A method for monitoring the direction of ground targets using an unattended single three-component geophone

Abstract

The present invention relates to a method for monitoring the direction of a ground target by an unattended single three-component geophone. In order to solve the problem that the noise in a complex environment is strong and it is difficult to estimate the azimuth of the earthquake source using seismic waves, the present invention extracts relatively ideal Rayleigh waves according to the law of polarization characteristic parameters, and then estimates the azimuth information of the ground earthquake source using the polarization characteristics of the Rayleigh waves. First, the frequency domain is preprocessed according to the time-frequency diagram of the vertical component signal, and then the reciprocal ellipticity in each frequency band is calculated, and a threshold is set to find the frequency band where the Rayleigh wave is most likely to exist. The azimuth is estimated from the three component data obtained after bandpass filtering. The method of the present invention realizes the estimation of the azimuth of the ground earthquake source by a single three-component geophone under low signal-to-noise ratio conditions, and is suitable for field environments with strong noise. After actual data analysis, this research method has obvious advantages over the traditional covariance matrix azimuth estimation method.

Description

Technical field:

[0001] The invention relates to the application field of seismic wave azimuth estimation, and in particular to a method for monitoring the direction of a ground target by an unattended single three-component seismic detector. Background technology:

[0002] In order to solve the practical problems of security protection in key areas such as borders, power plants, oil transmission lines, and highway intersections, it is necessary to monitor static ground sources such as ground explosions and intruders around the clock. Since these targets can cause ground vibrations, unattended distributed passive seismic sensor networks can effectively detect targets. The azimuth indicates the direction of the target relative to the sensor and is an important parameter of the target's location information. By monitoring the target's azimuth, the target's whereabouts can be grasped more quickly, key facilities can be protected, and unnecessary losses can be reduced.

[0003] At present, people mainly use seismic detector arrays to estimate the azimuth of ground targets by obtaining signal arrival time differences for positioning, such as comparative document 1 (CN201910823790.6). This method is generally suitable for strong-amplitude targets at a long distance, because the strong-amplitude target signals at a long distance will be received by multiple seismic sensors. However, the array method of estimating azimuth based on arrival time difference is effective only if all detectors in the array can extract arrival time information. When the vibration signal caused by the target is weak, the sensor network can only receive the signal when the target is close to several or a certain sensor. The detectors farther away in the array cannot extract arrival time information due to the low signal-to-noise ratio, which makes it impossible to estimate the azimuth of the entire array.

[0004] In order to solve the above problems, the direction of the ground vibration target can be estimated by analyzing the polarization characteristics of the seismic wave. In the seismic wave phase, the energy of the Rayleigh wave is strong, and the direction of its particle motion can reflect the azimuth information of the earthquake source. Reference 2 (CN113093093A) successfully estimates the azimuth of the source using the covariance method based on the characteristics of the elliptical motion of the Rayleigh wave. However, the covariance method for obtaining the azimuth is mainly suitable for the situation where the environmental noise is small or the target signal energy is particularly strong. When the environmental noise is large, the interference of seismic noise and Rayleigh wave dispersion will seriously affect the azimuth estimation accuracy, making the covariance method no longer applicable. Therefore, there is an urgent need for a method that can monitor the direction (azimuth) of the ground target using a single three-component detector in a strong noise environment. Summary of the invention:

[0005] The object of the present invention is to provide a method for monitoring the direction of a ground target by using an unattended single three-component geophone in view of the above-mentioned deficiencies in the prior art.

[0006] The inventive concept of the present invention is that in an actual environment with strong noise, seismic signals and noise are overlapped in the time domain, and effective signal detection and separation are considered in the frequency domain. However, the energy of low-frequency interference in the frequency domain may be higher than the signal, and it is unreasonable to only use the frequency domain energy to detect the effective signal. Since the particle motion trajectory of the Rayleigh wave presents an ellipsoidal surface in space, the inverse ellipticity of the polarization parameter can show the characteristics of the ellipsoid, so consider using the inverse ellipticity to detect a relatively ideal Rayleigh wave, and then according to the elliptic equation satisfied between the horizontal component signal and the vertical component signal, deduce the relationship between the three-component signal and the azimuth angle, and then calculate the azimuth angle.

[0007] The objective of the present invention is achieved through the following technical solutions:

[0008] A method for monitoring the direction of a ground target using an unattended single three-component seismic detector comprises the following steps:

[0009] a. Determine the initial frequency range of the useful signal, the original horizontal component signal X, Y and the vertical component signal Z, make a time-frequency diagram of Z, and determine the initial frequency range of the Rayleigh wave based on the energy distribution of the signal and the frequency range of the low-frequency interference [F 1 , F 2 ]Hz, and F 2 It should not be greater than the frequency corresponding to the maximum signal energy in the vertical component time-frequency diagram;

[0010] b. In [F 1 , F 2 ], select the zero phase shift filter to perform multiple bandpass filtering on X, Y and Z, the bandwidth and the moving step are both W Hz, that is, the filtering frequency band range Π(n) is [F 1 , F 1 +W]Hz,[F 1 +W,F 1 +2W]Hz, [F 1 +2W, F 1 +3W]Hz, ..., n is the serial number of the filter band, n = 1, 2, ..., N, N is the total number of filter bands;

[0011] c. Find the reciprocal ellipticity ρ of each frequency band. Taking the Jth filter frequency band in Π(n) as an example, the three components after Π(J) filtering are X J , Y J and Z J , let the matrix ξ=[Z J , X J , Y J ] T , find the 3×3 covariance matrix (ξξ T ) and perform eigenvalue decomposition on it, the eigenvalue λ 1 <λ2 <λ 3 , find ρ(J),

[0012]

[0013] Then find the reciprocal ellipticity ρ(n) in all frequency bands, which contains N elements;

[0014] d. Let the threshold of ρ(n) be L, and find the frequency range corresponding to the part of the ρ(n) discrete curve that is continuously less than L [F a , F b ], in [F a , F b ] to find the minimum point of ρ, the frequency band number corresponding to the minimum point in Π(n) is K. We believe that there is a relatively ideal Rayleigh wave in Π(K), and the horizontal component X after extraction is K , Y K and the vertical component Z K ;

[0015] e. Use calculus to transform X K , Y K and Z K Transform to S x , S y and V z , where S x , S y and V z They are the acceleration signals along the x and y directions and the velocity signal along the z direction;

[0016] f. Estimate the direction (azimuth) of the ground target, the calculation formula is:

[0017]

[0018] Where θ is the angle system and “·” represents the dot product operation between two one-dimensional arrays.

[0019] Beneficial effects:

[0020] The present invention can estimate the azimuth of the ground source using a single three-component detector, which is more suitable for application scenarios with a small number of deployment requirements. It can also be used in urban and outdoor environments with large noise interference. It has strong applicability and can still accurately estimate the azimuth of close-range ground sources in a strong noise environment. The present invention is mainly suitable for estimating the azimuth of static ground sources such as walking people and small explosions, and also has important reference value for dynamic ground sources such as various mobile vehicles. Description of the drawings:

[0021] Figure 1The covariance matrix method and the method of the present invention are used to estimate the absolute error diagram of the azimuth angles of 30 hammer signals.

[0022] The following is a further detailed description of a method for monitoring the direction of a ground target by an unattended single three-component geophone proposed by the present invention in conjunction with the accompanying drawings and embodiments.

[0023] This embodiment was carried out in a stone brick open space, using a velocity type three-component seismic detector arranged on the ground. A small rubber hammer was selected as the ground source, and hammered 30 times continuously at one point.

[0024] A method for monitoring the direction of a ground target using an unattended single three-component seismic detector comprises the following steps:

[0025] a. Determine the initial frequency range of the useful signal, the original horizontal component signal X, Y and the vertical component signal Z, and make a time-frequency diagram of Z. The low-frequency interference is mainly below 20Hz, and the frequency range of the signal energy distribution is [20, 150]Hz. Since the frequency corresponding to the largest part of the signal energy in the vertical component time-frequency diagram is 120Hz, the initial frequency range of the Rayleigh wave is determined to be [20, 120]Hz;

[0026] b. In the range of [20, 120] Hz, a zero-phase shift filter is selected to perform multiple bandpass filtering on X, Y and Z, and the bandwidth and the step size of the movement are both 5 Hz, that is, the filter band range Π(n) is [20, 25] Hz, [25, 30] Hz, [30, 35] Hz, [35, 40] Hz, ..., n is the sequence number of the filter band, n = 1, 2, ..., N, N is the total number of filter bands;

[0027] c. Find the reciprocal ellipticity ρ of each frequency band. Taking the Jth filter frequency band in Π(n) as an example, the three components after Π(J) filtering are X J ,

[0028] Y J and Z J , let the matrix ξ=[Z J , X J , Y J ] T , find the 3×3 covariance matrix (ξξ T ) and perform eigenvalue decomposition on it, the eigenvalue λ 1 <λ 2 <λ 3 , find ρ(J),

[0029]

[0030] Then find the reciprocal ellipticity ρ(n) in all frequency bands, which contains N elements;

[0031] d. Assume that the threshold of ρ(n) is 0.2, find the frequency range [20, 45] Hz corresponding to the part of the discrete curve of ρ(n) that is continuously less than 0.2, and find the minimum point of ρ within [20, 45] Hz. The frequency band number of the minimum point in Π(n) is K. We believe that there is a relatively ideal Rayleigh wave in Π(K). The extracted horizontal component X K , Y K and the vertical component Z K ;

[0032] e. Use calculus to transform X K , Y K and Z K Transform to S x , S y and V z , where S x , S y and V z They are the acceleration signals along the x and y directions and the velocity signal along the z direction;

[0033] f. Estimate the direction (azimuth) of the ground target, the calculation formula is:

[0034]

[0035] Where θ is the angle system and “·” represents the dot product operation between two one-dimensional arrays.

[0036] contrast Figure 1 The absolute error of the azimuth angle of the drum is estimated using the covariance matrix method and the method of the present invention. It can be clearly seen that the error of the covariance matrix method is very large, with an average error of about 50 degrees, which is almost unusable. The azimuth angle estimation accuracy obtained by the method of the present invention has been significantly improved. Under low signal-to-noise ratio conditions, the average absolute error can still be stabilized at 8 degrees, which shows that the method of the present invention has relatively ideal effectiveness and robustness.

Claims

1. A method for monitoring the direction of a ground target using an unattended single three-component seismic detector. The following steps are involved: a. Determine the initial frequency range of the useful signal, the original horizontal component signal X, Y and the vertical component signal Z, make a time-frequency diagram of Z, and determine the initial frequency range of the Rayleigh wave based on the energy distribution of the signal and the frequency range of the low-frequency interference [F 1 , F 2 ]Hz, and F 2 It should not be greater than the frequency corresponding to the maximum signal energy in the vertical component time-frequency diagram; b. In [F 1 , F 2 ], select the zero phase shift filter to perform multiple bandpass filtering on X, Y and Z, the bandwidth and the moving step are both WHz, that is, the filtering band range ∏(n) is [F 1 , F 1 +W]Hz,[F 1 +W,F 1 +2W]Hz, [F 1 +2W, F 1 +3W]Hz, ..., n is the serial number of the filter band, n = 1, 2, ..., N, N is the total number of filter bands; c. Find the reciprocal ellipticity ρ of each frequency band. Taking the Jth filter frequency band in ∏(n) as an example, the three components after ∏(J) filtering are X J , Y J and Z J , let the matrix ξ=[Z J , X J , Y J ] T , find the 3×3 covariance matrix (ξξ T ) and perform eigenvalue decomposition on it, the eigenvalue λ 1 <λ 2 <λ 3 , find ρ(J), Then find the reciprocal ellipticity ρ(n) in all frequency bands, which contains N elements; d. Let the threshold of ρ(n) be L, and find the frequency range corresponding to the part of the ρ(n) discrete curve that is continuously less than L [F a , F b ], in [F a , F b ] to find the minimum point of ρ, the frequency band number of the minimum point corresponding to ∏(n) is K. We believe that there is a relatively ideal Rayleigh wave in ∏(K), and the horizontal component X after extraction is K , Y K and the vertical component Z K ; e. Use calculus to transform X K , Y K and Z K Transform to S x , S y and V z , where S x , S y and V z They are the acceleration signals along the x and y directions and the velocity signal along the z direction; f. Estimate the direction of the ground target, the calculation formula is Where θ is the angle system and "·" represents the dot product operation between two one-dimensional arrays.

Images