A method for locating the source of an intrusion event using distributed optical fiber monitoring

Abstract

The application provides a distributed optical fiber monitoring intrusion event source location method, comprising the following steps: reading a distributed optical fiber modulation demodulation signal, screening acoustic signals of the same intrusion event; picking up P wave first arrivals of each acoustic signal by using multiple known methods; establishing a distance-decay intrusion event source location target function; solving the intrusion event source location target function by using a Bayesian model, picking up sampling by using a Markov chain Monte Carlo method in the solving process, iteratively updating Bayesian model parameters, obtaining adaptively screened P wave first arrivals and source initial location results; adding noise to the adaptively screened P wave first arrivals, and iteratively obtaining source location results with added uncertainty by using the Markov chain Monte Carlo method in combination with the Bayesian model. The source location method provided by the application has high sensitivity, can adaptively screen P wave first arrivals, and reduces the problem of low picking accuracy of a single picking method.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of intrusion event source location, and particularly relates to a distributed optical fiber monitoring intrusion event source location method. BACKGROUND

[0002] Distributed optical fiber acoustic sensing technology, with its high sensitivity and wide coverage, plays an important role in engineering fields such as seismology, urban safety monitoring, and slope engineering. In these applications, acoustic signal detection, classification, and source location of intrusion events are key. Key technologies for source location include P-wave first arrival picking, construction of a source location objective function, and solution methods.

[0003] Currently, P-wave first arrival picking technology has made many advances, such as improved methods such as long-short time window mean ratio, kurtosis, Akaike information criterion, and waveform-based machine learning picking. These methods have achieved positive results in improving the accuracy and efficiency of P-wave first arrival picking. However, there is currently no method that can guarantee absolute accurate picking of P-wave first arrivals in all cases, which may lead to failure in source location. Therefore, it is necessary to develop a source location method that can adaptively screen P-wave first arrivals to improve the accuracy of source location.

[0004] In terms of source location of intrusion events, existing methods often assign the same weight to all P-wave first arrivals. However, signals propagating over long distances are affected by waveform attenuation, and P-wave first arrivals often have greater errors, which means that these data should be given lower weights in source location. To solve this problem, a new source location objective function is needed that can consider the effect of P-wave first arrivals becoming worse as the propagation distance increases.

[0005] At the same time, traditional source location methods often treat a picked P-wave first arrival as a fixed value, ignoring the uncertainty caused by factors such as sampling frequency limitations and picking errors. In fact, a picked P-wave first arrival should be considered as a reference value with a certain uncertainty interval around it. Therefore, developing a source location method that considers the uncertainty of P-wave first arrivals and gives the uncertainty of the source location result is of great significance to determining the reliability of the source location result.

[0006] In summary, to improve the effectiveness of distributed optical fiber acoustic sensing technology in engineering applications, it is crucial to develop a source location method that can adaptively screen P-wave first arrivals, reasonably consider the weights of P-wave first arrivals, and fully consider the uncertainty of P-wave first arrivals. This not only greatly improves the accuracy of source location, but also enhances the practical application value of distributed optical fiber technology in fields such as seismology, urban safety monitoring, and slope engineering. SUMMARY

[0007] In view of the problems in the prior art, a distributed optical fiber monitoring intrusion event source positioning method is provided, which is realized based on multiple P-wave first arrival pickups and Bayesian theory to improve the intrusion event source positioning accuracy and give the uncertainty of the source positioning result.

[0008] The technical scheme adopted by the present application is as follows: a distributed optical fiber monitoring intrusion event source positioning method, comprising:

[0009] Reading the distributed optical fiber modulation and demodulation signal, screening the acoustic signals of the same intrusion event;

[0010] Using multiple known methods to pick up the P-wave first arrival of each acoustic signal;

[0011] Establishing an intrusion event source positioning target function with distance attenuation;

[0012] Establishing a Bayesian model for the intrusion event source positioning target function, using the Markov chain Monte Carlo method to sample the P-wave first arrival of each acoustic signal and the source parameters picked up by each method, iteratively updating the Bayesian model parameters to obtain the adaptively screened P-wave first arrival and the source initial positioning result;

[0013] Adding noise to the adaptively screened P-wave first arrival, and then using the Markov chain Monte Carlo method combined with the Bayesian model to iteratively determine the source positioning result with added uncertainty.

[0014] As a preferred scheme, the reading of the distributed optical fiber modulation and demodulation signal, screening the acoustic signals of the same intrusion event, specifically comprises:

[0015] Using the long-short time window mean ratio method to detect the intrusion event acoustic signal, and then using the adjacent channel cross-correlation function to screen the same event acoustic signal.

[0016] As a preferred scheme, the using of the adjacent channel cross-correlation function to screen the same event acoustic signal, specifically comprises:

[0017] Calculating the maximum value of the cross-correlation function, wherein the cross-correlation function calculation formula is:

[0018]

[0019] wherein, is the cross-correlation function value of the signal x and the signal y , and are the time series of the signal x and the signal y , respectively;

[0020] If the maximum difference of the cross-correlation function of adjacent signals is greater than a preset threshold, it means that the two acoustic signals do not belong to the same intrusion event; otherwise, they belong to the same intrusion event.

[0021] As a preferred solution, the method of picking up the first arrival of each acoustic signal P wave by using a variety of known methods specifically includes:

[0022] The P-wave first arrival was picked using manual picking method, STA / LTA method, high-order statistics method, Akaike information criterion method, cross-correlation method, fractal dimension method and artificial neural network method.

[0023] As a preferred solution, the establishment of an intrusion event source location objective function that attenuates with distance specifically includes:

[0024] Calculate the straight-line distance between the current earthquake source and each sensor and the P-wave propagation time;

[0025] Calculate the P-wave first arrival coefficient of each sensor;

[0026] Calculate the objective function of earthquake source location for intrusion events that decays exponentially with distance.

[0027] As a preferred solution, the calculation method of the P wave first arrival coefficient of each sensor is:

[0028]

[0029] in, For the k The P-wave first arrival coefficient of each sensor, is the number of sensors, is the variance of the propagation distance, The current earthquake source and the k The straight line distance between the sensors.

[0030] As a preferred solution, the intrusion event source location objective function that decays exponentially with distance is specifically:

[0031]

[0032] in, is the objective function for locating the source of the intrusion event, x 0 、y 0. z 0 is the earthquake source x 、 y 、 z The coordinates of the direction, t 0 is the time when the earthquake source occurs, For the k The acoustic signal is given by i The first arrival of the P wave determined by the picking method, P-wave propagation time corresponding to the kth sensor.

[0033] As a preferred solution, the Markov Chain Monte Carlo method is used to sample each acoustic signal P-wave first arrival and source parameter picked up by each method in the solving process, and the iterative updating of the Bayesian model parameters specifically includes:

[0034] The model parameters are numbered, wherein the model parameters include each acoustic signal P-wave first arrival and source parameter picked up by each method;

[0035] An integer in the numbering range is randomly generated, if the integer is within the acoustic signal number range, a P-wave first arrival is randomly selected from the acoustic signal pickup data corresponding to the integer for updating; if the integer is greater than the acoustic signal number, a pre-set updating step and a Gaussian distribution random number satisfying the mean value of 0 and the standard deviation of 1 are used to update the corresponding source parameter.

[0036] As a preferred solution, the adaptive screened P-wave first arrival and source initial positioning result are obtained, specifically including:

[0037] Based on the updated model parameters, the Bayesian model maximum likelihood function before and after the model parameter updating is calculated, and then the receiving probability of the Bayesian model before and after the model parameter updating is calculated;

[0038] If the receiving probability is greater than or equal to a random number satisfying the 0-1 uniform distribution, it means that the current iteration of the Bayesian model parameter updating is successful, otherwise the Bayesian model parameters before updating are kept;

[0039] The sampling iteration process is repeated, the last acoustic signal P-wave first arrival selected in the iteration process is taken as the adaptive screened P-wave first arrival, and the mean value of the last N source parameters updated in the iteration process is taken as the source initial positioning result.

[0040] As a preferred solution, the noise is added to the adaptive screened P-wave first arrival, specifically including:

[0041] The screened P-wave first arrival is taken as the mean value, and noise with a standard deviation of is added to represent the uncertainty of the P-wave first arrival, wherein a is a coefficient related to the sensor response, is the straight-line distance between the current source and the kth sensor. k

[0042] ​Compared with the prior art, the beneficial effects of the above technical solutions are: the application uses adjacent channel cross-correlation coefficients to screen the same invasion event signal, and has high sensitivity; the established invasion event source positioning target function with distance attenuation can reduce the influence of the usually poor quality of the far distance P wave first arrival; the developed multiple P wave first arrival MCMC sampling solving source initial positioning method can adaptively screen the P wave first arrival, and reduce the problem of low picking accuracy of a single picking method; and the proposed source Bayesian positioning method considering the uncertainty of the P wave first arrival can give the uncertainty of the source positioning result, and has good guiding significance for the reliability analysis of the source positioning result. BRIEF DESCRIPTION OF DRAWINGS

[0043] Figure 1 is a distributed optical fiber monitoring invasion event source positioning method flow chart proposed by the application.

[0044] Figure 2 is an example of acoustic signals monitored by DAS.

[0045] Figure 3 is a sequence of adjacent acoustic signal cross-correlation maximum values.

[0046] Figure 4 is the position of the optical fiber sensor and the invasion event source.

[0047] Figure 5 is the theoretical propagation time of the invasion event and the P wave first arrival picking time of different methods.

[0048] Figure 6 is the positioning weight of each sensor data at the last iteration of MCMC.

[0049] Figure 7 is a sequence of MCMC iteration maximum likelihood functions of multiple picking source initial positioning.

[0050] Figure 8 is a sequence of MCMC iteration source positions of multiple picking source initial positioning.

[0051] Figure 9 is the adaptive optimal time of MCMC iteration of source initial positioning.

[0052] Figure 10 is a P wave first arrival noise data distribution diagram.

[0053] Figure 11 is a sequence of MCMC iteration maximum likelihood functions of noise data source positioning.

[0054] Figure 12 is a sequence of MCMC iteration source positions of noise data source positioning.

[0055] Fig. 13(a), Fig. 13(b), Fig. 13(c) are the uncertainty of the source location result of the intrusion event in the xy, xz, yz directions respectively. DETAILED DESCRIPTION

[0056] Embodiments of the present application are described in detail below with reference to examples illustrated in the accompanying drawings, in which the same or similar components or components having the same or similar functions are denoted by the same or similar reference numerals throughout. The embodiments described below with reference to the accompanying drawings are exemplary and are intended only for explanation of the present application, and cannot be understood as limiting the present application. On the contrary, the embodiments of the present application include all changes, modifications and equivalents falling within the spirit and scope of the appended claims.

[0057] Since the conventional P-wave first arrival picking method cannot ensure the accuracy of the P-wave first arrival, the source location does not consider the uncertainty of the P-wave first arrival, etc. The embodiments of the present application propose a distributed optical fiber monitoring intrusion event source location method based on multiple P-wave first arrival picking and Bayesian theory. The method integrates the same intrusion event screening, source location target function optimization, multiple method P-wave first arrival picking data adaptive screening, considering the P-wave first arrival uncertainty, etc. to improve the accuracy of the source location of the intrusion event, and gives the uncertainty of the source location result.

[0058] Referring to Figure 1 , the distributed optical fiber monitoring intrusion event source location method comprises the following steps:

[0059] Step 1, screen the acoustic signals belonging to the same intrusion event.

[0060] Through the screening of the acoustic signals, the signals not belonging to the intrusion event can be effectively excluded to improve the positioning accuracy. In an embodiment, the adjacent channel cross-correlation function is used for screening, and in other embodiments, the adjacent channel P-wave first arrival time difference can also be used to screen the acoustic signals of the same intrusion event. In this embodiment, the adjacent channel cross-correlation function is used for illustration, and the specific process is as follows:

[0061] For the obtained distributed optical fiber modulation demodulation signal, the STA / LTA method is used to detect the intrusion event acoustic signal, and then the adjacent channel cross-correlation function is used to screen the acoustic signal of the same intrusion event. The cross-correlation function calculation expression is:

[0062]

[0063] Wherein, is the cross-correlation function value of signal x and signal y, and The time series of signal x and signal y respectively. At this time, the cross-correlation function of signal 1 and 2, signal 3 and 4, …, etc. is calculated according to the sensor number in the system. When the maximum value of the cross-correlation function of the adjacent on-off is greater than 5%, it is considered that the two acoustic signals do not belong to the same intrusion event; otherwise, it is considered that they belong to the same intrusion event, so as to screen out the acoustic signals belonging to the same intrusion event.

[0064] Step 2, pick up the P-wave first arrival of each acoustic signal by using various known methods.

[0065] Since the traditional P-wave first arrival picking method cannot guarantee absolute accurate picking in all cases, it may lead to failure of the source location, therefore, in the embodiment, various methods are used to pick up the P-wave first arrival of each acoustic signal, and adaptive screening is carried out in the subsequent process. Preferably, the methods used in the embodiment include artificial picking method, STA / LTA method, high-order statistics (PAI-K) method, Akaike information criterion (AIC) method, cross-correlation method, fractal dimension method and artificial neural network method. For the P-wave first arrival of the i-th acoustic signal determined by the j-th method, it is denoted as k i , K and are the number of acoustic signals and the picking method number respectively.

[0066] Step 3, establish an intrusion event source location target function with distance attenuation.

[0067] Since the signal transmitted at a long distance is affected by waveform attenuation image, the P-wave first arrival often has a large error, therefore, these data should be given a lower weight. In the embodiment, an intrusion event source location ray travel time target function with exponential attenuation of propagation distance is established to realize adaptive weighting of P-wave first arrival. The target function is calculated as follows:

[0068] Suppose the current source position coordinates are , the i-th sensor coordinates are k , , k =1, 2, …, K , K is the number of sensors, then the straight-line segment distance between the current source and the i-th sensor is: k

[0069]

[0070] The P-wave propagation time is , wherein is the P-wave propagation velocity, and the variance of the propagation distance is , wherein ​​​is the average distance between all sensors and the current earthquake source, then k The P wave first arrival coefficient of each sensor is

[0071]

[0072] in, is the P wave first arrival coefficient of the kth sensor, is the number of sensors, which is consistent with the number of acoustic signals obtained, is the variance of the propagation distance, is the straight line distance between the current earthquake source and the kth sensor.

[0073] Thus, the objective function of earthquake source location for intrusion events that decays exponentially with distance is obtained:

[0074] .

[0075] in, is the objective function for locating the source of the intrusion event, x 0 、y 0. z 0 are the x, y, and z coordinates of the earthquake source, t 0 is the time when the earthquake source occurs, For the k The acoustic signal is given by i The first arrival of the P wave determined by the picking method, For the k The P-wave propagation time corresponding to each sensor.

[0076] It should be noted that this embodiment only takes the establishment of an intrusion event source location objective function that decays exponentially with distance as an example. In other embodiments, the objective function may also be established in the form of power law decay, logarithmic decay, etc.

[0077] Step 4: Use MCMC to sample the P-wave first arrival data from various methods to solve the initial earthquake source location.

[0078] A Bayesian model is established to solve the objective function of earthquake source location for intrusion events. The maximum likelihood function is calculated as follows:

[0079]

[0080] in, for The matrix formed, for The matrix formed, k =1,2,…, K ; T is the matrix transpose symbol; for The covariance matrix formed; refer to The determinant of Refers to the current model; Refers to the updated model; No. k The elements are .

[0081] In this embodiment, Markov chain Monte Carlo (MCMC) is used to analyze each model parameter ( 、 x 0, y 0, z 0, t 0) Sampling, sampling only one parameter at a time, and solving the objective function through sampling data and Bayesian model. The specific process is as follows:

[0082] Set the source parameters ( x 0, y 0, z 0, t 0) initial value, the model parameters ( 、 x 0, y 0, z 0, t 0) are numbered 1, 2, ..., K , K +1, K +2, K +3, K +4; where K corresponds to K acoustic signals; K+1~K+4 correspond to 4 source parameters;

[0083] Randomly generate [1, K +4], if the integer belongs to [1, K ], then randomly select a data from the acoustic signal picked up by the integer to update; if the number belongs to [ K +1, K+4 ], then use Update the corresponding source parameters, where p 0. mv and g are the current value of the source parameter, the update step size, and a Gaussian distribution random number with a mean of 0 and a standard deviation of 1. During the update process, one model parameter is updated each time.

[0084] In one embodiment, xy The update step size of the direction is 0.5, z The update step size of the direction is 0.001, t The update step size of 0 is 0.0005.

[0085] On this basis, the maximum likelihood function after updating the parameters is calculated , and calculate the reception probability based on the maximum likelihood function before and after the update , and a random number that satisfies a uniform distribution of 0 to 1 u For comparison, if , then the model parameters of this iteration ( 、 x 0, y 0, z 0, t 0) Update is successful, otherwise the model parameters before the update are kept. Repeat the sampling iteration process, and the number of iterations is set according to the requirements. The last value obtained in the iterative sequence is the first arrival of the P wave after adaptive screening, and the last 5000 values ​​updated in the iterative process are x 0, y 0, z The mean of the 0 values ​​is used as the initial source location result. In other embodiments, a grid search method can also be used to screen the P wave first arrival.

[0086] Based on the initial earthquake source location results, this embodiment also proposes a Bayesian earthquake source location method that takes into account the uncertainty of the first arrival of the P wave, namely:

[0087] Step 5: Bayesian earthquake source location considering the uncertainty of the first arrival of the P wave.

[0088] The first arrival of the adaptively screened P wave is taken as the mean, and the standard deviation is added on this basis. The noise is used to characterize the uncertainty of the P-wave first arrival. a is the coefficient related to the sensor response. Then, using the MCMC sampling method in step 4, continue to iteratively calculate using the Baye model x 0, y 0, z 0 to complete the earthquake source location. It should be noted that in each iteration, each P wave first arrival is based on the above noise data (that is, based on the mean of the P wave first arrival, with a mean of 0 and a standard deviation of The data after noise is randomly sampled. x 0, y 0, z The mean of the last 5000 values ​​is used as the earthquake source location result, and the x 0, y 0, z The last 5000 values ​​of 0 are plotted as density clouds to represent the uncertainty of the earthquake source location results.

[0089] In this embodiment, the Bayesian positioning method is still used to complete the source positioning after adding noise. In other embodiments, the optimization method can also be used to position each P-wave first arrival sample, and multiple first positioning results are used to represent the uncertainty of the source positioning result.

[0090] The following results Figure 2~Figure 1 3. The distributed optical fiber monitoring intrusion event source positioning method proposed in the present application is verified.

[0091] Figure 2 is an example of acoustic signals monitored by DAS. As can be seen from the figure, the signal-to-noise ratio of the DAS signal is usually low, and the 593~599, 600~612, and 613~620 sensors have similar waveforms and can be considered as different intrusion events. Intrusion event 1 has a clear similar waveform segment (signal in the dashed box), but its P-wave first arrival is not obvious, and it is difficult to use the P-wave first arrival time difference to screen the acoustic signals of the same intrusion event; Intrusion event 2 has a relatively obvious P-wave first arrival and similar waveform characteristics, which provides convenience for both P-wave first arrival time difference and waveform cross-correlation function to screen the acoustic signals of the same event; Intrusion event 2 is mixed in the tail of intrusion event 3, which increases the difficulty of screening the acoustic signals of the same intrusion event.

[0092] Figure 3 is a sequence of adjacent acoustic signal cross-correlation maximum values. The cross-correlation maximum value of signal pair No. i corresponds to the maximum value of the cross-correlation sequence of the i and i +1 acoustic signals. As can be seen from the figure, the cross-correlation values of the 599th and 600th, 613th and 614th acoustic signals are obvious local minimum values, and the cross-correlation maximum value decreases by more than 5%, so the 593~599, 600~612, and 613~620 acoustic signals are divided into three types of intrusion events, which is consistent with the manual analysis of Figure 2 , verifying the effectiveness of the adjacent channel cross-correlation function in screening the acoustic signals of the same event.

[0093] Figure 4 is the optical fiber sensor and the source position of the intrusion event. The coordinates of each number of the optical fiber sensor are shown in Table 1, and the coordinates of the intrusion event are Ex =327955.64, Ey =4407652.27, Ez =1228.00, and the P-wave propagation speed is 3200 m / s. As can be seen from the figure, the middle number sensor is close to the position of the intrusion event, that is, when the sensor number expands to both sides, the P-wave propagation time shows an increasing trend.

[0094] Table 1 Coordinates of optical fiber sensors

[0095]

[0096] Figure 5 is the theoretical propagation time of the invasion event and the P-wave first arrival picked by different methods. The theoretical propagation time curve in the figure is obtained by dividing the straight-line distance between the invasion event location and each sensor by the P-wave propagation velocity, and then the P-wave first arrival is picked by using the long-short time window average ratio method (STA / LTA), the Akaike information criterion method (AIC) and the kurtosis method (PAI-K). For easy comparison, the P-wave first arrival picking time is subtracted by a base. As can be seen from the figure, the STA / LTA method has better P-wave first arrival picking effect when the acoustic signal propagation distance is relatively short, and the AIC method and the PAI-K method may have larger errors due to picking stability problems when the acoustic signal propagation distance is relatively short; the three picking methods have larger P-wave first arrival picking errors when the acoustic signal propagation distance is relatively long.

[0097] Figure 6 is the positioning weight of each sensor data at the last iteration of MCMC. As can be seen from the figure, the sensor data close to the invasion event location has obviously larger positioning weight; and for the sensor data with a long propagation distance, the positioning weight tends to zero, which is adapted to the smaller weight of the source location when the data quality of the sensor with a long propagation distance is poor in Figure 5 .

[0098] Figure 7 is the maximum likelihood function sequence of MCMC iteration of the multiple picked source initial location. As can be seen from the figure, the maximum likelihood function increases rapidly at the beginning of iteration, and tends to be stable when the iteration is about 5000 times, indicating the effectiveness of the established Bayesian model. The maximum likelihood function sequence oscillates in the whole iteration process, which is due to the fact that a random number of 0-1 is selected in each iteration of the Bayesian model, that is, the poor model after parameter updating may also be received, which provides convenience for finding the global optimal source location result.

[0099] Figure 8 is the source location sequence of MCMC iteration of the multiple picked source initial location. As can be seen from the figure, the source location changes greatly at the beginning of iteration, x and y coordinates, while the z coordinate changes little in the whole iteration process, which is due to the fact that the sensor network has good distribution in the xy direction and poor distribution in the z direction. The source location tends to be stable when the MCMC iteration is about 1000 times, and the initial location result is (327955.58, 4407652.30, 1228.49) m, and the location error is 0.49 m.

[0100] Figure 9is the adaptive optimal time of the initial location of the source by MCMC iteration. It can be seen from the figure that for sensors with a relatively short propagation distance, the P-wave first arrival time after adaptive optimization is in good agreement with the theoretical propagation time; for sensors with a relatively long propagation distance, adaptive screening can remove P-wave first arrival picking data with large errors, but it may still screen P-wave first arrival picking data with large errors. This is because, during source location, the data of the remote sensor has a weight that tends to zero. The P-wave first arrival picking data with large errors obtained by adaptive screening has little effect on the source location, indicating the effectiveness of the P-wave first arrival adaptive screening of the patent.

[0101] Figure 10 is a P-wave first arrival noise data distribution diagram. A Gaussian noise is added to the distance between the initial location result and the sensor and the response characteristics of the sensor, with the screened P-wave first arrival as the mean, to obtain Figure 10 . It can be seen from the figure that the 20000 added Gaussian noises have good Gaussian distribution, and the noises added to the data of sensors with a larger propagation distance are generally larger.

[0102] Figure 11 is the MCMC iteration maximum likelihood function sequence of the source location of the noise data. Compared with the MCMC iteration maximum likelihood function sequence of the source location of the multiple picked data Figure 7 , the maximum likelihood function sequence with the initial location result as the input Figure 11 has a faster overall convergence speed, but its iteration sequence increases overall during the entire iteration process, which is due to the addition of Gaussian noise to the P-wave travel time during each iteration.

[0103] Figure 12 is the MCMC iteration source location sequence of the noise data source location. It can be seen from the figure that the MCMC iteration source location sequence of the noise data source location tends to be stable after about 2000 iterations, which is due to the good initial value of the iteration location, but the stability of the MCMC iteration source location sequence is slightly worse than that of the MCMC iteration source location sequence of the multiple picked data source location, which is due to the addition of Gaussian noise to the P-wave travel time during each iteration. The source location result is (327954.89, 4407652.80, 1228.34) m, and the location error is 0.98 m, which is slightly larger than the initial location error, which is due to the decrease in the convergence of the Bayesian model after adding Gaussian noise. The location error of the intrusion event is within 1 m, and the location effect is good.

[0104] Fig. 13(a), Fig. 13(b), Fig. 13(c) are the uncertainty of the source location of the intrusion event in xy, xz, yz directions respectively. As shown in the figures, the uncertainty of the source location in xy direction is circularly distributed, and the uncertainty in z direction is elliptically distributed, which is due to the good distribution of the sensor network in xy direction and poor distribution in z direction. The distribution of the uncertainty of the intrusion event location is relatively concentrated, indicating that the location result has good reliability.

[0105] In particular, the processes described above with reference to the flowcharts can be implemented as computer software programs in accordance with embodiments of the present application. For example, embodiments of the present application include a computer program product comprising a computer program carried on a computer readable medium, the computer program comprising program code for executing the methods illustrated by the flowcharts.

[0106] It should be noted that the computer readable medium shown in the embodiments of the present application can be a computer readable signal medium or a computer readable storage medium or any combination of the two. The computer readable storage medium may, for example, but is not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or apparatus, or any combination of the above. More specific examples of the computer readable storage medium can include, but are not limited to, an electrical connection having one or more wires, a portable computer diskette, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM), a flash memory, an optical fiber, a portable compact disk read-only memory (Compact Disc Read-Only Memory, CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In the present application, the computer readable storage medium can be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, device or apparatus. In the present application, the computer readable signal medium can include a data signal carried in a baseband or as a part of a carrier wave, which carries the computer readable program code. Such a propagated data signal can take many forms, including but not limited to an electromagnetic signal, an optical signal or any suitable combination of the above. The computer readable signal medium can also be any computer readable medium other than the computer readable storage medium, which can send, propagate or transmit the program for use by or in conjunction with an instruction execution system, device or apparatus. The program code contained on the computer readable medium can be transmitted by any suitable medium, including but not limited to wireless, wired, or the like, or any suitable combination of the above.

[0107] The flow and block diagrams in the drawings represent possible architectural, functional, and operational architectures of systems, methods, and computer program products according to various embodiments of the present application. Each block in the flow and block diagrams can represent a module, a segment, or a portion of code, which comprises one or more executable instructions for implementing the specified logical function(s). It should also be noted that in some alternative implementations, the functions noted in the blocks can occur out of the order noted in the figures. For example, two blocks shown in succession may, in fact, be executed substantially concurrently or the blocks may sometimes be executed in the reverse order, depending upon the functionality involved. It will also be noted that each block of the block diagrams and / or flowchart illustrations, and combinations thereof, can be implemented by a dedicated hardware-based system that performs the specified functions or operations, or combinations of hardware and software.

[0108] The units involved in the embodiments of the present application can be implemented by software, or can be implemented by hardware, and the described units can also be arranged in a processor. In some cases, the names of the units do not constitute a limitation on the units themselves.

[0109] As another aspect, the present application also provides a computer program product or computer program, which includes computer instructions stored in a computer readable storage medium. The processor of the computer device reads the computer instructions from the computer readable storage medium, and the processor executes the computer instructions, so that the computer device executes the distributed optical fiber monitoring intrusion event source positioning method described in the above embodiments.

[0110] As another aspect, the present application also provides a computer readable medium, which can be included in the electronic device described in the above embodiments, or can exist separately without being assembled into the electronic device. The computer readable medium carries one or more programs, which, when executed by the electronic device, enable the electronic device to implement the distributed optical fiber monitoring intrusion event source positioning method described in the above embodiments.

[0111] It should be noted that although several modules or units of the device for action execution are mentioned in the above detailed description, such division is not mandatory. In fact, according to the embodiments of the present application, the features and functions of two or more modules or units described above can be embodied in one module or unit. Conversely, the features and functions of one module or unit described above can be further divided into multiple modules or units.

[0112] Through the above description of the embodiments, those skilled in the art can easily understand that the example embodiments described herein can be implemented by software, or by software in combination with necessary hardware. Therefore, the technical solutions according to the embodiments of the present application can be embodied in the form of a software product. The software product can be stored in a non-volatile storage medium (which can be a CD-ROM, a U disk, a mobile hard disk, etc.) or on a network, and includes a plurality of instructions to enable a computing device (which can be a personal computer, a server, a touch terminal, or a network device, etc.) to perform the method according to the embodiments of the present application.

[0113] For those skilled in the art, the specific meanings of the above terms in the present application can be understood according to the specific circumstances; the drawings in the embodiments are used to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, not all the embodiments. The components of the embodiments of the present application described and shown in the drawings herein can be arranged and designed in various different configurations.

[0114] Although the embodiments of the present application have been shown and described above, it can be understood that the above embodiments are exemplary and cannot be understood as limiting the present application. Those skilled in the art can make changes, modifications, replacements and variations to the above embodiments within the scope of the present application.

Claims

1. A method for distributed optical fiber monitoring of an intrusion event hypocenter location, the method comprising: The application relates to an adaptive method for picking up P-wave first arrivals of acoustic signals of an intrusion event. ​ The method comprises the following steps: reading distributed fiber modulation demodulation signals, screening acoustic signals of the same intrusion event; adopting multiple known methods to pick up P-wave first arrivals of each acoustic signal; establishing a distance-decay intrusion event source positioning target function; solving the distance-decay intrusion event source positioning target function by establishing a Bayesian model, in the solving process, the Markov Chain Monte Carlo method is used to sample each P-wave first arrival of each acoustic signal picked up by each method and source parameters, and the Bayesian model parameters are iteratively updated to obtain adaptively screened P-wave first arrivals and source preliminary positioning results; The noise is added in the P-wave first arrival after adaptive screening, including: taking the screened P-wave first arrival as mean value, adding noise with mean value of 0 and standard deviation of to represent the uncertainty of the P-wave first arrival, wherein, a is a coefficient related to the sensor response, is the straight-line segment distance between the current seismic source and the first k sensor. adding noise to the adaptively screened P-wave first arrivals, and iteratively obtaining source positioning results with added uncertainty by using the Markov Chain Monte Carlo method in combination with the Bayesian model; the distance-decay intrusion event source positioning target function is established, and the establishment specifically comprises the following steps: calculating the linear segment distance of the current source from each sensor and the P-wave propagation time; calculating the P-wave first arrival coefficient of each sensor; calculating the distance-decay intrusion event source positioning target function; in, For the k The P-wave first arrival coefficient of each sensor, is the number of sensors, is the variance of the propagation distance, The current earthquake source and the k The straight line distance between the sensors; the calculation method of the P-wave first arrival coefficient of each sensor is as follows: in the solving process, the Markov Chain Monte Carlo method is used to sample each P-wave first arrival of each acoustic signal picked up by each method and source parameters, and the Bayesian model parameters are iteratively updated, and the specific steps comprise the following steps: numbering model parameters, wherein the model parameters include P-wave first arrivals of all acoustic signals and source parameters; randomly generating an integer within the numbering range, if the integer is within the acoustic signal number range, randomly selecting a P-wave first arrival in the acoustic signal picked up data corresponding to the integer for updating, if the integer is greater than the acoustic signal number, using a pre-set updating step and a Gaussian distribution random number satisfying the mean value of 0 and the standard deviation of 1 to update the corresponding source parameters; the adaptively screened P-wave first arrivals and source preliminary positioning results are obtained, and the obtaining specifically comprises the following steps: based on the updated model parameters, calculating the Bayesian model maximum likelihood function before and after the model parameter updating, and then calculating the receiving probability of the Bayesian model before and after the model parameter updating; The sampling iteration process is repeated, the last P-wave first arrival selected in the iteration process is taken as the P-wave first arrival after adaptive screening, and the last mean of the source parameters updated in the iteration process is taken as the source initial result. N ​ 2. The method of claim 1, wherein, if the receiving probability is greater than or equal to a random number satisfying the 0-1 uniform distribution, it indicates that the current iteration of the Bayesian model parameter updating is successful, otherwise the Bayesian model parameters before the updating are kept. the reading of the distributed fiber modulation demodulation signals and the screening of the acoustic signals of the same intrusion event specifically comprise the following steps:

3. The method of claim 2, wherein, using a long-short time window mean ratio method to detect intrusion event acoustic signals, and then using an adjacent channel cross-correlation function to screen the same event acoustic signals. the screening of the same event acoustic signals by using the adjacent channel cross-correlation function specifically comprises the following steps: wherein is a signal x and a signal y cross-correlation function value, and are time series of a signal x and a signal y respectively; calculating a cross-correlation function maximum value, wherein the cross-correlation function calculation formula is as follows:

4. The method of claim 1 or 2, wherein, if the difference between the adjacent signal cross-correlation function maximum values is greater than a pre-set threshold, it indicates that the two acoustic signals do not belong to the same intrusion event; otherwise, it indicates that the two acoustic signals belong to the same intrusion event. the picking up of P-wave first arrivals of each acoustic signal by using multiple known methods specifically comprises the following steps: using an artificial picking method, an STA / LTA method, a high-order statistic method, an AIC method, a cross-correlation method, a fractal dimension method and an artificial neural network method to pick up P-wave first arrivals.

5. The method of claim 1, wherein, The invasion event source positioning target function with the exponential decay with distance is specifically: wherein, is an objective function for locating the source of an intrusion event, x 0 、y 0, z 0 are the coordinates of the source x , y , z in the direction of the source, t 0 is the time of occurrence of the source, is the P-wave first arrival determined by the k th acoustic signal using the i th picking method, is the P-wave travel time corresponding to the k th sensor.

Images