the benchmark driving vibration signal sequence based on the probability density function comparison graph.

4. The traffic road optical cable measurement area consistency detection system according to claim 1, characterized in that: Dynamic time warping can lengthen or shorten the test vehicle vibration signal sequence in each measurement area until the length of the test vehicle vibration signal sequence in each measurement area is consistent with the reference vehicle vibration signal sequence. During this process, the time axis of the test vehicle vibration signal sequence in each measurement area will be distorted or bent so that the sum of the Euclidean distances of all corresponding points in the two signals can be minimized after alignment.

5. The traffic road optical cable measurement area consistency detection system according to claim 1 or 4, characterized in that: The specific method for performing dynamic time warping is: For the reference vehicle vibration signal sequence q with a length of n0 n0 , and the test driving vibration signal sequence c of the test area with a length of m0 m0 , when m0 and n0 are not equal, construct an n0×m0 matrix and calculate the element q at (i,j) in the n0×m0 matrix i and c j Euclidean distance between two points For the reference driving vibration signal sequence q n0 And the test area driving vibration signal sequence c m0 Alignment is performed, that is, finding a monotonically continuous path starting from the (1,1) element in the matrix and ending at the (n0,m0) element, and the sum of the elements on the path is the smallest among all possible paths, so that the reference driving vibration signal sequence q indicated by the elements passed by the path is n0 And the test area driving vibration signal sequence c m0 The points in the path should appear in the same position in time. The monotonicity and continuity of the path mean that when an element (a′, b′) in the n0×m0 matrix is ​​selected as a point on the path, the next point (a, b) on the path should satisfy 0≤(aa′)≤1 and 0≤(bb′)≤1.

6. The traffic road optical cable measurement area consistency detection system according to claim 1, characterized in that: It also includes a distribution map drawing module, and the signal similarity calculation module is further used to calculate the Euclidean distance between the test driving vibration signal sequence of each measurement area and the reference driving vibration signal sequence after length alignment; The distribution map drawing module is used to use the measurement area sequence number of the measurement area test vehicle vibration signal sequence as the horizontal coordinate, and then the Euclidean distance and correlation coefficient between each measurement area test vehicle vibration signal sequence and the reference vehicle vibration signal sequence as the vertical coordinate, to obtain the distance distribution map and correlation coefficient distribution map of all measurement area test vehicle vibration signal sequences and the reference vehicle vibration signal sequence, use the Euclidean distance and correlation coefficient formula to calculate the distance and correlation coefficient between each test sequence and the reference sequence, and use the measurement area number of the test sequence as the horizontal coordinate and the distance and correlation coefficient as the vertical coordinate to form the distance distribution map between the signals of each measurement area of ​​the optical cable and the reference measurement area; The specific method for calculating the Euclidean distance between the test vehicle vibration signal sequence and the reference vehicle vibration signal sequence in each measurement area after length alignment is as follows: For the reference vehicle vibration signal sequence q with a length of n n and the length of the test area driving vibration signal sequence c n , the calculation formula of the Euclidean distance d(q,c) between them is: Where i represents the i-th point in the sequence, and n represents the reference vehicle vibration signal sequence q n And the test area driving vibration signal sequence c n The length n.

7. The traffic road optical cable measurement area consistency detection system according to claim 1, characterized in that: The specific calculation method for calculating the correlation coefficient r(q,c) between the test vehicle vibration signal sequence and the reference vehicle vibration signal sequence in the remaining measurement areas after length alignment is as follows: Where Cov(q,c) represents the reference vehicle vibration signal sequence q n And the test area driving vibration signal sequence c n The covariance between them, Var[q] is the reference driving vibration signal sequence q n Variance of the test vehicle vibration signal sequence c in the test area. n The variance of .

8. The traffic road optical cable measurement area consistency detection system according to claim 1, characterized in that: The probability density function is determined as follows: Determine the corresponding probability density function type according to the shape distribution of the correlation coefficient histogram, and use the maximum likelihood estimation method to determine the parameters of the probability density function when the probability distribution D of the correlation coefficient is determined. D The parameter θ, for the probability distribution of the correlation coefficient population Z, its distribution law P{Z=z}=p(z;θ), where P{Z=z} represents the probability that the correlation coefficient takes the value z, and p(z;θ) represents the probability that the correlation coefficient value is equal to z. For n1 samples, the event {Z1=z1,Z2=z2,...,Z n1 =z n1 The probability of occurrence is Among them, Z1 represents the first correlation coefficient value, z1 represents the value z1, Z n1 Indicates the correlation coefficient value taken for the n1th time, z n1 Indicates the value is z n1 , i1 represents the i1th value, p(z i1 ; θ) indicates that the value of i1 is z i1 The probability of L(θ) is the likelihood function, and θ when the likelihood function reaches its maximum value is the parameter of the probability density function.

9. The traffic road optical cable measurement area consistency detection system according to claim 1, characterized in that: In the probability density function comparison diagram, the closer the peak point of the probability density function corresponding to the test vehicle vibration signal in the measurement area is to the right, the more the correlation coefficient of the test vehicle vibration signal in the measurement area tends to be concentrated in the direction of 1, and the corresponding test vehicle vibration signal in the measurement area is more consistent with the reference vehicle vibration signal sequence.

10. A method for detecting consistency of a traffic road optical cable measurement area, characterized in that: It includes the following steps: Step 1: Obtain the driving vibration signal sequences of each measurement area along the traffic road to be measured, calculate the spectral entropy of the driving vibration signal sequences of each measurement area, and randomly select a driving vibration signal sequence of the measurement area from the driving vibration signal sequences of the measurement area whose spectral entropy is within a preset spectral entropy interval as the reference driving vibration signal sequence, and the remaining driving vibration signal sequences of the measurement area as the test driving vibration signal sequences of the measurement area; Among them, two adjacent ultra-weak fiber Bragg gratings and the optical fiber between them in the traffic road distributed acoustic wave sensing system constitute a vehicle vibration signal measurement area. Each vehicle vibration signal measurement area can sense the corresponding vehicle vibration signal sequence in the measurement area. Step 2: Perform dynamic time warping operations on the test vehicle vibration signal sequences in each measurement area and the reference vehicle vibration signal sequence, so that the length of the test vehicle vibration signal sequence in each measurement area is aligned with the length of the reference vehicle vibration signal sequence; Step 3: Calculate the similarity between the length-aligned test driving vibration signal sequences of each measurement area and the reference driving vibration signal sequence, and calculate the correlation coefficient between the length-aligned test driving vibration signal sequences of each measurement area and the reference driving vibration signal sequence; Step 4: Plot the correlation coefficients between the test vehicle vibration signal sequences and the reference vehicle vibration signal sequences in all measurement areas and the frequencies of the corresponding correlation coefficients into a correlation coefficient histogram; Step 5: Fit each correlation coefficient histogram with a probability density function to obtain a corresponding probability density function comparison diagram. Compare the consistency of the test vehicle vibration signal in each measurement area with the reference vehicle vibration signal sequence based on the probability density function comparison diagram.

Images