JP6637592B2 - Signal processing system, signal processing method, signal processing device, and program - Google Patents

Description

  The present invention relates to a signal processing system, a signal processing method, a signal processing device, and a program, and more particularly to a technique for processing a signal based on vibration such as seismic exploration.

  2. Description of the Related Art In ground surveys and damage surveys inside existing structures, nondestructive measurement is performed to measure the characteristics of a medium to be investigated without destroying the medium. Patent Literature 1 discloses a method of measuring a characteristic of a medium by installing a plurality of exciters so as to be in contact with a medium and receiving and analyzing signals transmitted from these exciters by a receiver. ing. In this method, each oscillator transmits a vibration wave based on a pseudo-random code signal having a different pattern, and a receiver receives a combined vibration wave obtained by multiplexing vibration waves transmitted from a plurality of oscillators. . By using the correlation between the vibration wave transmitted from the oscillator and the synthetic vibration wave received by the receiver, information on a specific vibration wave is separated.

  Such a technique is a technique of discriminating after simultaneously receiving signals transmitted from a plurality of epicenters. In particular, a technique called stacking, in which a plurality of continuous transmission signals are added in a transmission cycle after reception, can obtain theoretically excellent correlation. This technique requires a long measurement time.

JP 2004-163322 A

  However, in the technique called stacking, when the number of times of stacking is small, the correlation between the vibration wave transmitted from the oscillator and the combined vibration wave received by the receiver deteriorates. In order to prevent this, a method of transmitting a dummy transmission signal in advance has been proposed, but there has been a problem that the accuracy of discrimination cannot be increased, for example, when the transmission source moves frequently.

  The present invention has been made in view of these points, and an object of the present invention is to improve the discrimination accuracy when discriminating signals received from a plurality of exciters simultaneously after reception.

  A first aspect of the present invention is a signal processing system. This system receives N (N is an integer of 1 or more) oscillators that transmit a vibration wave to a medium, and receives a combined wave of the vibration waves transmitted from each of the N oscillators and propagated through the medium. And a receiver that cuts out data of a predetermined length D from the end of the multiplexed time series data received by the receiver, and converts the data into time series data from the beginning of the time series data to the predetermined length D. A correction data generation unit that generates the corrected time series data that has been added; and a data indicating a correlation between each of the time series data of the N vibration waves transmitted by the N pieces of the shakers and the corrected time series data. And a correlation calculating unit for calculating.

  Each of the N oscillators may transmit a vibration wave based on a pseudo-random signal that is uncorrelated with each other within the range of the predetermined length D.

  Each of the N oscillators may transmit the vibration wave in synchronization with each other, and the receiver may control the time-series data in synchronization with the transmission of the vibration wave by each of the N oscillators. May be started.

  Each of the N vibrators outputs a pseudo-random signal having a period equal to or longer than the predetermined length D multiplied by N + 1 by M times the predetermined length (M is 0 or more and N-1 or less) (Different for each of the N oscillators), and a vibration wave based on a signal obtained by cyclically shifting by a length is transmitted.

  The correlation calculating section sequentially shifts the corrected time-series data from 0 to a length obtained by multiplying the predetermined length D by N, and sequentially calculates an autocorrelation coefficient with the pseudo-random signal. It may be calculated.

  The data calculated by the correlation calculation unit is analyzed, and the delay time of each vibration wave, which is the time required for each of the N oscillators to transmit the vibration wave and reach the receiver, is calculated. The computer may further include a delay time calculating unit for calculating the delay time.

  Each of the N oscillators is configured such that the length of the vibration wave to be transmitted until the time of the maximum value of the delay time of each of the vibration waves elapses is equal to or less than the predetermined length D. A vibration wave may be transmitted.

  A second aspect of the present invention is a signal processing method. The method comprises the steps of transmitting N (N is an integer of 1 or more) vibration waves to a medium, receiving a multiplex of the N vibration waves propagated through the medium, and converting time-series data of the received waves. Generating, and generating corrected time-series data by cutting out data of a predetermined length D from the end of the time-series data and adding it to time-series data from the beginning of the time-series data to the predetermined length D. And executing a step of calculating data indicating a correlation between each of the transmitted time series data of the N vibration waves and the corrected time series data.

  A third aspect of the present invention is a signal processing device. This device includes a plurality of N (N is an integer of 1 or more) oscillators that transmit a vibration wave to a medium, and each of the N oscillators transmits and transmits a vibration wave transmitted through the medium. A time-series data generation unit that generates time-series data, and cuts out data having a predetermined length D from the end of the time-series data, and converts the data into time-series data from the beginning of the time-series data to the predetermined length D. A correction data generation unit that generates the corrected time series data that has been added; and a data indicating a correlation between each of the time series data of the N vibration waves transmitted by the N pieces of the shakers and the corrected time series data. And a correlation calculating unit for calculating.

  A fourth aspect of the present invention is a program. This program is obtained by receiving, by a computer, each of N (N is an integer of 1 or more) vibrators that transmit a vibration wave to a medium and receives a combined wave of the vibration waves transmitted through the medium. A function of generating time-series data of a received wave, extracting data of a predetermined length D from the end of the time-series data, and adding the data to the time-series data from the head of the time-series data to the predetermined length D A function of generating corrected time-series data, and a function of calculating data indicating a correlation between each of the N time-series data of the N vibration waves transmitted by each of the N oscillators and the corrected time-series data. To realize.

  ADVANTAGE OF THE INVENTION According to this invention, the discrimination precision at the time of discriminating after receiving the signal simultaneously transmitted from the several shaking devices after receiving can be improved.

FIG. 1 is a diagram for describing an overview of a signal processing system according to an embodiment. It is a figure which shows the example of arrangement | positioning of the vibrator and receiver which concerns on embodiment. FIG. 2 is a diagram schematically illustrating a functional configuration of a vibrator and a receiver according to the embodiment. FIG. 4 is a diagram illustrating an example of a pseudo-random signal generated by a pseudo-random number generation unit and a random number shift unit according to the embodiment. It is a figure for explaining a vibration wave which a vibration wave generation part concerning an embodiment generates. FIG. 3 is a diagram schematically illustrating a waveform of a vibration wave transmitted by the oscillator according to the embodiment, transmitted through the medium, and received by the receiver. FIG. 7 is a diagram for describing a process of generating correction time-series data by a correction data generation unit according to the embodiment. It is a schematic diagram for demonstrating the correction | amendment time series data regarding the received waveform of the vibration wave which the 1st vibration generator sent. FIG. 5 is a diagram illustrating a graph of a time change of an autocorrelation coefficient calculated by a correlation calculation unit according to the embodiment. FIG. 9 is a diagram for explaining a method of calculating an autocorrelation coefficient according to a comparative example. FIG. 9 is a diagram illustrating a graph of a time change of an autocorrelation coefficient calculated using an autocorrelation coefficient calculation method according to a comparative example. It is a graph which shows the example of the actual measurement data of the autocorrelation function in seismic survey.

  FIG. 1 is a diagram for describing an overview of a signal processing system 100 according to an embodiment. The signal processing system 100 includes N (N is an integer of 1 or more) oscillators 1 and at least one receiver 2. In the example shown in FIG. 1, the value of N is 4, and four exciters 1 from the first exciter 1a to the fourth exciter 1d are illustrated.

  Each of the plurality of oscillators 1 generates a pseudo-random signal that is uncorrelated with each other within a range of a predetermined length D, and transmits a vibration wave generated based on the generated signal into a medium. The exciters 1 are synchronized with each other, and simultaneously transmit the vibration waves.

  The “predetermined length D” is the length of the vibration wave transmitted by each exciter 1 until the “predetermined time T” elapses. Although the details will be described later, the “predetermined time T” is expected as the time required for the vibration waves transmitted by each of the N oscillators 1 to propagate through the medium to be measured and reach the receiver 2. This is the largest time of the day. The predetermined time T is, as it were, the expectation of the maximum value of the “delay time” defined as the time required from the transmission of the vibration wave by the oscillator 1 to the arrival at the receiver 2.

  Each of the N oscillators 1 transmits a vibration wave such that the length of the vibration wave to be transmitted until the predetermined time T elapses becomes less than or equal to the predetermined length D. Note that the predetermined time T may not be the same as the maximum value as long as it is longer than the maximum value in the delay time.

  The receiver 2 receives the vibration wave propagated in the medium. The receiver 2 starts receiving the vibration wave at the same time that the vibration generator 1 transmits the vibration wave in synchronization with the transmission of the vibration wave by each of the N number of the vibration generators 1. Here, the receiver 2 continues receiving from the time when each of the oscillators 1 has finished transmitting the vibration wave until at least the vibration wave having the predetermined length D is received.

  Since each of the oscillators 1 simultaneously transmits a vibration wave to the same medium, the receiver 2 receives a combined vibration wave in which a plurality of vibration waves generated by the plurality of vibration generators 1 are combined. . The synthetic vibration received by the receiver 2 is converted into time-series data corresponding to the synthetic vibration wave in the signal processing device 3 (for example, a computer) and analyzed.

  The signal processing device 3 converts the synthetic vibration wave received by the receiver 2 into time-series data. The signal processing device 3 further cuts out data having a length corresponding to the predetermined length D from the end of the time-series data. The signal processing device 3 generates corrected time-series data by adding the cut-out data to time-series data from the beginning of the time-series data to a length corresponding to a predetermined length D.

  The signal processing device 3 calculates a correlation between the corrected time-series data and each of the N vibration waves transmitted from the N receivers 2. As a result, the signal processing device 3 can separate the component based on the vibration wave transmitted from the specific vibrator 1 from the received signal based on the synthetic vibration wave received by the receiver 2, and the N vibration generators 1 The characteristics of the medium between them can be specified.

  More specifically, the signal processing device 3 specifies the timing at which the calculated correlation value reaches a peak and the magnitude of the correlation coefficient at the peak. The timing at which the correlation coefficient reaches a peak corresponds to the above-described delay time, from which the transmission speed of the vibration wave in the medium can be estimated. Also, from the magnitude of the peak of the correlation coefficient, it is possible to estimate the amount of attenuation of the vibration wave due to propagation through the medium. From these results, the signal processing device 3 can specify the characteristics of the medium between the specific oscillator 1 and the receiver 2.

The above is the outline of the signal processing executed by the signal processing device 3 in the signal processing system 100 according to the embodiment. The signal processing device 3 cuts out a signal having a predetermined length D from the end of the synthetic vibration wave received by the receiver 2 and adds the signal to the head of the synthetic vibration wave to generate a signal as a correlation calculation target. This is one feature of the signal processing system 100 according to the embodiment.
Hereinafter, the signal processing system 100 will be described in more detail.

  FIG. 2 is a diagram illustrating an example of the arrangement of the oscillator 1 and the receiver 2 according to the embodiment. In FIG. 2, four exciters 1 (first to fourth exciters 1a to 1d) and a receiver 2 are shown. The vibrating waves transmitted by the respective exciters 1 propagate through the ground, which is an example of the medium M, and reach the receiver 2 after being reflected and scattered according to the characteristics of the substance in the ground.

  The satellite S is a navigation satellite such as a GPS (Global Positioning System). Each of the oscillator 1 and the receiver 2 can receive the radio wave transmitted by the satellite S and synchronize with each other by using time information included in the received radio wave.

  The receiver 2 is connected to the signal processing device 3 via a cable or a wireless line, and the receiver 2 sends a signal based on the received vibration wave to the signal processing device 3. The signal processing device 3 specifies the characteristics of the medium M by analyzing the signal received from the receiver 2.

FIG. 3 is a diagram schematically illustrating a functional configuration of the oscillator 1 and the receiver 2 according to the embodiment.
The oscillator 1 includes a pseudo random number generation unit 11, a random number shift unit 12, a vibration wave generation unit 13, a clock generation unit 14, and a vibration wave storage unit 15. The exciter 1 is, for example, vibroseis.

  The clock generation unit 14 generates a clock based on a radio wave received from the satellite S. The clock generation unit 14 outputs the generated clock to the pseudo random number generation unit 11. Since the clock generated by the clock generation unit 14 is synchronized with the time information included in the radio wave received from the satellite S, all the oscillators 1 included in the signal processing system 100 can operate in synchronization.

  The pseudo-random number generator 11 generates a pseudo-random signal having a predetermined period P based on the clock input from the clock generator 14. More specifically, the pseudo-random number generation unit 11 generates an auto-correlation pseudo-random number pattern selected from an M-sequence signal having a predetermined period P. Such a pseudo-random signal expressing the autocorrelation pseudo-random pattern in an appropriate modulation scheme is 1 at a lag time of 0, and becomes -1 / P when the lag time is other than 0.

Here, the predetermined period P is a value equal to or greater than the length obtained by multiplying the predetermined length D described above by the number N + 1 of the oscillators 1 included in the signal processing system 100. That is,
P ≧ (N + 1) D (1)
Is the length that satisfies

  The random number shift unit 12 generates a signal obtained by cyclically shifting the pseudo random signal generated by the pseudo random number generation unit 11 by a length M times the predetermined length D. Here, M is an integer equal to or greater than 0 and equal to or less than N−1, and differs for each of the N oscillators 1. In the case of the signal processing system 100 according to the embodiment, M = 0 to 3 are assigned to the first shaker 1a to the fourth shaker 1d, respectively.

  FIGS. 4A and 4B are diagrams illustrating an example of a pseudo random signal generated by the pseudo random number generation unit 11 and the random number shift unit 12 according to the embodiment. More specifically, FIG. 4A shows an example of a pseudo-random signal commonly generated by the pseudo-random number generation units 11 of the respective oscillators 1, and FIG. 4B illustrates the first oscillator 1a. , The second shaker 1b, the third shaker 1c, and the fourth shaker 1d show pseudo-random signals S1 to S4 output from the random number shift units 12 respectively.

  Although the details will be described later, the vibration wave generator 13 generates a vibration wave having a positive amplitude when the value of the pseudo-random signal is 1 and a negative amplitude when the value of the pseudo random signal is 0 at 0.1 [s] = 1 [bit]. Generated at intervals of 100 [ms].

  In the signal processing system 100 according to the embodiment, the maximum value of the “delay time” defined as the time required for the oscillator 1 to transmit the vibration wave to reach the receiver 2 is expected to be 10 [s]. Have been. Therefore, the predetermined time T is also set to 10 [s], and as a result, the predetermined length D is 100 bits (10 [s] /0.1 [s] = 100).

Since the number of the vibrators 1 included in the signal processing system 100 according to the embodiment is four, the predetermined period P is obtained from Equation (1) as
P ≧ (4 + 1) × 100 = 500 (2)
Is set as a value that satisfies.

  In FIG. 4 (a), the rectangles indicated by symbols A to E each include a signal having a length of 100 bits, that is, a signal having the same length as a predetermined length D. The rectangle indicated by the symbol F includes an 11-bit signal. The signal generated by the oscillator 1 is a signal having a length of 511 bits obtained by connecting signals included in rectangles indicated by symbols A to F. Since the vibration wave generation unit 13 generates vibration waves at intervals of 0.1 [s] per bit, the signal generated by the vibrator 1 is a signal having a length of 51.1 [s].

  Here, the period of the M-sequence generated by the pseudo random number generation unit 11 is set to 511 bits or more. Therefore, the correlation of the signal generated by the first shaker 1a is 1 at a lag time of 0, and becomes almost 0 at other times. Therefore, the signal generated by the first shaker 1a satisfies Expression (2).

  Hereinafter, for convenience of description, signals included in rectangles indicated by symbols A to F may be described as signals A to F, respectively. Further, 511 bits, which is the length of the pseudo random signal generated by the pseudo random number generation unit 11 and output by the random number shift unit 12, may be described as a signal length L.

  The above-described value of M is set in advance in a storage unit (not shown) of the random number shift unit 12 in each shaker 1. For example, since the value of M set in the random number shift unit 12 of the first oscillator 1a is 0, the random number shift unit 12 of the first oscillator 1a does not shift the pseudo random signal generated by the pseudo random number generator 11. As a result, the pseudo random signal S1 output from the random number shift unit 12 of the first shaker 1a becomes the same as the pseudo random signal generated by the pseudo random number generation unit 11, and as shown in FIG. The signals are connected in this order.

  M = 1 is set in the random number shift unit 12 of the second shaker 1b. For this reason, the random number shift unit 12 of the second oscillator 1b outputs a signal obtained by cyclically shifting the pseudo random signal generated by the pseudo random number generation unit 11 by a predetermined length D. Therefore, as shown in FIG. 4B, the pseudo random signal S2 output from the random number shift unit 12 of the second shaker 1b is a signal in which the signals B to F and the signal A are connected in this order.

  Similarly, M = 2 and 3 are set in the random number shift unit 12 of the third oscillator 1c and the random number shift unit 12 of the fourth oscillator 1d, respectively. Therefore, the pseudo random signal S2 output from the random number shift unit 12 of the third oscillator 1c becomes a signal in which the signals C to F and the signals A and B are connected in this order, and the random number shift unit of the fourth oscillator 1d. The pseudo random signal S2 output by the signal 12 is a signal in which the signals D to F and the signals A to C are connected in this order.

  As described above, the pseudo-random signals S1 to S4 are signals generated by cyclically shifting the same pseudo-random signal having a period equal to or longer than the signal length L by at least a predetermined length D. Therefore, the pseudo-random signals S1 to S4 are pseudo-random signals that are uncorrelated with each other at least within a range of a predetermined length D.

  FIGS. 5A to 5C are diagrams for explaining a vibration wave generated by the vibration wave generation unit 13 according to the embodiment. The vibration wave generator 13 drives a hydraulically driven actuator (not shown) based on the pseudo random signal output from the random number shifter 12 to generate a vibration wave.

  FIG. 5A is a diagram illustrating a relationship between the amount of oil supplied to the actuator and the displacement of the actuator when the value of the signal is “1”. In FIG. 5A, the horizontal axis indicates the passage of time, and the vertical axis indicates the displacement of the actuator or the amount of oil supply. In FIG. 5A, the broken line indicates the time variation of the oil supply amount for driving the actuator, and the solid line indicates the time variation of the displacement of the actuator.

  The vibration wave generator 13 starts oil supply to the actuator at time 0 [ms], and increases the oil supply amount until time 25 [ms]. The vibration wave generator 13 reduces the amount of oil supply after time 25 [ms], and sets the amount of oil supply to 0 at time 50 [ms]. Thereafter, the vibration wave generating unit 13 continues to reduce the oil supply amount until the time 75 [ms], and after the time 75 [ms], increases the oil supply amount again. At time 100 [ms], the vibration wave generation unit 13 sets the oil supply amount to 0. The above is the control of the amount of oil supplied to the actuator and the displacement of the actuator performed by the vibration wave generation unit 13 when the value of the signal is “1”. The vibration wave generator 13 drives the actuator at a time of 100 [ms] per signal bit.

  Here, the displacement of the actuator is a value obtained by time-integrating the oil supply amount for driving the actuator. For this reason, as shown in FIG. 5A, the displacement of the vibration wave generating unit 13 becomes maximum at the time 50 [ms], and the displacement speed is 0 at the times 0 [ms] and 100 [ms] (that is, FIG. 5 (a) is 0). As described above, the vibration wave generator 13 sets the displacement speed to 0 at the start and end of the displacement of the actuator, thereby reducing the load on the actuator and reducing the failure rate of the vibrosize. .

  FIG. 5B schematically shows the outline of the vibration wave generated by the vibration wave generator 13 based on the first 5 bits of the pseudo random signal S1 generated by the random number shifter 12 of the first oscillator 1a. FIG. Since the first 5 bits of the pseudo-random signal S1 are "10001", as shown in FIG. 5 (b), the vibration wave generator 13 generates a vibration wave having a positive amplitude three times in succession. A vibration wave having a positive amplitude is generated, and then a vibration wave having a positive amplitude is generated again. As described above, the vibration wave generating unit 13 drives the actuator based on the pseudo random signal to generate a vibration wave, thereby converting the pseudo random signal which is a discrete signal into a continuous vibration wave.

  FIG. 5C is a diagram illustrating an outline of an autocorrelation function of the vibration wave generated by the vibration wave generation unit 13 based on the pseudo random signal. As shown in FIG. 5C, the vibration wave generated by the vibration wave generator 13 has good autocorrelation.

  The vibration wave storage unit 15 samples the vibration wave generated by the vibration wave generation unit 13 and generates and stores time-series data of the vibration wave. The vibration wave storage unit 15 is realized by a known AD converter and storage device.

  More specifically, the vibration wave storage unit 15 samples the vibration waves generated by the vibration wave generation unit 13 at a sampling interval of 6.25 [ms], and accumulates them in time series in a 64-bit floating point format. I do. The vibration wave generation unit 13 generates a vibration wave with a length of 100 [ms] per bit of the pseudo random signal output from the random number shift unit 12. Therefore, as a result, the data stored in the vibration wave storage unit 15 is data having 511 × (100 / 6.25) = 511 × 16 = 8176 elements. Hereinafter, in this specification, the time-series data stored by the vibration wave generator 13 may be described as a transmission signal array S (j), (j = 0,..., 8175).

  FIGS. 6A to 6E are diagrams schematically illustrating waveforms of vibration waves transmitted by the oscillator 1 according to the embodiment, propagated through the medium M, and received by the receiver 2. The vibration wave generated by the vibration wave generator 13 of the vibrator 1 is delayed and attenuated according to the distance that the medium M existing between the vibrator 1 and the receiver 2 propagates and the velocity of the vibration wave in the medium M. And reaches the receiver 2. FIGS. 6A to 6E are diagrams showing waveforms of vibration waves generated by simulation. Hereinafter, the figures shown in FIGS. 7 and 9 similarly show calculation results by simulation.

  FIGS. 6A to 6D show received waveforms Sr1 to Sr4 of the vibration waves transmitted from the first to fourth oscillators 1a to 1d, respectively. 6A to 6D, the horizontal axis indicates time, and the vertical axis indicates the amplitude of the vibration wave when the vibration wave reaches the receiver 2. 6 (a) to 6 (d), τ1 to τ4 are time periods until the vibration wave transmitted from the first oscillator 4a to the fourth oscillator 1d to the medium M reaches the receiver 2, that is, the delay described above. Indicates time.

  As shown in FIG. 2, the paths through which the vibration waves transmitted by the respective exciters 1 propagate through the medium M until reaching the receiver 2 are different from each other. In the example shown in FIG. 2, the length of the path through which the vibration wave transmitted from each oscillator 1 propagates through the medium M until it reaches the receiver 2 is the first oscillator 1a, the second oscillator 1b, and the third oscillator. The first and fourth exciters 1c and 1d are shorter in this order. Therefore, the delay time is also shortened in the order of τ1, τ2, τ3, and τ4. The vibration wave is attenuated while propagating through the medium M. For this reason, the amplitudes of the received waveforms Sr1 to Sr4 of the vibration wave when reaching the receiver 2 decrease in this order.

  Each of the oscillators 1 is synchronized, and each of the oscillators 1 simultaneously transmits a vibration wave into the medium M. For this reason, the receiver 2 does not individually receive the vibration waves shown in FIGS. 6A to 6D, but combines the vibration waves transmitted from each of the oscillators 1 and propagated through the medium M. Receive a synthetic vibration wave.

  FIG. 6E is a diagram illustrating a waveform of the synthetic vibration wave received by the receiver 2. As shown in FIG. 6 (e), the synthetic vibration wave received by the receiver 2 is a superposition of the received waveforms Sr1 to Sr4 shown in FIGS. 6 (a) to 6 (d). )-The vibration wave shown in (d) cannot be separated. The signal processing device 3 uses the statistical properties of the pseudo-random signals S1 to S4 to extract information indicating the delay time τ1 to τ4 and the amplitude of each vibration wave from the synthetic vibration wave.

Returning to the description of FIG. The signal processing device 3 includes a reception unit 31, a time-series data generation unit 32, a vibration wave reproduction unit 33, a clock generation unit 34, a correction data generation unit 35, a correlation calculation unit 36, and a delay time calculation unit 37.
The clock generator 34 has a function equivalent to that of the clock generator 14 and generates a clock based on radio waves received from the satellite S. The clock generator 34 outputs the generated clock to the receiver 31. The receiver 31 receives the received signal of the synthetic vibration wave generated by the receiver 2 in synchronization with the clock acquired from the clock generator 34.

  Here, it is considered that the receiving unit 31 receives a signal including receiving a received signal from the receiver 2. In this case, since the clock generator 14 of each oscillator 1 and the clock generator 34 of the receiver 2 are both synchronized with the time information of the satellite S, they can output a clock signal on a common time axis. As a result, the signal processing device 3 can start receiving signals at the same time as each of the oscillators 1 sends out a vibration wave.

  Since it takes some time from the start of the vibration of the oscillator 1 to the arrival of the vibration wave to the arrival at the receiver 2, the vibration wave cannot be measured immediately after the receiver 2 starts the reception. Therefore, when the clock generation unit 34 starts receiving the signal from the receiver 2 and continues receiving the vibration wave transmission time of 51.1 [s], it receives all the vibration waves transmitted by the vibration generator 1. I can't do it.

  Here, since the maximum delay time is expected to be 10 [s], the clock generation unit 34 starts receiving a signal from the receiver 2 and then 51.1 [s] +10 [s] = 61.1. [S] = 61100 [ms] If the reception is continued, all the vibration waves transmitted by the exciter 1 can be completely received.

  The time-series data generation unit 32 generates time-series data of the synthetic vibration wave received by the reception unit 31. The time-series data generation unit 32 is a known AD converter, and converts a synthetic vibration wave into digital data at a sampling interval of 25 [ms]. As a result, the data generated by the time-series data generator 32 is data having 61100 [ms] / 25 [ms] = 2444 elements. This data is finally converted to 64-bit floating-point format data, and becomes 64-bit × 2444 record data. Hereinafter, in this specification, the time-series data generated by the time-series data generation unit 32 may be referred to as a received signal array R (j), (j = 0,..., 2443).

  The correction data generation unit 35 cuts out a signal having a length corresponding to a predetermined length D from the end of the reception signal array R (j) generated by the time-series data generation unit 32. The correction data generating unit 35 adds correction time-series data Cd (j), (j) obtained by adding the cut-out signal to time-series data from the beginning of the received signal array R (j) to a length corresponding to the predetermined length D. = 0,..., 2043).

Here, the predetermined length D in the pseudo random signal generated by the pseudo random number generation unit 11 corresponds to 100 bits. The time series data generation unit 32 performs sampling four times while the vibration wave generation unit 13 generates a 1-bit vibration wave. Therefore, the predetermined length D of the corrected time series data Cd (j) is 400 data numbers. Would be equivalent to As described above, the corrected time-series data Cd (j) can be expressed by the following equation (3).

  FIGS. 7A and 7B are diagrams illustrating a process of generating the correction time-series data Cd (j) by the correction data generating unit 35 according to the embodiment. FIG. 7A is a diagram illustrating a process of correcting the received signal array R (j) by the correction data generation unit 35. FIG. 7B is a diagram illustrating the correction time-series data Cd (j) generated by the correction data generation unit 35 as a result of the correction processing. As shown in FIG. 7, the correction data generation unit 35 adds the correction time-series data Cd (j) obtained by adding the last 400 data of the reception signal array R (j) to the head of the reception signal array R (j). Generate. Finally, the number of elements of the corrected time-series data Cd (j) is 2044.

  FIGS. 8A and 8B are schematic diagrams for explaining the modified time-series data Cd (j) relating to the received waveform Sr1 of the vibration wave transmitted by the first vibration generator 1a. As shown in FIG. 8A, the reception waveform Sr1 reaches the receiver 2 after a delay time τ1 has elapsed since the clock generation unit 34 started generating the reception signal array R (j). Therefore, until the delay time τ1 elapses after the clock generation unit 34 starts generating the reception signal array R (j), the reception waveform Sr1 includes a signal derived from the vibration wave transmitted by the first oscillator 1a. Is not included. The correction data generation unit 35 adds the last 400 pieces of data of the reception signal array R (j) to the head of the reception signal array R (j), so that the correction time-series data Cd (j) can be obtained over the entire area. This includes a signal originating from the vibration wave transmitted from one shaker 1a.

  The vibration wave transmitted by the oscillator 1 basically keeps the shape of the vibration wave generated by the vibration wave generation unit 13 even when the vibration wave propagates through the medium M. Therefore, as shown in FIG. 8B, the correlation is calculated while relatively cyclically shifting the corrected time-series data Cd (j) and the transmission signal array S (j) representing the waveform of the vibration wave. , The delay time τ.

  Returning to the description of FIG. The vibration wave reproducing unit 33 acquires the transmission signal array S (j) stored in the vibration wave storage unit 15. The vibration wave reproducing unit 33 may acquire the transmission signal array S (j) from the vibration wave storage unit 15 by wire communication or wireless communication, for example, or transmit the transmission signal array S (j) copied to a portable recording medium. ) May be obtained by reading out from the recording medium.

  The correlation calculation unit 36 calculates data indicating a correlation between each of the transmission signals S (j) of the N vibration waves transmitted by the N oscillators 1 and the corrected time-series data Cd (j). More specifically, the correlation calculator 36 sequentially shifts the transmission time series of the corrected time series data Cd (j) while cyclically shifting the corrected time series data Cd (j) from 0 to a length obtained by multiplying the predetermined length D by N. An autocorrelation coefficient with the array S (j) is calculated. That is, the correlation calculation unit 36 calculates a time change of the autocorrelation coefficient with the corrected time-series data Cd (j) and the transmission signal array S (j).

More specifically, the correlation calculator 36 calculates the autocorrelation coefficient ρ (0) when the delay time is 0 according to the following equation (4).

In Equation (4), the index of the transmission signal array S (j) is multiplied by 4 because the transmission signal array S (j) has a sampling rate (ie, the sampling rate) four times that of the modified time-series data Cd (j). This is because the interval is sampled at 1/4). Cav represents the time average of the corrected time series data Cd (j), and Sav represents the time average of the transmission signal array S (j). That is, Cav and Sav are calculated according to the following equations (5) and (6), respectively.

Subsequently, the correlation calculator 36 cyclically shifts the transmission signal array S (j) by one data according to the following equations (7-1) and (7-2).
S (0) ← S (8175) (7-1)
S (j) ← S (j−1), (j = 1,..., 8175) (7-2)

  The correlation calculation unit 36 substitutes the transmission signal array S (j) cyclically shifted according to the equations (7-1) and (7-2) into the equation (4) to obtain the transmission signal array S ( The autocorrelation coefficient ρ (1) at a delay time of 6.25 [ms] corresponding to one sampling of j) can be calculated.

  In the signal processing system 100 according to the embodiment, the maximum value in the delay time defined as the time required for the oscillator 1 to transmit the vibration wave to reach the receiver 2 is 10 [s] = 10000 [ms]. It is expected. Therefore, in the transmission signal array S (j), the data of 10,000 [ms] /6.25 [ms] = 1600 corresponds to the predetermined length D.

  The correlation calculating section 36 cyclically shifts the transmission signal array S (j) one data at a time according to the equations (7-1) and (7-2), and increases the autocorrelation coefficient ρ of the equation (4) 1600 times. Ρ (0) to ρ (1599) are obtained. ρ (0) to ρ (1599) are time changes of the autocorrelation coefficient ρ from the start of transmission of the exciter 1 to the time corresponding to the predetermined length D. Of the autocorrelation coefficients ρ (0) to ρ (1599) calculated by the correlation calculation unit 36, the self-correlation coefficient corresponding to the delay time τ1, which is the time when the vibration wave transmitted from the first oscillator 1a arrives at the receiver 2, is used. In the correlation coefficient ρ, the autocorrelation coefficient ρ generates a steep peak.

  As described with reference to FIG. 4, when the pseudo random signal S1 output from the random number shift unit 12 of the first oscillator 1a is cyclically shifted by a predetermined length D, the random number shift unit of the second oscillator 1b is shifted. 12 coincides with the output pseudo-random signal S2. Similarly, when the pseudo random signal S2 output from the random number shift unit 12 of the second oscillator 1b is cyclically shifted by a predetermined length D, the pseudo random signal S3 output from the random number shift unit 12 of the third oscillator 1c is output. Matches. When the pseudo random signal S3 output by the random number shift unit 12 of the third oscillator 1c is cyclically shifted by a predetermined length D, the pseudo random signal S4 coincides with the pseudo random signal S4 output by the random number shift unit 12 of the fourth oscillator 1d. I do.

  This means that the transmission signal array S (j) obtained by the correlation calculator 36 cyclically shifting the transmission signal array S (j) 1600 times according to the equations (7-1) and (7-2) is , The transmission signal array S (j) of the vibration wave transmitted by the vibration wave generation unit 13 of the second oscillator 1b. Therefore, the correlation calculation unit 36 calculates the autocorrelation coefficients ρ (1600) to ρ (3199) according to the equations (7-1), (7-2), and (4), thereby obtaining the second quake. It is possible to calculate the time change of the autocorrelation coefficient ρ between the transmission signal array S (j) and the corrected time-series data Cd (j) in the vibration wave transmitted by the detector 1b.

  Similarly, the correlation calculation unit 36 calculates the autocorrelation coefficients ρ (3200) to ρ (4799), thereby obtaining the transmission signal array S (j) in the vibration wave transmitted from the third oscillator 1c and the corrected time series The time change of the autocorrelation coefficient ρ with the data Cd (j) can be calculated. Further, the correlation calculating unit 36 calculates the autocorrelation coefficients ρ (4800) to ρ (6399), whereby the transmission signal array S (j) in the vibration wave transmitted by the fourth oscillator 1d and the corrected time-series data The time change of the autocorrelation coefficient ρ with Cd (j) can be calculated. As a result, the correlation calculation unit 36 calculates the autocorrelation coefficients ρ (0) to ρ (6399) according to Expressions (7-1), (7-2), and (4), thereby obtaining The time change of the autocorrelation coefficient ρ between the transmission signal array S (j) and the corrected time-series data Cd (j) in the vibration wave transmitted by each of the first to fourth oscillators 1a to 1d is calculated at once. be able to.

  FIG. 9 is a diagram illustrating a graph of a time change of the autocorrelation coefficient ρ calculated by the correlation calculator 36 according to the embodiment. As shown in FIG. 9, a steep peak is recognized in the time change of the autocorrelation coefficient ρ at the positions of T1, T2, T3, and T4 on the horizontal axis. This is a position corresponding to τ1 to τ4, which are the delay times of the vibration waves transmitted by the first oscillator 1a to the fourth oscillator 1d, respectively.

  In FIG. 9, T1 represents the delay time τ1 itself of the vibration wave transmitted by the first vibrator 1a. T2 is a time corresponding to 1600 data in the transmission signal array S (j), that is, a predetermined time T corresponding to a predetermined length D in the delay time τ2 of the vibration wave transmitted from the second oscillator 1b. = 10 [s]. Therefore, T2 = τ2 + T. At T3, a bias of 20 [s] corresponding to twice the predetermined length D is applied to the delay time τ3 of the vibration wave transmitted by the third oscillator 1c, and T3 = τ3 + 2T. Similarly, T4 = τ4 + 3T.

  The delay time calculation unit 37 analyzes the autocorrelation coefficient ρ, which is the data calculated by the correlation calculation unit 36, identifies a peak, and each of the oscillators 1 transmits a vibration wave before reaching the receiver 2. Calculate the delay time of each quake wave, which is the time required until. This makes it possible to estimate the velocity of each vibration wave in the medium M that has become the propagation path of the vibration wave transmitted from each of the exciters 1. The user of the signal processing system 100 can estimate the characteristics of the medium M through which the vibration wave has propagated, based on the estimated velocity of the vibration wave.

<Comparative example>
FIG. 10 is a diagram for explaining a method of calculating the autocorrelation coefficient according to the comparative example. In the method of calculating the autocorrelation coefficient according to the comparative example, a signal having a predetermined length D is cut out from the end of the synthetic vibration wave received by the receiver 2, and the signal is generated by adding the signal to the head of the synthetic vibration wave. This signal is not used for correlation calculation. Instead, the autocorrelation coefficient is calculated without correcting the synthetic vibration wave received by the receiver 2.

  As shown in FIG. 10, it is assumed that the vibration wave transmitted by the first vibration generator 1a arrives at the receiver 2 at the delay time τ1. In this case, a signal derived from the vibration wave transmitted by the first oscillator 1a arrives after a lapse of the delay time τ1 from the start of the generation of the reception waveform Sr1 by the time-series data generation unit 32.

  In the calculation method of the autocorrelation coefficient according to the comparative example, the correlation coefficient with the reception waveform Sr1 is calculated while simply shifting the transmission signal array S (j). In this case, as shown in FIG. 10, when transmission signal array S (j) is shifted by an amount corresponding to delay time τ1, signals constituting reception waveform Sr1 and transmission signal array S (j) are different. Matches. However, in other cases, the signal for which the transmission signal array S (j) calculates the correlation coefficient does not partially include the information of S (j), and thus the correlation coefficient is zero. Is not always the case. Therefore, in the temporal change of the calculated correlation coefficient, in addition to each peak indicating the delay time, a fine dummy peak may occur.

  FIG. 11 is a diagram illustrating a graph of a time change of the autocorrelation coefficient calculated by using the autocorrelation coefficient calculation method according to the comparative example. FIG. 11 shows that, in comparison with the time change graph of the autocorrelation coefficient ρ calculated by the correlation calculator 36 according to the embodiment shown in FIG. It indicates that.

  FIG. 12 is a graph showing an example of measured data of the autocorrelation function in the seismic survey. The signal propagating underground includes a plurality of peaks indicating a plurality of delay times reflecting the underground structure in the middle and including phase inversion. In FIG. 12, a minute peak appearing in an ellipse E1 indicated by a broken line is a so-called initial movement that is particularly important in seismic exploration, and is a peak indicating a vibration that reaches a receiver first. The peaks appearing in the ellipse E2 indicated by the broken line indicate that among the vibrations that arrive directly, the vibrations delayed by refraction or reflection at the impedance interface underground and the reflections appearing in the ellipse E3 are repeated a plurality of times. Followed by each peak indicating the signal reached.

  As described above, it is important to detect a small peak such as the initial motion of a seismic wave in a geological survey in the exploration of petroleum or the like, a ground survey in the civil engineering field, and a damage survey in an existing structure. In addition, in methods such as full wave inversion which are expected to be used in the field of seismic exploration, detection of minute peaks is more important. The calculation method of the autocorrelation coefficient according to the embodiment includes the initial movement, and also suppresses noise other than the peak indicating the delay time in the region where all peaks appear. The accuracy of peak detection can be increased as compared with the calculation method.

<Effects of Embodiment>
As described above, according to the signal processing system 100 according to the embodiment, it is possible to improve the discrimination accuracy when discriminating after simultaneously receiving signals transmitted from a plurality of shakers. In particular, the initial motion (the vibration that first arrives at the receiver from the transmitter), which is important in seismic exploration, is often much smaller than the subsequent vibration. In addition, it is possible to easily discriminate a small signal.

  As described above, the present invention has been described using the embodiments, but the technical scope of the present invention is not limited to the scope described in the above embodiments. It is apparent to those skilled in the art that various changes or improvements can be made to the above-described embodiment, for example, using the analysis in the frequency domain instead of the processing in the time domain to calculate the autocorrelation coefficient. It is. It is apparent from the description of the appended claims that embodiments with such changes or improvements can be included in the technical scope of the present invention.

<Modification>
In the above description, the case where the signal processing system 100 includes one receiver 2 has been described, but a plurality of receivers 2 may be provided. In seismic surveys, etc., it is common to make measurements at multiple points. Therefore, the measurement is performed by deploying the plurality of receivers 2 at different points. If the number of the receivers 2 is not enough, the receiver 2 is moved and the measurement is repeated. By using a plurality of receivers 2, the measurement time can be reduced.

  In the above description, specific numerical values such as the signal length L and the sampling intervals 100 [ms], 25 [ms], 6.25 [ms] have been described, but the present invention is not limited to these. These specific values may be appropriately determined by experiments in consideration of prior information of the medium M to be measured by the signal processing system 100, actual measurement results, and the like.

DESCRIPTION OF SYMBOLS 1 ... Exciter 2 ... Receiver 3 ... Signal processing apparatus 11 ... Pseudo random number generation part 12 ... Random number shift part 13 ... Vibration wave generation part 14 ... Clock generation part 15 ··· Vibration wave storage unit 31 ··· Receiving unit 32 ··· Time series data generation unit 33 ··· Vibration wave reproduction unit 34 ··· Clock generation unit 35 ··· Correction data generation unit 36 ··· Correlation Calculator 37 ... Delay time calculator 100 ... Signal processing system

Claims (10)

N (N is an integer of 1 or more) oscillators for transmitting a vibration wave to a medium;
A receiver for receiving a combined vibration wave transmitted by the N oscillators and transmitted through the medium;
A modified time series obtained by cutting out data of a predetermined length D from the end of the multiplexed time series data received by the receiver and adding it to the time series data from the beginning of the time series data to the predetermined length D. A correction data generation unit that generates data;
A correlation calculation unit that calculates data indicating a correlation between each of the time series data of the N vibration waves transmitted by each of the N oscillators and the corrected time series data;
Equipped with a,
The predetermined length D is a time period between the time when the maximum value of the delay time, which is the time required for each of the N oscillators to transmit the vibration wave and reach the receiver, elapses. the signal processing system of the length der wherein Rukoto seismic waves the focal unit is sent to. Each of the N oscillators transmits a vibration wave based on a pseudo-random signal that is uncorrelated with each other within the range of the predetermined length D.
The signal processing system according to claim 1, wherein: Each of the N oscillators transmits the vibration wave in synchronization with each other;
The receiver starts receiving the time-series data in synchronization with transmission of the vibration wave by each of the N oscillators.
The signal processing system according to claim 1 or 2, wherein: Each of the N vibrators outputs a pseudo-random signal having a period equal to or longer than the predetermined length D multiplied by N + 1 by M times the predetermined length (M is 0 or more and N-1 or less) , Different from each of said N generators), and transmits a vibration wave based on a signal obtained by cyclically shifting by a length.
The signal processing system according to claim 1, wherein: The correlation calculating section sequentially shifts the corrected time-series data from 0 to a length obtained by multiplying the predetermined length D by N, and sequentially calculates an autocorrelation coefficient with the pseudo-random signal. calculate,
The signal processing system according to claim 4, wherein: Data were analyzed by the correlation calculation unit has calculated, further comprising a delay time calculation section for calculating the N focal vessel their Re respectively the delay time of,
The signal processing system according to claim 1, wherein: Each of the N oscillators is configured such that the length of the vibration wave to be transmitted until the time of the maximum value of the delay time of each of the vibration waves elapses is equal to or less than the predetermined length D. Sending out a tremor wave,
The signal processing system according to claim 6, wherein: A signal processing method executed by a processor,
A step of transmitting a seismic Doha the medium in each of the N focal instrument (N is an integer of 1 or more),
And generating time-series data of the received and the received wave at the receiver a multiplexed of relaxin Doha before propagating through said medium,
Cutting out data of a predetermined length D from the end of the time-series data, and generating corrected time-series data added to the time-series data from the beginning of the time-series data to the predetermined length D;
Calculating data indicating a correlation between each of the transmitted time series data of the N vibration waves and the corrected time series data;
Including
The predetermined length D is equal to or less than the maximum value of the delay time, which is the time required from when each of the N vibration waves is transmitted to when the vibration waves reach the receiver. signal processing method but which is characterized in length der Rukoto seismic waves to be sent. For each of the N (N is an integer of 1 or more) oscillators that transmit a vibration wave to a medium , the receiver receives a combined wave of the vibration waves transmitted by the receiver and received by the receiver. A time-series data generation unit that generates series data,
A correction data generation unit that cuts out data of a predetermined length D from the end of the time-series data, and generates corrected time-series data added to the time-series data from the beginning of the time-series data to the predetermined length D When,
A correlation calculation unit that calculates data indicating a correlation between each of the time series data of the N vibration waves transmitted by each of the N oscillators and the corrected time series data;
Equipped with a,
The predetermined length D is a time required for the N number of oscillators to transmit the vibration wave and reach the receiver until the maximum value of the delay time elapses. signal processing apparatus according to claim length der Rukoto seismic waves the focal unit is sent to. On the computer,
For each of the N (N is an integer of 1 or more) oscillators that transmit a vibration wave to a medium , the receiver receives a combined wave of the vibration waves transmitted by the receiver and received by the receiver. The ability to generate series data,
A function of cutting out data of a predetermined length D from the end of the time-series data and generating corrected time-series data added to the time-series data from the beginning of the time-series data to the predetermined length D;
A function of calculating data indicating a correlation between each of the N time-series data of the N vibration waves transmitted by each of the N number of oscillators and the corrected time-series data;
To achieve,
The predetermined length D is a time required for the N number of oscillators to transmit the vibration wave and reach the receiver until the maximum value of the delay time elapses. program characterized lengths der Rukoto seismic waves the focal unit is sent to.