Urban road multi-risk evaluation method based on DAS and acoustic emission signal characteristics

Abstract

The invention discloses an urban road multi-risk evaluation method based on DAS and acoustic emission signal characteristics, and relates to the technical field of urban road safety monitoring. The method comprises the following steps: receiving voiceprint signals around existing communication optical fibers of urban roads acquired based on a plurality of channels of a DAS (Data Acquisition System); extracting a time domain, a frequency domain and composite features based on the acquired voiceprint signals around the existing communication optical fiber of the urban road, removing redundant features of the time domain, the frequency domain and the composite features through a Pearson correlation coefficient matrix, and generating optimized time domain, frequency domain and composite features; and for the optimized time domain, frequency domain and composite features, performing anomaly judgment by adopting an area outside a confidence interval of single feature density distribution and two-dimensional Gaussian distribution, and positioning an abnormal DAS channel in combination with principal component analysis dimension reduction and a DBSCAN clustering method. According to the invention, underground cavity hidden dangers of urban roads can be effectively detected, ground risk events can be classified, and urban road safety risk monitoring and evaluation are realized.

Description

Technical Field

[0001] The invention relates to the technical field of urban road safety monitoring, and in particular to an urban road multi-risk assessment method based on DAS and acoustic emission signal characteristics. Background Art

[0002] The long-term performance of urban roads is affected by a variety of external factors, leading to degradation and potential safety hazards. Some underground cavities or ground risks are particularly serious, which may cause irreversible damage to traffic facilities and people's lives and property. Traditional road hazard detection methods rely on local sensors, with limited monitoring range and difficulty in dealing with complex underground environments. Distributed fiber acoustic sensing (DAS) technology has been applied in many fields due to its advantages such as low cost, high sensitivity, anti-electromagnetic interference and corrosion resistance. However, DAS technology faces challenges such as low signal-to-noise ratio, depth uncertainty and uneven fiber response in urban road monitoring. Current research is mostly based on signal processing of a single spatial point, ignoring the spatial correlation between the positions of the optical fiber, and not considering the unevenness of the optical fiber response, which makes it difficult to accurately identify underground cavities and ground risk hazards, reducing the accuracy and efficiency of urban road safety monitoring. To this end, the present invention proposes a multi-risk assessment method for urban roads based on DAS and acoustic emission signal characteristics. Summary of the invention

[0003] The purpose of the present invention is to provide a multi-risk assessment method for urban roads based on DAS and acoustic emission signal characteristics. By combining DAS technology and acoustic emission signal characteristics, it can effectively detect underground cavity hazards and classify ground risk events.

[0004] According to a first aspect of the present invention, in order to achieve the above-mentioned purpose, the present invention provides the following technical solution: a multi-risk assessment method for urban roads based on DAS and acoustic emission signal characteristics, comprising the following steps: Receive voiceprint signals around existing communication optical fibers on urban roads collected through multiple DAS channels; Based on the collected voiceprint signals around the existing communication optical fibers in urban roads, time domain, frequency domain and composite features are extracted, and the redundant features of the extracted time domain, frequency domain and composite features are removed through the Pearson correlation coefficient matrix to generate optimized time domain, frequency domain and composite features; For the optimized time domain, frequency domain and composite features, single feature density distribution and two-dimensional Gaussian distribution are used to judge the abnormality of the area outside the confidence interval, and the principal component analysis dimensionality reduction and DBSCAN clustering method are combined to locate the abnormal DAS channel for identifying underground cavity hazards; The phase cross-correlation function between multiple DAS channels is calculated, and the sharpness correlation peak index κ is used to compare the similarity between multiple DAS channels, and finally the reliability index β of each channel is obtained; The calculated reliability index β is distributed and statistically analyzed, and the risk event characteristics are quantified using mean and variance parameters. The principal component dimension reduction and K-means clustering method are combined to achieve ground risk event classification and early warning. Urban road safety risk monitoring and evaluation can be achieved based on the identification and positioning of underground cavity hazards and the classification and early warning of ground risk events.

[0005] Furthermore, the voiceprint signals around the existing communication optical fibers on urban roads are collected using optical fiber sensors based on a distributed optical fiber acoustic wave sensing system.

[0006] Furthermore, the phase cross-correlation function between DAS channels is calculated, and the sharpness correlation peak index κ is used to compare the similarity between the fiber channels, and finally the reliability index β of each channel is obtained, as follows: Each DAS channel i With all other channels j For comparison, j =1,…, N , N is the total number of DAS longitudinal channels, and the channel pairs are calculated Phase cross-correlation function PCCF ij And using the sharpness correlation peak index κ ij For comparison, the peak-to-mean-square ratio is used to estimate the sharpness of the PCCF, and the similarity indicator is defined as κ ,Right now: (1) Where: , , represents each independent variable in the W window, represents the RMS value of the window W around the main correlation peak, Represents the sampling period, that is, through a length of 2 L , centered at Window W , and excludes the main correlation peak at In this way, we get the value of i Channel vector κ i =[ κ i,1 , κ i,2 ,…, κ i,N ]; In order to estimate each DAS channel i Reliability, calculation κ iThe RMS value of the vector and the DAS channel i The reliability is defined as β i ,Right now: (2) in, j ≠ i , and this process is repeated for all channels in turn, resulting in a 1× N vector β i =[ β 1 , β 2 ,…, β N ], which contains the reliability indicators of all channels, by comparing the reliability indicators of each channel with other channels or waveforms , and obtain a preliminary judgment on the reliability of each channel.

[0007] Furthermore, based on the collected voiceprint signals around the existing communication optical fibers on urban roads, time domain, frequency domain and composite features are extracted, and the redundant features of the extracted time domain, frequency domain and composite features are removed through the Pearson correlation coefficient matrix to generate optimized time domain, frequency domain and composite features, as follows: (41) Extract time domain, frequency domain and composite features, including duration, amplitude, energy, zero crossing rate, number of zero crossings, rise time, decay time, time centroid, entropy, count, peak count, rise angle, decay angle, rise time to amplitude ratio, rise time to duration ratio, duration to amplitude ratio, energy to amplitude ratio, dimensionless amplitude, count to duration ratio, partial power, average frequency, frequency centroid, peak frequency, spectrum spread, spectrum skewness, spectrum kurtosis, upper roll-off frequency, lower roll-off frequency, start frequency, reverberation frequency, weighted peak frequency; (42) The Pearson correlation coefficient matrix is ​​used to screen out features with low correlation and independent information, retain a number of the most discriminative features, and reduce redundant features.

[0008] Furthermore, the single feature density distribution and two-dimensional Gaussian distribution are used to determine the anomaly in the area outside the confidence interval, and the principal component analysis dimensionality reduction and DBSCAN clustering method are combined to locate the abnormal DAS channel, which includes the following: (51) Abnormal channel identification based on two-dimensional Gaussian distribution fitting; (51.1) Firstly, the characteristic index calculation method of acoustic emission signal processing is introduced to select the wave and extract the characteristics of the signal of each channel of DAS; (51.2) Use kernel density estimation to plot the probability density distribution of a single feature for each channel, and obtain the eigenvalue and probability density corresponding to the peak point of the probability density curve; (51.3) K-S Gaussian distribution test and data transformation; Perform one-dimensional Gaussian distribution fitting and discrimination on the eigenvalue and probability density of the peak point respectively. If it does not conform to the one-dimensional Gaussian distribution, the square root transformation method needs to be used for transformation. If it conforms to the one-dimensional Gaussian distribution, the original data is retained without adjustment; (51.4) Perform two-dimensional Gaussian distribution fitting on the eigenvalue and its probability density of the peak point, calculate its confidence interval, and determine the channels outside the confidence interval as abnormal channels; (52) Identification of abnormal channels based on DBSCAN clustering; (52.1) Use principal component analysis to perform dimensionality reduction analysis on multivariate features, solve the eigenvalues and arrange them in descending order according to the eigenvalue size, and select the first three principal components for three-dimensional visualization; Through orthogonal transformation, the relevant eigenvalues are transformed into a set of independent and uncorrelated variables, that is, principal components. The analysis process includes normalization of eigenvalue data, calculation of covariance matrix, solution of eigenvectors and eigenvalues, and projection of data into a new coordinate system composed of principal components; (52.2) Based on the distribution of the first three principal components, apply the DBSCAN clustering algorithm to perform clustering analysis on the data; By specifying the distance threshold and the minimum number of samples, identify the high-density clustering regions from the data, and at the same time identify the sparsely distributed samples as noise points, so as to further identify the noise points and determine the corresponding channels.

[0009] Further, perform distribution statistics on the calculated reliability index β, quantify the characteristics of risk events using the mean and variance, and combine the PCA dimensionality reduction and K-means clustering methods to achieve accurate classification and early warning of multiple risk event categories, specifically as follows: (61) Classification of ground risk events based on distribution parameter statistics of the reliability index β; (61.1) For the reliability index calculated for a period of data β , its numerical distribution is respectively fitted with normal distribution, lognormal distribution, exponential distribution, Weibull distribution and gamma distribution, and the Kolmogorov-Smirnov hypothesis test is used to verify the fitting results, and the mean and variance statistical parameters of each distribution are obtained. The calculation formulas of the above normal distribution, lognormal distribution, exponential distribution, Weibull distribution and gamma distribution are shown as follows: (3) (4) (5) (6) (7) Wherein, x is the fitted data, and the μ and σ of the normal distribution in formula (3) are the mean and standard deviation respectively; the and of the lognormal distribution in formula (4) are the mean and standard deviation after logarithmic transformation; the λ of the exponential distribution in formula (5) is the rate parameter; the of the Weibull distribution in formula (6) is the scale parameter, k is the shape parameter; the α of the gamma distribution in formula (7) is the shape parameter, β is the scale parameter, Γ Γ(α) is the gamma function; (61.2) Analyze the mean and variance statistics obtained from the distribution fitting in step (61.1), and calculate the probability density histogram distribution and the fitting distribution curve of the β value; (61.3) Then, statistically analyze the dispersion degree of the β value distribution of different events through a box plot, and the discrimination and distinction of data of different events can be initially realized; (62) Classification of ground risk events based on k-means clustering; For the data of different events, select the maximum reliability channel in their respective β curves as the subsequent calculation channel, and perform wave selection and multi-feature calculation on it (62.1) Use principal component analysis to reduce the dimension of multivariate features. Through orthogonal transformation, the relevant features are converted into a set of mutually independent principal components, so as to extract the main information of the data and reduce the dimension; (62.2) Apply the unsupervised learning algorithm k-means to perform clustering analysis on the dimension-reduced data; The k-means algorithm divides the data into a specified number of clusters through iterative optimization, and each cluster is centered on its centroid, so as to further realize the classification and recognition of different events.

[0010] According to the second aspect of the present invention, the present invention provides an urban road multi-risk evaluation system based on DAS and acoustic emission signal characteristics, which is used to implement the above-mentioned urban road multi-risk evaluation method based on DAS and acoustic emission signal characteristics, including: A receiving module, which receives the acoustic fingerprint signals around the existing communication optical fibers of urban roads collected by DAS; The feature extraction module is used to extract time domain, frequency domain and composite features based on the collected voiceprint signals around the existing communication optical fibers on urban roads, and remove redundant features of the extracted time domain, frequency domain and composite features through the Pearson correlation coefficient matrix to generate optimized time domain, frequency domain and composite features; The abnormal channel identification module is used to judge the abnormality of the area outside the confidence interval using the single feature density distribution and two-dimensional Gaussian distribution for the optimized time domain, frequency domain and composite features, and to locate the abnormal DAS channel by combining the principal component analysis dimensionality reduction and DBSCAN clustering method to identify the hidden dangers of underground cavities; A calculation module is used to calculate the phase cross-correlation function between DAS channels and compare the similarity between DAS channels using the sharpness correlation peak index κ, and finally obtain the reliability index β of each channel; The ground risk event classification module is used to perform distribution statistics on the calculated reliability index β, quantify the risk event characteristics using mean and variance parameters, and combine principal component dimension reduction and K-means clustering methods to achieve ground risk event classification and early warning; The monitoring and evaluation module is used to monitor and evaluate urban road safety risks based on the identification and location of underground cavity hazards and the classification and early warning of ground risk events.

[0011] Furthermore, it also includes a collection module, which is configured as a distributed optical fiber acoustic wave sensing system.

[0012] According to a third aspect of the present invention, the present invention provides a terminal device, comprising a memory, a processor, and a computer program stored in the memory and capable of running on the processor, wherein the memory stores a computer program capable of running on the processor, and when the processor loads and executes the computer program, the above-mentioned urban road multi-risk assessment method based on DAS and acoustic emission signal characteristics is adopted.

[0013] According to a fourth aspect of the present invention, the present invention provides a storage medium comprising computer executable instructions, which, when executed by a computer processor, are used to execute the above-mentioned urban road multi-risk assessment method based on DAS and acoustic emission signal characteristics.

[0014] The present invention has at least the following beneficial effects: 1. The present invention can effectively detect underground cavity hazards in urban roads and classify ground risk events by combining DAS technology and acoustic emission signal characteristics, thereby improving the accuracy and efficiency of urban road safety monitoring; 2. The present invention calculates the channel reliability index by using the phase cross-correlation function (PCCF), and combined with the acoustic emission characteristics and clustering algorithm, can accurately identify abnormal channels, quantify the characteristics of risk events, and provide accurate classification of multiple risk events; 3. The present invention has a high level of intelligence, can monitor the safety status of urban roads in real time, provide early warning information, and is suitable for road risk assessment in a variety of complex scenarios.

[0015] Of course, any product implementing the present invention does not necessarily need to achieve all of the advantages described above at the same time. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 It is a schematic diagram of the process of the method of the present invention; Figure 2 It is a schematic diagram of the calculation process of the reliability index β in the present invention; Figure 3 It is a schematic diagram of the abnormal channel identification process in the present invention.

[0017] Figure 4 It is a schematic diagram of the ground risk event classification process in the present invention. DETAILED DESCRIPTION

[0018] The following will be combined with the drawings in the embodiments of the present disclosure to clearly and completely describe the technical solutions in the embodiments of the present disclosure. Obviously, the described embodiments are only part of the embodiments of the present disclosure, rather than all the embodiments. Based on the embodiments in the present disclosure, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present disclosure.

[0019] See also Figure 1 The present invention provides a technical solution: a multi-risk assessment method for urban roads based on DAS and acoustic emission signal characteristics, comprising the following steps: S1. Receive the voiceprint signal around the existing communication optical fiber in the urban road based on DAS collection; The acoustic wave signals propagated by optical fibers are collected based on the distributed optical fiber acoustic wave sensing system. The distributed optical fiber acoustic wave sensing (DAS) system uses optical fiber sensors to monitor the acoustic wave signals propagated along the optical fiber, realizing real-time monitoring of underground and ground vibrations of urban roads, and providing data support for subsequent risk assessment; S2. Extract the time domain, frequency domain and composite features based on the collected voiceprint signals around the existing communication optical fiber in the urban road, and remove the redundant features of the extracted time domain, frequency domain and composite features through the Pearson correlation coefficient matrix to generate optimized time domain, frequency domain and composite features, as follows: Extract 31 time-domain, frequency-domain, and composite features, specifically including duration, amplitude, energy, zero-crossing rate, number of zero crossings, rise time, decay time, temporal centroid, entropy, count, peak count, rise angle, decay angle, ratio of rise time to amplitude, ratio of rise time to duration, ratio of duration to amplitude, ratio of energy to amplitude, non-dimensional amplitude, ratio of count to duration, partial power, average frequency, frequency centroid, peak frequency, spectral spread, spectral skewness, spectral kurtosis, upper frequency roll-off frequency, lower frequency roll-off frequency, start frequency, reverberation frequency, weighted peak frequency; (S22)Screen redundant features through the Pearson correlation coefficient, and retain 16 most discriminative features, including Entropy, Duration, Spectral Spread, Amplitude, Zero Crossing Rate, Zero Crossings, Temporal Centroid, Rise Angle, Decay Angle, RA, ratio of rise time to duration (Risetime / Duration), ratio of duration to amplitude (Duration / Amplitude), non-dimensional amplitude (Non-dimensional Amplitude), ratio of duration to count (Counts / Duration), spectral kurtosis (Spectral Kurtosis), and roll-off frequency (Roll-off frequency); S3. For the optimized time-domain, frequency-domain, and composite features, use the single-feature density distribution and the region outside the confidence interval of the two-dimensional Gaussian distribution to judge anomalies, and combine principal component analysis for dimensionality reduction and the DBSCAN clustering method to locate the abnormal channels, as Figure 3 shown below: (S31)Identification of abnormal channels based on the two-dimensional Gaussian distribution fitting; (S31.1)First, introduce the feature index calculation method for acoustic emission signal processing, and perform wave selection and feature extraction on the signals of each channel of DAS; (S31.2)Use kernel density estimation to plot the probability density distribution of a single feature for each channel (kernel density estimation of a single feature for N channels), and obtain the eigenvalue and probability density corresponding to the peak point of the probability density curve; (S31.3)K-S Gaussian distribution test and data transformation; Perform one-dimensional Gaussian distribution fitting and discrimination on the eigenvalue and probability density of the peak point respectively. If it does not conform to the one-dimensional Gaussian distribution, the square root transformation method needs to be used for transformation. If it conforms to the one-dimensional Gaussian distribution, the original data is retained without adjustment; (S31.4) Perform a two-dimensional Gaussian distribution fitting on the eigenvalue and its probability density of the peak point, calculate its confidence interval, and determine the channels outside the confidence interval as abnormal channels (calculate the confidence interval by two-dimensional Gaussian distribution fitting); (S32) Abnormal channel identification based on DBSCAN clustering; (S32.1) Use principal component analysis to perform dimensionality reduction analysis on multivariate features, arrange them in descending order according to the eigenvalue size, and select the first three principal components for three-dimensional visualization; Through orthogonal transformation, transform the relevant eigenvalues (the selected M feature indicators) into a set of mutually independent and uncorrelated variables (the waveform of each channel is represented by an M-dimensional vector), that is, the principal components. The analysis process includes the normalization of eigenvalue data, the calculation of the covariance matrix, the solution of eigenvectors and eigenvalues, and the projection of data into the new coordinate system composed of the principal components (principal component analysis feature dimensionality reduction). For the convenience of visualization and subsequent analysis; Specifically, first standardize each feature, then calculate the covariance matrix of the standardized features, decompose the covariance matrix, then obtain the eigenvalues, arrange them in descending order according to the eigenvalue size, and select the first three principal components; (S32.2) Based on the distribution of the first three principal components, apply the DBSCAN clustering algorithm to perform clustering analysis on the data; By specifying the distance threshold and the minimum number of samples, identify the clustering regions with higher density from the data, and at the same time identify the sparsely distributed samples as noise points, so as to further identify the noise points and determine the corresponding channels (DBSCAN identifies the channels corresponding to the noise points), regarded as abnormal channels, for identifying potential hidden dangers of underground cavities; S4. Calculate the phase cross-correlation function between multiple channels of DAS, and use the sharpness correlation peak index κ to compare the similarity between each DAS channel, and finally obtain the reliability index β of each channel, as Figure 2 shown, specifically as follows: For each DAS input channel i Compare with all other channels j where j = 1, …, N , N is the total number of DAS longitudinal channels, calculate the phase cross-correlation function PCCF between channels ij and use the sharpness correlation peak index κ ij to compare, where the peak root mean square ratio (PRMSR) is used to estimate the sharpness of PCCF, and the similarity indicator κ is defined, that is: (1) In the formula: , , represents each independent variable within window W, represents the root mean square value of window W around the main correlation peak, represents the sampling period, that is, through a window of length 2 L and centered at , and excluding the value of the main correlation peak at W ; in this way, the th channel's i is obtained vector κ i = κ i,1 , κ i,2 ,…, κ i,N ; to estimate the reliability of each DAS channel i , calculate the root mean square (RMS) value of the κ i vector, and define the reliability of DAS channel i as β i , that is: (2) wherein, j ≠ i , in addition, this process is repeated sequentially for all channels, thereby obtaining a vector βi = β 1 , β 2 ,…, β N , which contains the reliability indices of all channels. By comparing the reliability indices of each channel with those of other channels or waveforms , a preliminary judgment on the reliability of each channel is obtained; S5. Conduct distribution statistics on the calculated reliability index β, quantify the characteristics of risk events using the mean and variance parameters, and combine PCA dimensionality reduction and K-means clustering methods to achieve precise classification and early warning of multiple risk event categories, as Figure 4 shown, specifically as follows: (S51) Classification of ground risk events based on the distribution parameter statistics of the reliability index β; (S51.1) For the reliability index calculated for a period of data , its numerical distribution is respectively fitted by normal distribution, lognormal distribution, exponential distribution, Weibull distribution and gamma distribution, and the Kolmogorov-Smirnov (K-S) hypothesis test is used to verify the fitting results, obtaining the mean and variance statistical parameters of each distribution. The calculation formulas of the above normal distribution, lognormal distribution, exponential distribution, Weibull distribution and gamma distribution are shown as follows: (3) (4) (5) (6) (7) In the formula, x is the fitted data. For the normal distribution in formula (3), μ and σ are the mean and standard deviation respectively; for the lognormal distribution in formula (4), and are the mean and standard deviation after logarithmic transformation; for the exponential distribution in formula (5), λ is the rate parameter; for the Weibull distribution in formula (6), is the scale parameter, k is the shape parameter; for the gamma distribution in formula (7), α is the shape parameter, β is the scale parameter, Γ (α) is the gamma function; (S51.2) By analyzing the mean and variance statistics obtained from the distribution fitting in step (S51.1) (specifically including normal distribution fitting, lognormal distribution fitting, exponential distribution, Weibull distribution fitting, gamma distribution fitting), calculate the probability density histogram distribution and the fitting distribution curve of the β value; (S51.3) Then, through the box plot, statistically analyze the dispersion degree of the β value distribution of different events, and initially realize the discrimination and differentiation of data of different events (β value - box plot distribution statistics); (S52) Classification of ground risk events based on k-means clustering; For the data of different events, select the maximum reliability channel in their respective β curves as the subsequent calculation channel, and perform wave selection and multi-feature calculation on it (S52.1) Use principal component analysis to reduce the dimension of multi-features, (the selected M feature indicators) are transformed into a set of independent and uncorrelated variables (each channel waveform is represented by an M-dimensional vector), that is, the principal components, so as to extract the main information of the data and reduce the dimension; (S52.2) Apply the unsupervised learning algorithm k-means to perform cluster analysis on the reduced dimensionality data; The k-means clustering algorithm divides the data into a specified number of clusters through iterative optimization, with each cluster centered on its centroid, thereby further achieving the classification and identification of different events, such as event 1, event 2, and event 3; S6. According to the results of step S3 to step S5, a multi-risk assessment model for urban roads is established; underground cavity hazards and ground risk event classification are visualized, and based on the identification and location of underground cavity hazards and the classification and early warning of ground risk events, urban road safety risk monitoring and assessment are achieved.

[0020] Next, the present invention is further described in conjunction with specific embodiments: In this embodiment, a test site was built on a section of urban road in City A that had been completed but not yet put into use. Four event conditions were set (fixed-point tamping by a rammer, fixed-point striking by a heavy hammer, heavy-loaded vehicle driving, and silent events). DAS data was collected and processed and analyzed. Through experimental data analysis, the effectiveness of the method in abnormal channel identification and ground risk event classification was verified.

[0021] This embodiment conducts actual road tests on a certain road section in City B. Real road data is collected by connecting the DAS device to the existing communication optical fiber, and a heavy hammer is used to actively stimulate the road surface to detect cavities or manhole covers around the optical fiber. The experimental data are analyzed to verify the effectiveness of the method in abnormal channel identification and ground risk event classification in an actual road environment.

[0022] Embodiment 2: This embodiment provides a system for evaluating multiple risks of urban roads based on DAS and acoustic emission signal characteristics, which is used to implement the above-mentioned method for evaluating multiple risks of urban roads based on DAS and acoustic emission signal characteristics, including: A receiving module receives voiceprint signals collected around existing communication optical fibers on urban roads using DAS; The feature extraction module is used to extract time domain, frequency domain and composite features based on the collected voiceprint signals around the existing communication optical fibers on urban roads, and remove redundant features of the extracted time domain, frequency domain and composite features through the Pearson correlation coefficient matrix to generate optimized time domain, frequency domain and composite features; The abnormal channel identification module is used to judge the abnormality of the area outside the confidence interval using the single feature density distribution and two-dimensional Gaussian distribution for the optimized time domain, frequency domain and composite features, and to locate the abnormal DAS channel by combining the principal component analysis dimensionality reduction and DBSCAN clustering method to identify the hidden dangers of underground cavities; A calculation module, which is used to calculate the phase cross-correlation function between DAS channels, compare the similarities between DAS channels by using the sharpness correlation peak index κ, and finally obtain the reliability index β of each channel; A ground risk event classification module, which is used to perform distribution statistics on the calculated reliability index β, quantify the characteristics of risk events by using the mean and variance parameters, and combine the principal component dimensionality reduction and K-means clustering methods to achieve ground risk event classification and early warning; A monitoring and evaluation module, which is used to realize the monitoring and evaluation of urban road safety risks based on the identification and location of underground cavity hidden dangers and the classification and early warning of ground risk events.

[0023] Specifically, the above receiving module, calculation module, feature extraction module, abnormal channel identification module, ground risk event classification module and monitoring and evaluation module can be embedded in a computer processing system. The computer, based on the above-provided multi-risk evaluation method for urban roads based on DAS and acoustic emission signal characteristics, calls the above modules to complete the task of precise classification and early warning of multi-risk event categories; the above receiving module, calculation module, feature extraction module, abnormal channel identification module, ground risk event classification module and monitoring and evaluation module can perform operations according to the specific steps given by the above-mentioned multi-risk evaluation method for urban roads based on DAS and acoustic emission signal characteristics.

[0024] It should be noted that it should be understood that the division of each module of the above system is only a logical function division. In actual implementation, it can be fully or partially integrated into a physical entity, or physically separated, and these modules can all be implemented in the form of software called by processing elements; they can also all be implemented in the form of hardware; they can also be partially implemented in the form of software called by processing elements and partially implemented in the form of hardware. For example, the receiving module can be a separately established processing element, or can be integrated in a certain chip of the above device. In addition, it can also be stored in the memory of the above device in the form of program code, and called and executed by a certain processing element of the above device to perform the functions of the above signal processing module, and the implementation of other modules is similar. In addition, all or part of these modules can be integrated together or independently implemented. The processing element mentioned here can be an integrated circuit with signal processing capabilities. In the implementation process, each step of the above method or each of the above modules can be completed by the hardware integrated logic circuit or software-form instructions in the processor element.

[0025] For example, the above-mentioned modules can be one or more integrated circuits configured to implement the above methods, such as: one or more Application Specific Integrated Circuits (ASICs), or, one or more Digital Signal Processors (DSPs), or, one or more Field Programmable Gate Arrays (FPGAs), etc. Again, when a certain module above is implemented in the form of a processing element scheduling program code, the processing element can be a general-purpose processor, such as a Central Processing Unit (CPU) or other processors that can call program code. Again, these modules can be integrated together and implemented in the form of a system-on-a-chip (SOC).

[0026] Embodiment 3: The present invention provides a terminal device, including a memory, a processor, and a computer program stored in the memory and capable of running on the processor. The computer program stored in the memory can run on the processor. When the processor loads and executes the computer program, the above-mentioned urban road multi-risk assessment method based on DAS technology and acoustic emission signals is adopted.

[0027] It should be noted that the terminal device can be a computer device such as a desktop computer, a laptop computer, or a cloud server. And the terminal device includes but is not limited to a processor and a memory. For example, the terminal device may further include input / output devices, network access devices, and a bus, etc.

[0028] Furthermore, the processor can adopt a Central Processing Unit (CPU). Of course, according to the actual usage situation, other general-purpose processors, Digital Signal Processors (DSPs), Application Specific Integrated Circuits (ASICs), off-the-shelf Programmable Gate Arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. can also be adopted. The general-purpose processor can adopt a microprocessor or any conventional processor, etc. This application does not make any restrictions in this regard.

[0029] Embodiment 4: The present invention provides a storage medium containing computer-executable instructions, and the computer-executable instructions are used to execute the above-mentioned urban road multi-risk assessment method based on DAS technology and acoustic emission signals when executed by a computer processor.

[0030] Among them, the computer program can be stored in a computer-readable medium. The computer program includes computer program code, which can be in the form of source code, object code, executable file, or some middleware form, etc. The computer-readable medium includes any entity or device capable of carrying the computer program code, recording medium, USB flash drive, mobile hard disk, magnetic disk, optical disc, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signal, telecommunication signal, and software distribution medium, etc. It should be noted that the computer-readable medium includes but is not limited to the above components.

[0031] It should be noted that in this article, relational terms such as first and second are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the term "comprising", "including" or any other variant thereof is intended to cover non-exclusive inclusion, so that a process, method, article or device comprising a series of elements not only includes those elements, but also includes other elements not expressly listed, or also includes elements inherent to such process, method, article or device.

[0032] For those of ordinary skill in the art, the specific meanings of the above terms in the present invention can be understood according to specific circumstances. When an element is referred to as being "assembled on", "installed on", "fixed to" or "disposed on" another element, it can be directly on the other element or there may also be an intermediate element. When an element is considered to be "connected" to another element, it can be directly connected to the other element or there may be an intermediate element at the same time. The terms "vertical", "horizontal", "upper", "lower", "left", "right" and similar expressions used herein are only for the purpose of illustration and do not represent the only implementation.

[0033] Although the embodiments of the present invention have been shown and described, for those of ordinary skill in the art, it can be understood that various changes, modifications, substitutions and variations can be made to these embodiments without departing from the principles and spirit of the present invention. The scope of the present invention is defined by the appended claims and their equivalents.

[0034] In the description of this specification, the description with reference to terms such as "one embodiment", "example", "specific example", etc. means that the specific features, structures, materials or characteristics described in connection with the embodiment or example are included in at least one embodiment or example of the present disclosure. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described can be combined in a suitable manner in any one or more embodiments or examples.

Claims

1. A multi-risk assessment method for urban roads based on DAS and acoustic emission signal characteristics, characterized in that: include; Receive voiceprint signals around existing communication optical fibers on urban roads collected through multiple DAS channels; Based on the collected voiceprint signals around the existing communication optical fibers in urban roads, time domain, frequency domain and composite features are extracted, and the redundant features of the extracted time domain, frequency domain and composite features are removed through the Pearson correlation coefficient matrix to generate optimized time domain, frequency domain and composite features; For the optimized time domain, frequency domain and composite features, single feature density distribution and two-dimensional Gaussian distribution are used to judge the abnormality of the area outside the confidence interval, and the principal component analysis dimensionality reduction and DBSCAN clustering method are combined to locate the abnormal DAS channel for identifying underground cavity hazards; The phase cross-correlation function between multiple DAS channels is calculated, and the sharpness correlation peak index κ is used to compare the similarity between multiple DAS channels, and finally the reliability index β of each channel is obtained; The calculated reliability index β is distributed and statistically analyzed, and the risk event characteristics are quantified using mean and variance parameters. The principal component dimension reduction and K-means clustering method are combined to achieve ground risk event classification and early warning. Urban road safety risk monitoring and evaluation can be achieved based on the identification and positioning of underground cavity hazards and the classification and early warning of ground risk events.

2. The urban road multi-risk assessment method based on DAS and acoustic emission signal characteristics according to claim 1 is characterized in that: The voiceprint signals around existing communication optical fibers on urban roads are collected using optical fiber sensors based on a distributed optical fiber acoustic wave sensing system.

3. The urban road multi-risk assessment method based on DAS and acoustic emission signal characteristics according to claim 2 is characterized in that: The phase cross-correlation function between DAS channels is calculated, and the sharpness correlation peak index κ is used to compare the similarity between the fiber channels, and finally the reliability index β of each channel is obtained, as follows: Each DAS channel i With all other channels j For comparison, j =1,…, N , N is the total number of DAS longitudinal channels, and the channel pairs are calculated Phase cross-correlation function PCCF ij And using the sharpness correlation peak index κ ij For comparison, the peak-to-mean-square ratio is used to estimate the sharpness of the PCCF, and the similarity indicator is defined as κ ,Right now: （1） Where: , , represents each independent variable in the W window, represents the RMS value of the window W around the main correlation peak, Represents the sampling period, that is, through a length of 2 L , centered at Window W , and excludes the main correlation peak at In this way, we get the value of i Channel vector κ i =[ κ i,1 , κ i,2 ,…, κ i,N ]; In order to estimate each DAS channel i Reliability, calculation κ i The RMS value of the vector and the DAS channel i The reliability is defined as β i ,Right now: （2） in, j ≠ i , and this process is repeated for all channels in turn, resulting in a vector βi =[ β 1, β 2,…, β N ], which contains the reliability indicators of all channels, by comparing the reliability indicators of each channel with other channels or waveforms , and obtain a preliminary judgment on the reliability of each channel.

4. The urban road multi-risk assessment method based on DAS and acoustic emission signal characteristics according to claim 3 is characterized in that: Based on the collected voiceprint signals around the existing communication optical fibers in urban roads, the time domain, frequency domain and composite features are extracted, and the redundant features of the extracted time domain, frequency domain and composite features are removed through the Pearson correlation coefficient matrix to generate optimized time domain, frequency domain and composite features, as follows: (41) The extracted time domain, frequency domain and composite features specifically include duration, amplitude, energy, zero crossing rate, number of zero crossings, rise time, decay time, time centroid, entropy, count, peak count, rise angle, decay angle, rise time to amplitude ratio, rise time to duration ratio, duration to amplitude ratio, energy to amplitude ratio, dimensionless amplitude, count to duration ratio, partial power, average frequency, frequency centroid, peak frequency, spectrum spread, spectrum skewness, spectrum kurtosis, upper roll-off frequency, lower roll-off frequency, start frequency, reverberation frequency, and weighted peak frequency; (42) The Pearson correlation coefficient matrix is ​​used to screen out features with low correlation and independent information, retain a number of the most discriminative features, and reduce redundant features.

5. The urban road multi-risk assessment method based on DAS and acoustic emission signal characteristics according to claim 4 is characterized in that: The single feature density distribution and two-dimensional Gaussian distribution are used to determine the anomaly in the area outside the confidence interval, and the principal component analysis dimensionality reduction and DBSCAN clustering method are combined to locate the abnormal DAS channel, including the following: (51) Abnormal channel identification based on two-dimensional Gaussian distribution fitting; (51.1) Firstly, the characteristic index calculation method of acoustic emission signal processing is introduced to select the wave and extract the characteristics of the signal of each channel of DAS; (51.2) Use kernel density estimation to draw the probability density distribution of a single feature of each channel, and obtain the eigenvalue and probability density corresponding to the peak point of the probability density curve; (51.3) KS Gaussian distribution test and data transformation; The characteristic value and probability density of the peak point are fitted and judged by one-dimensional Gaussian distribution respectively. If they do not conform to the one-dimensional Gaussian distribution, they need to be transformed by the square root transformation method. If they conform to the one-dimensional Gaussian distribution, the original data is retained without adjustment. (51.4) Perform two-dimensional Gaussian distribution fitting on the characteristic value of the peak point and its probability density, and calculate its confidence interval. Channels outside the confidence interval are identified as abnormal channels. (52) Abnormal channel identification based on DBSCAN clustering; (52.1) Use principal component analysis to reduce the dimension of multivariate features, solve the eigenvalues ​​and arrange them in descending order according to the size of the eigenvalues, and select the first three principal components of the arrangement for three-dimensional visualization; The related eigenvalues ​​are transformed into a set of independent and unrelated variables, namely, principal components, through orthogonal transformation. The analysis process includes normalization of eigenvalue data, calculation of covariance matrix, solution of eigenvectors and eigenvalues, and projection of data into a new coordinate system composed of principal components. (52.2) Based on the distribution of the first three principal components, the DBSCAN clustering algorithm was used to perform cluster analysis on the data; By specifying the distance threshold and the minimum number of samples, high-density clustering areas are identified from the data, and sparsely distributed samples are identified as noise points, so as to further identify the noise points and determine the corresponding channels.

6. The urban road multi-risk assessment method based on DAS and acoustic emission signal characteristics according to claim 5 is characterized in that: The calculated reliability index β is distributed and statistically analyzed, and the mean and variance parameters are used to quantify the risk event characteristics. The PCA dimension reduction and K-means clustering method are combined to achieve accurate classification and early warning of multiple risk event categories, as follows: (61) Classification of ground risk events based on the statistics of reliability index β distribution parameters; (61.1) Reliability index calculated for a piece of data β , and its numerical distribution is fitted by normal distribution, lognormal distribution, exponential distribution, Weibull distribution and gamma distribution respectively, and the fitting results are verified by Kolmogorov-Smirnov hypothesis test to obtain the mean and variance statistical parameters of each distribution. The calculation formulas of the above normal distribution, lognormal distribution, exponential distribution, Weibull distribution and gamma distribution are as follows: （3） （4） （5） （6） （7） In the formula, x For the fitted data, formula (3) is normally distributed μ and σ are the mean and standard deviation respectively; formula (4) lognormal distribution and are the mean and standard deviation after logarithmic transformation; Formula (5) is the exponential distribution λ is the rate parameter; Formula (6) Weibull distribution is the scale parameter, k is the shape parameter; Formula (7) Gamma distribution α is the shape parameter, β is the scale parameter, Γ (α) is the gamma function; (61.2) Calculate the probability density histogram and fitted distribution curve of the β value by analyzing the mean and variance statistics obtained from the distribution fitting in step (61.1); (61.3) Then, by using box plots to statistically analyze the discreteness of the β value distribution of different events, we can preliminarily distinguish and differentiate the data of different events; (62) Classification of ground risk events based on k-means clustering; For data of different events, select β The channel with the maximum reliability in the curve is used as the subsequent calculation channel, and wave selection and multi-feature calculation are performed on it; (62.1) Use principal component analysis to reduce the dimensionality of multivariate features, and transform related features into a set of independent principal components through orthogonal transformation, so as to extract the main information of the data and reduce the dimension; (62.2) Use the unsupervised learning algorithm k-means to perform cluster analysis on the reduced-dimensional data; The k-means algorithm divides the data into a specified number of clusters through iterative optimization, with each cluster centered on its centroid, thereby further achieving the classification and identification of different events.

7. A system for evaluating multiple risks of urban roads based on DAS and acoustic emission signal characteristics, used to implement the method for evaluating multiple risks of urban roads based on DAS and acoustic emission signal characteristics as claimed in any one of claims 1 to 6, characterized in that: include: A receiving module receives voiceprint signals collected around existing communication optical fibers on urban roads using DAS; The feature extraction module is used to extract time domain, frequency domain and composite features based on the collected voiceprint signals around the existing communication optical fibers on urban roads, and remove redundant features of the extracted time domain, frequency domain and composite features through the Pearson correlation coefficient matrix to generate optimized time domain, frequency domain and composite features; The abnormal channel identification module is used to judge the abnormality of the area outside the confidence interval using the single feature density distribution and two-dimensional Gaussian distribution for the optimized time domain, frequency domain and composite features, and to locate the abnormal DAS channel by combining the principal component analysis dimensionality reduction and DBSCAN clustering method to identify the hidden dangers of underground cavities; A calculation module is used to calculate the phase cross-correlation function between DAS channels and compare the similarity between DAS channels using the sharpness correlation peak index κ, and finally obtain the reliability index β of each channel; The ground risk event classification module is used to perform distribution statistics on the calculated reliability index β, quantify the risk event characteristics using mean and variance parameters, and combine principal component dimension reduction and K-means clustering methods to achieve ground risk event classification and early warning; The monitoring and evaluation module is used to monitor and evaluate urban road safety risks based on the identification and location of underground cavity hazards and the classification and early warning of ground risk events.

8. The urban road multi-risk assessment system based on DAS and acoustic emission signal characteristics according to claim 7 is characterized by: It also includes a collection module, which is configured as a distributed optical fiber acoustic wave sensing system.

9. A terminal device, comprising a memory, a processor, and a computer program stored in the memory and capable of running on the processor, characterized in that: The memory stores a computer program that can be run on the processor. When the processor loads and executes the computer program, the urban road multi-risk assessment method based on DAS and acoustic emission signal characteristics as described in any one of claims 1 to 6 is adopted.

10. A storage medium containing computer executable instructions, characterized in that: The computer executable instructions are used to execute the urban road multi-risk assessment method based on DAS and acoustic emission signal features as described in any one of claims 1 to 6 when executed by a computer processor.

Images