Recursion method for online automatic identification of seismic event

Abstract

The invention discloses a recursion method for on-line automatic identification of a seismic event, and the method comprises the steps: carrying out the phase-space reconstruction of a time-history signal of seismic oscillation, and carrying out the visual presentation of a reconstructed phase space through a recursion plot; and performing quantitative analysis on the recursion plot information, and determining an alarm threshold value when an earthquake occurs, thereby identifying an earthquake event according to a recursion quantitative index. According to the method, due to the fact that the recursion plot and the improved recursion quantitative index are adopted, the problems that an existing seismic event recognition result is not visual and the recognition timeliness is low are solved, the recursion plot can graphically display the seismic event, and the improved recursion quantitative index can achieve real-time online early warning of the seismic event; the method can be used for real-time monitoring and alarming of the bridge seismic oscillation event, and the intuition and recognition timeliness of the recognition result of the bridge seismic event are improved.

Description

Technical Field

[0001] The invention belongs to the field of bridge health monitoring seismic event recognition and early warning, and relates to an automatic seismic event recognition and early warning method. Background Art

[0002] Real-time online monitoring and early warning of earthquake events are important contents of bridge health monitoring. Existing research methods mainly include the long-short time window ratio method, Akaike information criterion, machine learning and comprehensive analysis algorithms of various methods. Existing methods have problems such as poor real-time processing of earthquake events and low degree of automation. It is difficult to monitor and alarm earthquake events in real time, and it is impossible to make timely decision-making suggestions for bridge maintenance. Summary of the invention

[0003] In view of the deficiencies in the prior art, the purpose of the present invention is to provide a recursive method for online automatic identification of earthquake events, introduce recursive graphs and improved recursive quantitative indicators to solve the problems of low earthquake event identification accuracy, non-intuitive results and low identification timeliness, so as to achieve graphical display and real-time warning of earthquake events.

[0004] In order to solve the above technical problems, the present invention adopts the following technical solutions to achieve the above problems:

[0005] A recursive method for online automatic identification of earthquake events comprises the following steps:

[0006] Step 1, phase space reconstruction: Use the delayed coordinate method to reconstruct the phase space of the earthquake event time series, and determine the optimal embedding dimension and optimal delay time required for phase space reconstruction;

[0007] Step 2, drawing a recursion graph: drawing a recursion graph of the phase space reconstruction result, and determining the parameters required for drawing the recursion graph;

[0008] Step 3, determine the recursive quantitative index: perform quantitative analysis on the recursive graph to obtain the earthquake identification quantitative index;

[0009] Step 4: Issue early warning for earthquake events based on the distribution pattern of earthquake identification quantitative indicators.

[0010] The present invention also includes the following technical features:

[0011] Specifically, in step 1, the phase space of the time series is reconstructed as follows:

[0012] X(i)={x(i),x(i+τ),...,x(i+(m-1)τ)}

[0013] Where X(i) is the time series matrix after phase space reconstruction, i represents the moment; x(i+(m-1)τ) represents the time series at the i+(m-1)τth moment; m is the embedding dimension, and τ is the delay time.

[0014] Specifically, in step 1, the process of determining the optimal embedding dimension includes: determining the embedding dimension m using the false neighbor point method: defining the time series x(t) as a series of phase point vectors in the m-dimensional phase space, solving its false neighbor points, and calculating the ratio of false neighbor points corresponding to different embedding dimensions to determine the optimal embedding dimension. When the ratio of false neighbor points is less than 5% or the false neighbor points no longer decrease with the increase of m, the corresponding dimension is the optimal embedding dimension.

[0015] Specifically, in step 1, the process of determining the optimal delay time includes: determining the delay time τ using the mutual information method: defining the time series x(t) and the time series x(t+τ) of the delay time τ as two sets of discrete sequences, and obtaining the mutual information I(τ) of the two sets of sequences based on the principles of information theory; if I(τ)=0, x(t+τ) is uncorrelated with x(t); the minimum value of I(τ) indicates that x(t+τ) is as uncorrelated as possible with x(t); the first minimum point in I(τ) is the optimal delay time sought.

[0016] Specifically, in step 2, the recursive graph is a graph composed of a 0-1 matrix described by different colors and two time axes, and the distance definition between any two vectors is as follows:

[0017]

[0018] Where: R i,j is a two-dimensional square matrix of order N-(m-1)·τ, i,j represents the i,jth moment, N is the state vector , m is the embedding dimension, τ is the delay time, ε is the cutoff threshold, ||·|| represents the Euclidean norm, Θ(·) represents the Heaviside function, Θ(x≤0)=0; Θ(x>0)=1; In the recursive graph, different R i,j Values ​​are represented in two easily distinguishable colors.

[0019] Specifically, the cutoff threshold ε is selected by a dynamic threshold method, and the selection standard is 20% of the maximum phase space diameter r m ;

[0020] When ε≥r m When , all points are recursive points, and the recursive graph is pure black; when ε<r m , the signal will be divided into recursive points and non-recursive points, and the recursive graph is composed of intersecting black and white lines.

[0021] Specifically, in step 3, the earthquake identification quantitative index T R :

[0022]

[0023] Where: N represents the number of sampling points in the recursive graph; TT is the average vertical line in the recursive graph, and RR is the recursive rate; the value range of TT is [0, N / 2], and the value range of RR is [0, 1].

[0024] Specifically, the TT calculation formula is as follows:

[0025]

[0026] Where N represents the number of sampling points in the recursive graph, v represents the length of the vertical line in the recursive graph, and v min represents the minimum vertical line length, and P(v) is the distribution probability of a vertical line of length ν in the recursive graph.

[0027] Specifically, the RR calculation formula is as follows:

[0028]

[0029] Where N represents the number of sampling points of the recursive graph, i,j represents the i,jth moment, R i,j (ε) is the recursive value.

[0030] Specifically, in step 4, when the earthquake identification quantitative index T R If the earthquake identification quantitative index T is distributed around 0, it is a random signal. R If it approaches 1 and the distribution range is (0.8,1], it is an earthquake signal and an earthquake early warning is carried out.

[0031] Compared with the prior art, the present invention has the following technical effects:

[0032] The present invention uses a recursive graph to visualize the phase space reconstruction result to analyze the characteristics of a dynamic system. The recursive graph can more intuitively observe and understand the nonlinear characteristics of a dynamic system. Earthquake event signals are identified based on an improved recursive quantification index. The index calculation is simple and the amount of calculation is small, and the earthquake event identification speed is fast. The recursive graph has strong adaptability to random signal data containing noise and the index is stable. The improved recursive quantification index can realize real-time online earthquake event identification without any human intervention.

[0033] The method of the present invention is a seismic signal recognition method based on a recursive graph, combines recursive quantitative indicators to provide early warning for earthquakes, and proposes new alarm indicators based on the relationship between earthquake intensity and acceleration division.

[0034] The present invention uses the earthquake time history signals collected by acceleration sensors placed at bridge abutments, anchors, etc. When an earthquake occurs, the time history signals change significantly, and obvious black and white stripes appear in the recursive graph. In contrast, the response of the structure under random environmental excitation is non-stationary, time-varying, and random, showing multi-modal random vibration. These differences are used as distinguishing conditions for monitoring and identifying earthquake events.

[0035] In the face of massive earthquake monitoring data, the present invention can automatically, quickly and accurately identify earthquake signals, meeting the basic requirements of real-time online bridge health monitoring. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] Figure 1 It is a principle framework diagram of the method of the present invention.

[0037] Figure 2 The false neighbor method is used to determine the embedding phase space dimension diagram.

[0038] Figure 3 The delay time map is determined using the mutual information method.

[0039] Figure 4 is the recurrence graph of a random signal.

[0040] Figure 5 It is the recursion diagram of seismic signal.

[0041] Figure 6 It is a comparison chart between the maximum earthquake acceleration amplitude and the recursive quantitative index.

[0042] Figure 7 This is a cloud map of the identification results of 21 earthquakes in the bridge site area from 2020 to 2023.

[0043] Figure 8 It is the seismic multi-channel sensor identification map. DETAILED DESCRIPTION

[0044] The present invention provides a recursive method for online automatic identification of earthquake events. First, the phase space of the earthquake event time series is reconstructed, and the parameters required for the phase space reconstruction are determined; then, a recursive graph is drawn for the phase space reconstruction result, and the parameters required for the recursive graph drawing are determined; then, the recursive quantitative index of the earthquake event is determined according to the characteristics of the recursive graph, and the size of the early warning threshold is determined by comparing the recursive quantitative index with the earthquake signal time course curve; finally, the earthquake event is warned according to the set early warning threshold. Figure 1 , specifically including the following steps:

[0045] Step 1: Phase space reconstruction

[0046] The delayed coordinate method is used to reconstruct the phase space of the chaotic time series and extract more information from the time series. The following formula is the reconstruction of the time series:

[0047] X(i)={x(i),x(i+τ),...,x(i+(m-1)τ)}

[0048] In the above formula, X(i) is the time series matrix after phase space reconstruction, i represents the moment; x(i+(m-1)τ) represents the time series at the i+(m-1)τth moment, m is the embedding dimension, and τ is the delay time. Therefore, the parameters required for the above phase space reconstruction are the embedding dimension m and the delay time τ.

[0049] From the coordinate delay method formula, it can be seen that phase space reconstruction requires two parameters to be determined, one is the delay time τ, and the other is the embedded phase space dimension m. The embedding dimension and time delay must be selected according to the actual situation, otherwise it will affect the quality of phase space reconstruction. The specific selection method is:

[0050] The embedding dimension m is determined by the false neighbor method. In this embodiment, the optimal embedding dimension m=4; the time series x(t) is defined as a series of phase point vectors in the m-dimensional phase space, and its false neighbor points are solved. The false neighbor method is to calculate the proportion of false neighbor points corresponding to different embedding dimensions to determine the optimal embedding dimension. It can be considered that when the proportion of false neighbor points is less than 5% or the false neighbor points no longer decrease with the increase of m, the corresponding dimension is the optimal embedding dimension, see Appendix Figure 2 .

[0051] The mutual information method is used to determine the delay time τ. The optimal delay time τ in this embodiment is 4. For a complex system like a bridge structure, its response evolution must be a nonlinear process, and the mutual information method is used to judge the correlation of nonlinear systems. Define the time series x(t) and the time series x(t+τ) of the delay time τ as two sets of discrete sequences. According to the principle of information theory, the mutual information I(τ) of the two sets of sequences is obtained. On the premise that x(t) is known, if the size of I(τ) is also known, the size of the certainty of x(t+τ) can be calculated. If I(τ) = 0, then x(t+τ) is uncorrelated with x(t). The minimum value of I(τ) indicates that x(t+τ) and x(t) are as uncorrelated as possible. The first minimum point in I(τ) is the optimal delay time τ sought, see attached. Figure 3 .

[0052] Step 2: Draw a recursion diagram

[0053] Phase space reconstruction is a prerequisite for drawing a recursion graph, and the recursion graph is a visual presentation of the reconstructed phase space. After step 1, the optimal embedding dimension and the optimal delay time are determined, and the phase space is reconstructed. The recursion graph is drawn below. The recursion graph is a graph composed of a 0-1 matrix described by different colors and two time axes. The distance between any two vectors is defined as follows:

[0054]

[0055] Where: R i,j is a two-dimensional square matrix of order N-(m-1)·τ, i,j represents the i,jth moment, N is the state vector , m is the embedding dimension, τ is the delay time, ε is the cutoff threshold, ||·|| represents the Euclidean norm, Θ(·) represents the Heaviside function, Θ(x≤0)=0; Θ(x>0)=1.

[0056] In the recurrence diagram, different R i,j Values ​​are represented in two easily distinguishable colors.

[0057] The cut-off threshold ε is an empirical value. Different values ​​will have a great impact on the drawing and quantitative analysis of the recursive graph. The recursive threshold selection method for seismic signal recognition is the dynamic threshold method. The selection standard is 20% of the maximum phase space diameter (r m In order to prevent the interference of mutation noise, for signals with amplitude less than 0.1, the 1% maximum phase space diameter (r m )

[0058] The maximum phase space diameter r m It describes the maximum distance between orbital i and orbital j in a system, that is:

[0059] r m =max|x i -x j |

[0060] When ε≥r m When , all points are recursive points and the recursive graph is pure black. m , the signal will be divided into recursive points and non-recursive points, and the recursive graph is composed of intersecting black and white lines.

[0061] Step 3: Determine the recursive quantitative index

[0062] The recursion diagram is a tool for qualitative analysis of signals. Quantitative analysis of the recursion diagram can quantify the earthquake signal and lay the foundation for the automatic identification of earthquake events. The average vertical line length (Trapping time, TT) was selected for research, and the earthquake identification quantitative index T was obtained by reducing it with the recurrence rate (RR). R :

[0063]

[0064] Where: N represents the number of sampling points of the recursive graph; TT is the average vertical line in the recursive graph, indicating that some states of this signal change slowly or do not change at some moments, that is, stagnant state. This indicator is calculated by all reaching the minimum length ν min The average length of the vertical line is used to estimate the average dead time of the signal in some states. The calculation formula is as follows:

[0065]

[0066] Where N represents the number of sampling points in the recursive graph, v represents the length of the vertical line in the recursive graph, and v min It represents the minimum vertical line length, which is 2. P(v) is the distribution probability of a vertical line with a length of ν in the recursive graph. R i,j is the recursive value, and k is a natural number.

[0067] RR is the recursion rate, which is calculated by calculating the percentage of the black points in the recursion graph to all the recursion points in the entire recursion graph to indicate the degree of aggregation of the trajectory of the phase point in the phase space and the recursion frequency. The calculation formula is as follows:

[0068]

[0069] Where N represents the number of sampling points of the recursive graph, i,j represents the i,jth moment, R i,j (ε) is the recursive value.

[0070] The value range of TT is [0, N / 2], and the value range of RR is [0, 1].

[0071] For random signals, the vertical line length distribution range is narrow and mainly concentrated in the short vertical line area. The average vertical line and the maximum vertical line length are both small, and the recursive rate RR approaches 0.

[0072] For seismic signals, the vertical line length distribution range is wide and the length distribution is uniform. The average vertical line length TT is close to The maximum vertical line value is close to N, and the recursion rate RR approaches 1.

[0073] In summary, the quantitative index T of random signal R It approaches 0 infinitely, while for seismic signals, it approaches 1.

[0074] Step 4: Determine and issue an early warning:

[0075] According to the distribution law of the above quantitative indicators, when a random signal occurs, the overall acceleration signal mutation degree is small, and the quantitative indicator T RDistributed around 0; when an earthquake occurs, the acceleration signal mutation degree intensifies, and the greater the mutation degree, the greater the quantitative index T R The larger the earthquake warning is, the R The range is (0.8,1].

[0076] Specific embodiments of the present invention are given below. It should be noted that the present invention is not limited to the following specific embodiments, and all equivalent changes made on the basis of the technical solution of this application fall within the protection scope of the present invention.

[0077] Example:

[0078] The recognition test was conducted based on the actual monitoring data of multiple earthquakes in the area near the site of a cross-sea bridge from 2020 to 2023, and the accuracy of the recognition method was verified. The statistics of earthquake information during this period are shown in the following table:

[0079] Table 1 Earthquake statistics in the area of ​​a cross-sea bridge from 2020 to 2023

[0080]

[0081]

[0082] Note 1: All the above data except the epicenter distance are from the seismic network; the epicenter distance is an approximate estimate.

[0083] Note 2: Event selection criteria: events within 300 km from the bridge site and with a magnitude of 2.0 or above.

[0084] Note 3: Earthquake intensity estimation formula: I = 4.493 + 1.454M - 1.792Ln (R + 16.0).

[0085] Note 4: The seismic station is a certain distance away from the bridge site. The monitoring data of the strong motion instrument at the bridge site has a certain lag, and the identification time lags behind the seismic network by 1-2 minutes.

[0086] The recognition results of the method of the present invention are as follows: Figure 5 As shown in Figure 2. After the earthquake signal is generated, its time period is fixed and its propagation range is wide, so it can be detected by multiple sensors; while the noise signal in different time periods is random and has a limited propagation range, so it can basically only be detected by a single sensor. Figure 6 As shown, comparative analysis can be performed through multi-channel sensors. When the following multi-channel verification conditions are met, an alarm is triggered, which can further improve the recognition accuracy.

[0087] e=sum([0 / 1] 1×N ) / N

[0088]

[0089] Where: N represents the number of sensors, and the multi-channel verification is triggered only when 3 sensors are met; 0 / 1 represents the earthquake criterion, 1 means that the earthquake signal is recognized; e represents the proportion of multiple sensors triggering alarms; Indicates the multi-sensor calibration result, 1 means the earthquake alarm is finally triggered.

[0090] The present invention combines the recursive graph and the recursive quantitative index to automatically identify special earthquake events. The false neighbor method is used to determine the optimal embedding dimension m, and the mutual information method is used to determine the delay time τ. For the identification of earthquake signals, a new quantitative index T is proposed by combining the average vertical line length and the recursive rate. R The accuracy of this indicator has been verified by measuring earthquakes at the bridge site of a cross-sea bridge from 2020 to 2023. When the multi-channel strong motion detectors are used for joint identification and calibration, the accuracy will be further improved.

[0091] The preferred embodiments of the present invention are described in detail above in conjunction with the accompanying drawings. However, the present invention is not limited to the specific details in the above embodiments. Within the technical concept of the present invention, a variety of simple modifications can be made to the technical solution of the present invention, and these simple modifications all belong to the protection scope of the present invention.

[0092] It should also be noted that the various specific technical features described in the above specific embodiments can be combined in any suitable manner without contradiction. In order to avoid unnecessary repetition, the present invention will not further describe various possible combinations.

[0093] In addition, various embodiments of the present invention may be arbitrarily combined, and as long as they do not violate the concept of the present invention, they should also be regarded as the contents disclosed by the present invention.

Claims

1. A recursive method for online automatic identification of earthquake events, characterized in that: The following steps are involved: Step 1, phase space reconstruction: Use the delayed coordinate method to reconstruct the phase space of the earthquake event time series, and determine the optimal embedding dimension and optimal delay time required for phase space reconstruction; Step 2, drawing a recursion graph: drawing a recursion graph of the phase space reconstruction result, and determining the parameters required for drawing the recursion graph; Step 3, determine the recursive quantitative index: perform quantitative analysis on the recursive graph to obtain the earthquake identification quantitative index; Step 4: Issue early warning for earthquake events based on the distribution pattern of earthquake identification quantitative indicators.

2. The recursive method for online automatic identification of earthquake events according to claim 1, characterized in that: In step 1, the phase space reconstruction of the time series is performed as follows: X(i)={x(i),x(i+τ),...,x(i+(m-1)τ)} Where X(i) is the time series matrix after phase space reconstruction, i represents the moment; x(i+(m-1)τ) represents the time series at the i+(m-1)τth moment; m is the embedding dimension, and τ is the delay time.

3. The recursive method for online automatic identification of earthquake events according to claim 2, characterized in that: In step 1, the process of determining the optimal embedding dimension includes: determining the embedding dimension m using the false neighbor point method: defining the time series x(t) as a series of phase point vectors in the m-dimensional phase space, solving its false neighbor points, and calculating the ratio of false neighbor points corresponding to different embedding dimensions to determine the optimal embedding dimension. When the ratio of false neighbor points is less than 5% or the false neighbor points no longer decrease with the increase of m, the corresponding dimension is the optimal embedding dimension.

4. The recursive method for online automatic identification of earthquake events according to claim 2, characterized in that: In step 1, the process of determining the optimal delay time includes: determining the delay time τ using the mutual information method: defining the time series x(t) and the time series x(t+τ) of the delay time τ as two sets of discrete sequences, and obtaining the mutual information I(τ) of the two sets of sequences based on the principle of information theory; if I(τ)=0, x(t+τ) is uncorrelated with x(t); the minimum value of I(τ) indicates that x(t+τ) is uncorrelated with x(t) as much as possible; the first minimum point in I(τ) is the optimal delay time sought.

5. The recursive method for online automatic identification of earthquake events according to claim 2, characterized in that: In step 2, the recursive graph is a graph composed of a 0-1 matrix described by different colors and two time axes. The distance definition between any two vectors is as follows: Where: R i,j is a two-dimensional square matrix of order N-(m-1)·τ, i,j represents the i,jth moment, N is the state vector , m is the embedding dimension, τ is the delay time, ε is the cutoff threshold, ||·|| represents the Euclidean norm, Θ(·) represents the Heaviside function, Θ(x≤0)=0; Θ(x>0)=1; In the recurrence diagram, different R i,j Values ​​are represented in two easily distinguishable colors.

6. The recursive method for online automatic identification of earthquake events according to claim 5, characterized in that: The method for selecting the cutoff threshold ε is the dynamic threshold method, and the selection standard is 20% of the maximum phase space diameter r m ; When ε≥r m When , all points are recursive points, and the recursive graph is pure black; when ε<r m , the signal will be divided into recursive points and non-recursive points, and the recursive graph is composed of intersecting black and white lines.

7. The recursive method for online automatic identification of earthquake events according to claim 1, characterized in that: In step 3, the earthquake identification quantitative index T R : Where: N represents the number of sampling points in the recursive graph; TT is the average vertical line in the recursive graph, and RR is the recursive rate; the value range of TT is [0, N / 2], and the value range of RR is [0, 1].

8. The recursive method for online automatic identification of earthquake events according to claim 7, characterized in that: The TT calculation formula is as follows: Where N represents the number of sampling points in the recursive graph, v represents the length of the vertical line in the recursive graph, and v min represents the minimum vertical line length, and P(v) is the distribution probability of a vertical line of length ν in the recursive graph.

9. The recursive method for online automatic identification of earthquake events according to claim 7, characterized in that: The RR calculation formula is as follows: Where N represents the number of sampling points of the recursive graph, i,j represents the i,jth moment, R i,j (ε) is the recursive value.

10. The recursive method for online automatic identification of earthquake events according to claim 7, characterized in that: In step 4, when the earthquake identification quantitative index T R If the earthquake identification quantitative index T is distributed around 0, it is a random signal. R If it approaches 1 and the distribution range is (0.8,1], it is an earthquake signal and an earthquake early warning is carried out.

Images