Method for estimating sound velocity and thickness of ice layer by using airwaves and ice layer bending waves

Abstract

The invention belongs to the field of polar acoustic measurement, and relates to a method for estimating the sound velocity and thickness of an ice layer by using airwaves and ice layer bending waves, which comprises the following steps of: acquiring broadband pulse sound signals by using a sensor fixed on an ice surface; recording the distance between the detonating device and the ice surface sensor; performing time-frequency analysis on the broadband pulse signal acquired by the ice surface sensor; obtaining a time-frequency spectrum of the signal through time-frequency analysis, calculating the propagation velocity of the airwave, converting the time-frequency spectrum into a group velocity spectrum, and extracting a group velocity dispersion curve of the bending wave; initializing the longitudinal wave velocity, the transverse wave velocity, the ice layer thickness and the seawater sound velocity of the ice layer; calculating an expected group velocity curve according to the sound propagation model, and constructing a cost function; and optimizing the cost function by adopting a particle swarm algorithm, and outputting a final parameter estimation value. The method can be realized only by using a single sensor, and the synchronization of transmitting equipment and receiving equipment is not required; the sound velocity and thickness of the ice layer can be inverted by using the relationship between the group velocity and the acoustic parameters.

Description

Technical Field

[0001] The invention belongs to the field of polar acoustic measurement and relates to a method for estimating the sound velocity and thickness of an ice layer by using air waves and ice layer bending waves. Background Art

[0002] Existing studies have shown that the impact of global climate change is strongest in the Arctic, which has now become the region with the strongest greenhouse effect on Earth. A typical feature is the accelerated decline in sea ice coverage, both in terms of ice extent and average thickness, which is declining faster than predicted by climate models. Therefore, a more detailed description of the physical processes involved in sea ice models is needed, which involves the accurate acquisition of parameters such as ice thickness, sound speed, salinity, and temperature.

[0003] Sea ice is a seismic elastic waveguide, and the elastic wave field consists of Lamb wave modes and horizontally polarized shear wave (SH) modes. Lamb waves are decomposed into two groups: antisymmetric modes An and symmetric modes Sn. n = 0 represents the fundamental mode, and n = 1, 2, 3 represents the higher-order modes. It will only propagate at frequencies higher than its cutoff frequency. The antisymmetric mode produces bending motion, and the symmetric mode produces traction-compression motion. However, the solid-liquid interface causes changes in boundary conditions. The main difference is that an additional Scholte wave similar to that propagating on a semi-infinite liquid-solid interface is generated, which behaves similarly to the A0 mode (bending wave) in the low frequency band and has no dispersion effect in the high frequency band.

[0004] The propagation of seismic waves in sea ice can be used to monitor sea ice properties, such as sea ice thickness, sound speed, Young's modulus, Poisson's ratio and other parameters. Traditional seismic methods usually require dozens or more seismic detectors to form an array, but the harsh polar environment limits the deployment of some large instruments and equipment. Using more intelligent, easy-to-operate miniaturized equipment to monitor sea ice properties is a promising method. Summary of the invention

[0005] In view of the above existing research and technology, the technical problem to be solved by the present invention is to provide a method for estimating the sound velocity and thickness of the ice layer by using air waves and ice bending waves based on a single sensor.

[0006] In order to solve the above technical problems, a method for estimating the sound velocity and thickness of ice layer by using air waves and ice layer bending waves of the present invention comprises the following steps:

[0007] Step 1: Use a sensor fixed on the ice surface to collect broadband pulse sound signals; sensors here include but are not limited to seismometers, accelerometers, microphones and other instruments that can monitor particle vibrations. Broadband pulse signals can usually be detonated on the ice surface using detonators, firecrackers or air cannons to stimulate bending wave signals propagating in the ice layer. Record the distance between the detonator and the ice surface sensor, set as R.

[0008] Step 2: Perform time-frequency analysis on the broadband pulse signal collected by the ice surface sensor; time-frequency analysis methods generally include short-time Fourier transform, wavelet transform and other methods. Wavelet transform is recommended here because it has high frequency resolution in the low frequency band. The time-frequency spectrum of the signal is obtained through time-frequency analysis, set as P(t,f), where P is the amplitude, t is the time, and f is the frequency. The bending wave arrives first in the time-frequency spectrum, and the air wave arrives later, and the time when the air wave arrives first is determined, set as t 1 .

[0009] Step 3: Measure the temperature of the experimental site and calculate the propagation speed of the air wave, set as c 1 ; The relationship between air propagation speed and temperature satisfies:

[0010] c 1 =331.6+0.6T

[0011] Where T represents temperature in degrees Celsius.

[0012] Step 4: Convert the time-frequency spectrum to the group velocity spectrum and extract the group velocity dispersion curve of the bending wave; the two coordinates of the time-frequency spectrum P(t,f) in step 2 are time t and frequency f, and the two coordinates of the group velocity spectrum P(c,f) are group velocity c and frequency f. The conversion relationship between the two coordinates of time t and group velocity is:

[0013]

[0014] The corresponding relationship between the time-frequency spectrum and the group velocity spectrum is obtained as follows: From c 1 is the propagation speed of air waves, t 1 represents the time when the air wave first arrives, R represents the distance between the detonator and the ice surface sensor, t represents time, f represents frequency, and P(c,f) is the group velocity spectrum.

[0015] Extract the bending waves in the group velocity spectrum P(c,f) to form a dispersion curve, denoted as c 2 (f), where f is the frequency of the bending wave, c 2 (f) represents the group velocity coordinate corresponding to the bending wave when the frequency coordinate is f in the group velocity spectrum P(c,f).

[0016] Step 5: Initialize the ice layer longitudinal wave velocity c s , shear wave velocity c t , ice thickness d, seawater sound speed c w ; The initial intervals of these four parameters are not fixed and can be modified using historical measurement data. The default settings are: the initial interval of longitudinal wave speed is 3000m / s-4500m / s, the initial interval of shear wave speed is 1600m / s-2000m / s, the initial interval of ice thickness is 0.1m-10.0m, and the initial interval of seawater speed is 1400m / s-1500m / s.

[0017] Step 6: Calculate the expected group velocity curve according to the sound propagation model and construct a cost function; the cost function is:

[0018]

[0019] Where N represents the number of frequency points of the dispersion curve extracted in step 4, c 2 (f n ) is when the nth frequency is f n The group velocity of the bending wave. n ) is the frequency f calculated using the sound propagation model n The theoretical group velocity of the bending wave at . The sound propagation model here uses KRAKEL in the simple normal wave model KRAKEN of underwater sound propagation, which can calculate the horizontal wave number k of the bending wave under the conditions of the longitudinal wave velocity of the ice layer, the shear wave velocity, the ice layer thickness, and the seawater sound velocity at a given frequency f. The relationship between the horizontal wave number and the group velocity is:

[0020]

[0021] in represents the derivative of frequency f with respect to horizontal wave number k.

[0022] Step 7: Use the particle swarm algorithm to optimize the cost function and output the final parameter estimation value; the particle swarm algorithm is a global optimization algorithm that is easy to calculate and can converge quickly. According to the parameters to be inverted set in step 5, the number of particle coordinates is set to 4, the number of particles is set to 100, and the upper and lower limits of the coordinates are the initialization settings of step 5. The particle swarm algorithm is used to minimize the cost function of step 6, and the four parameter coordinates of the particles are continuously updated until convergence is finally achieved. The longitudinal wave velocity, transverse wave velocity, ice thickness, and seawater sound velocity of the ice layer are estimated based on the average value of the 100 particle coordinates.

[0023] The beneficial effects of the present invention are as follows: the present invention can be realized by using only a single sensor, and does not require more sensors to form an array; this method does not require the synchronization of the transmitting device and the receiving device; it is proposed to use the sound speed and arrival time of the air wave to calculate the group velocity of the ice layer bending wave; the relationship between the group velocity and the acoustic parameters can be used to invert the longitudinal wave velocity, transverse wave velocity, ice layer thickness, and seawater sound speed of the ice layer; and finally the effectiveness is verified through ice experiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] Figure 1 is a parameter inversion flow chart of the present invention;

[0025] Figure 2 It is a diagram of the field experiment scene in the present invention;

[0026] Figure 3 It is the field experiment acquisition signal in the present invention;

[0027] Figure 4 is the signal time-frequency spectrum in the present invention;

[0028] Figure 5 is the group velocity spectrum in the present invention;

[0029] Figure 6 is the change of the cost function of the particle swarm optimization in the present invention;

[0030] Figure 7 is the parameter estimation result in the present invention;

[0031] Figure 8 This is the comparison result of the bending wave group velocity in the present invention. DETAILED DESCRIPTION

[0032] The present invention will be further described below in conjunction with the accompanying drawings and embodiments.

[0033] like Figure 1 As shown, the present invention provides a method for estimating the sound velocity and thickness of an ice layer by using air waves and ice layer bending waves, comprising the following steps:

[0034] like Figure 2 The figure shows a diagram of the field experiment scene, which includes a sensor fixed on the ice surface and an explosion sound source that can excite air waves and ice bending waves.

[0035] like Figure 3As shown, step 1 uses a sensor fixed on the ice surface to collect broadband pulse sound signals; the sensors here include but are not limited to seismic detectors, accelerometers, microphones and other instruments and equipment that can monitor particle vibrations. Broadband pulse signals can usually be detonated on the ice surface using detonators, firecrackers or air cannons to stimulate bending wave signals propagating in the ice layer. Record the distance between the detonator and the ice surface sensor, set to R. The ice experiment uses firecrackers to detonate on the ice surface, and a vertical accelerometer is fixed on the ice surface to collect vibration signals. The distance R = 200m, and the collected signal is as follows Figure 2 As shown in Figure 1, the first wave to arrive is the ice bending wave, followed by the air wave, which has a larger amplitude than the bending wave.

[0036] like Figure 4 As shown, step 2 performs time-frequency analysis on the explosion sound collected by the ice surface sensor; time-frequency analysis methods generally include short-time Fourier transform, wavelet transform and other methods. Wavelet transform is used here because it has a higher frequency resolution ability in the low frequency band. The time-frequency spectrum of the signal is obtained through time-frequency analysis, set as P(t,f), where P is the amplitude, t represents time, and f represents frequency. In the time-frequency spectrum, the bending wave arrives first, and the air wave arrives later, and the time when the air wave arrives first is determined, set as t 1 =0.665s.

[0037] Step 3: Measure the temperature of the experimental site and calculate the propagation speed of the air wave, set as c 1 ; The relationship between air propagation speed and temperature satisfies:

[0038] c 1 =331.6+0.6T

[0039] T represents temperature in degrees Celsius. The ambient temperature is about minus 20 degrees Celsius, so the air propagation speed is about 320m / s.

[0040] like Figure 5 As shown, step 4 converts the time-frequency spectrum into a group velocity spectrum and extracts the group velocity dispersion curve of the bending wave; in step 2, the two coordinates of the time-frequency spectrum P(t,f) are time t and frequency f, and the two coordinates of the group velocity spectrum P(c,f) are group velocity c and frequency f. The conversion relationship between the two coordinates of time t and group velocity is:

[0041]

[0042] The corresponding relationship between the time-frequency spectrum and the group velocity spectrum is obtained as follows: From c 1 is the propagation speed of air waves, t 1 represents the time when the air wave first arrives, R represents the distance between the detonator and the ice surface sensor, t represents time, f represents frequency, and P(c,f) is the group velocity spectrum.

[0043] Extract the bending waves in the group velocity spectrum P(c,f) to form a dispersion curve, denoted as c 2 (f) If Figure 4 As shown by the dotted line, f is the frequency of the bending wave, c 2 (f) represents the group velocity coordinate corresponding to the bending wave when the frequency coordinate is f in the group velocity spectrum P(c,f).

[0044] Step 5: Initialize the ice layer longitudinal wave velocity c s , shear wave velocity c t , ice thickness d, seawater sound speed c w ; The initial intervals of these four parameters are not fixed and can be modified using historical measurement data. The default settings are: the initial interval of longitudinal wave speed is 3000m / s-4500m / s, the initial interval of shear wave speed is 1600m / s-2000m / s, the initial interval of ice thickness is 0.1m-10.0m, and the initial interval of seawater speed is 1400m / s-1500m / s.

[0045] Step 6: Calculate the expected group velocity curve according to the sound propagation model and construct a cost function; the cost function is:

[0046]

[0047] Where N represents the number of frequency points of the dispersion curve extracted in step 4, c 2 (f n ) is when the nth frequency is f n The group velocity of the bending wave. n ) is the frequency f calculated using the sound propagation model n The theoretical group velocity of the bending wave when the wave is moving. The sound propagation model here uses KRAKEL in the simple normal wave model KRAKEN of underwater sound propagation, which can calculate the horizontal wave number k of the bending wave under the given ice layer longitudinal wave velocity, shear wave velocity, ice layer thickness, and seawater sound velocity. The calculation relationship between the horizontal wave number and the group velocity is:

[0048]

[0049] in represents the derivative of frequency f with respect to horizontal wave number k.

[0050] like Figure 6As shown, step 7 uses particle swarm algorithm to optimize the cost function and output the final parameter estimation value; particle swarm algorithm is a global optimization algorithm that is convenient to calculate and can converge quickly. According to the parameters to be inverted set in step 5, the number of particle coordinates is set to 4, the number of particles is set to 100, and the upper and lower limits of the coordinates are the initialization settings of step 5. The cost function of step 6 is minimized by particle swarm algorithm, and the four parameter coordinates of the particles are continuously updated until convergence is finally achieved. The longitudinal wave velocity, transverse wave velocity, ice thickness, and seawater sound velocity of the ice layer are estimated based on the probability density of the 100 particle coordinates. Figure 6 It shows that the optimization function value of particle swarm optimization is decreasing with the number of iterations, and the final cost function tends to a stable minimum value, indicating that the current coordinate parameters of the particles are consistent with the actual parameters.

[0051] like Figure 7 As shown in the figure, the final coordinate parameters of the particles are statistically analyzed, the probability density is plotted, and the relative relationship between the four parameters is shown. The final estimated parameter value is: ice layer longitudinal wave speed c s =4195m / s, shear wave velocity c t =1712m / s, seawater sound speed c w =1434m / s, ice layer thickness d=0.77m.

[0052] like Figure 8 As shown in the figure, the dotted line is the theoretical group velocity calculated using the parameter values ​​estimated in step 7, which is compared with the solid line in the figure based on the experimental observation value given in step 4, showing that the theoretical group velocity is consistent with the observed group velocity, reflecting the accuracy of the parameter estimation method.

[0053] It should be understood that the application of the present invention is not limited to the above examples. For ordinary technicians in this field, improvements or changes can be made based on the above description. All these improvements and changes should fall within the scope of protection of the claims attached to the present invention.

Claims

1. A method for estimating the sound velocity and thickness of ice layer using air waves and ice bending waves, characterized in that: The following steps are involved: Step 1: Use a sensor fixed on the ice surface to collect broadband pulse sound signals; record the distance between the detonator and the ice surface sensor, set as R; Step 2: Perform time-frequency analysis on the broadband pulse signal collected by the ice surface sensor; obtain the time-frequency spectrum of the signal through time-frequency analysis, set as P(t,f), where P is the amplitude, t is the time, and f is the frequency; Step 3: Measure the temperature of the experimental site and calculate the propagation speed of the air wave, set as c1; the relationship between the propagation speed of the air and the temperature satisfies: c1=331.6+0.6T Where T represents temperature in degrees Celsius; Step 4: Convert the time-frequency spectrum to the group velocity spectrum and extract the group velocity dispersion curve of the bending wave; the two coordinates of the time-frequency spectrum P(t,f) in step 2 are time t and frequency f, and the two coordinates of the group velocity spectrum P(c,f) are group velocity c and frequency f; the conversion relationship between the two coordinates of time t and group velocity is: The corresponding relationship between the time-frequency spectrum and the group velocity spectrum is obtained as follows: Where c1 is the propagation speed of the air wave, t1 represents the time when the air wave first arrives, R represents the distance between the detonator and the ice surface sensor, t represents time, f represents frequency, and P(c,f) is the group velocity spectrum; Extract the bending wave in the group velocity spectrum P(c,f) to form a dispersion curve, set as c2(f), where f is the frequency point of the bending wave, and c2(f) represents the group velocity coordinate corresponding to the bending wave in the group velocity spectrum P(c,f) when the frequency coordinate is f; Step 5: Initialize the ice layer longitudinal wave velocity c s , shear wave velocity c t , ice thickness d, seawater sound speed c w ; Step 6: Calculate the expected group velocity curve according to the sound propagation model and construct the cost function; Step 7: Use the particle swarm algorithm to optimize the cost function and output the final parameter estimation value; according to the parameters to be inverted set in step 5, set the number of particle coordinates to 4, the number of particles to 100, and the upper and lower limits of the coordinates are the initialization settings of step 5; minimize the cost function of step 6 through the particle swarm algorithm, continuously update the four parameter coordinates of the particles, and finally reach convergence, and estimate the longitudinal wave velocity, shear wave velocity, ice thickness, and seawater sound speed based on the mean value of the 100 particle coordinates.

2. The method for estimating the sound velocity and thickness of ice layer using air waves and ice bending waves according to claim 1, characterized in that: The sensor in step 1 is a seismometer, an accelerometer or a microphone; the broadband pulse signal is detonated on the ice surface using a detonator, a firecracker or an air cannon to excite a bending wave signal propagating in the ice layer.

3. The method for estimating the sound velocity and thickness of ice layer by using air waves and ice bending waves according to claim 1, characterized in that: In step 2, the time-frequency analysis method uses wavelet transform.

4. The method for estimating the sound velocity and thickness of ice layer using air waves and ice bending waves according to claim 1, characterized in that: The initial intervals of the four parameters in step 5 are modified using historical measurement data; the default settings are: the initial interval of longitudinal wave speed is 3000m / s-4500m / s, the initial interval of shear wave speed is 1600m / s-2000m / s, the initial interval of ice thickness is 0.1m-10.0m, and the initial interval of seawater sound speed is 1400m / s-1500m / s.

5. The method for estimating the sound velocity and thickness of ice layer by using air waves and ice bending waves according to claim 1, characterized in that: The cost function in step 6 is: Where N represents the number of frequency points of the dispersion curve extracted in step 4, c2(f n ) is when the nth frequency is f n The group velocity of the bending wave when n ) is the frequency f calculated using the sound propagation model n The theoretical group velocity of the bending wave at the time; the sound propagation model uses KRAKEL in the simple normal wave model KRAKEN of underwater sound propagation, and calculates the horizontal wave number k of the bending wave under the conditions of the longitudinal wave velocity of the ice layer, the shear wave velocity, the ice layer thickness, and the seawater sound speed at a given frequency f; the relationship between the horizontal wave number and the group velocity is: in represents the derivative of frequency f with respect to horizontal wave number k.

Images