JP3164761B2 - Sound ray identification method for ocean acoustic tomography system - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a sound ray identification method for a marine acoustic tomography system using a computer.

[0002]

2. Description of the Related Art A marine acoustic tomography system includes a plurality of transmitters 11 anchored at an intermediate depth in a predetermined arrangement as shown in FIG. It is to estimate the state of the ocean within the predetermined area AR from the sound propagation times between the receivers 14 and 21 and the plurality of receivers 21 to 25. In practice, the transmitter / receiver is used as a transmitter or a receiver in a time-division manner.

[0003] Reference 1 “BDCornuelle,“ Simulation o
f Acoustic Tomography Array Performance With Untra
cked or Drifting Sources and Receivers '', JOURNAL
OFGEOPHYSICAL RESEARCH, VOL.90, NO.C5, pp.9079-9088
Each of the transducers 1i and 1j is, as shown in FIG.
It is positioned at an intermediate depth in the ocean by anchors 21i, 21j and bladders 31i, 31j. Floating bag 31i,
Each transducer 1i, via an electric cable via 31j,
Communication devices 51i and 51j connected to 1j are floating on the sea surface 4, and each communication device 51i and 51j communicates with an analysis center via an artificial satellite (not shown).

[0004] The ocean has the following properties with respect to sound waves. In the ocean, the higher the temperature, the faster the sound speed, and the higher the pressure (water pressure), the faster the sound speed.
Furthermore, the speed of sound also changes depending on the relationship (forward or reverse) between the flow of seawater and the direction of propagation of sound waves. Therefore, the sound speed distribution of the ocean (hereinafter, the term "sound speed" may also mean such a sound speed distribution) is not uniform. Sound waves, like light, refract at boundaries of different speeds of sound in the ocean.

[0005] Here, consider the case where one transducer 1i operates as a transducer and the other transducer 1j operates as a transducer. Sound waves radiated in all directions from the transmitter 1i propagate in the ocean and reach the receiver 1j, which is, for example, several hundred km away. Here, as shown in FIG. 3, the propagation path of the sound wave reaching the receiver 1j (referred to as a sound ray path if necessary) is a sound velocity c (x, y, z) that varies with time as shown in FIG. , t) field (sound speed field
), There are a plurality of paths Γ _k (k is 1 to K).

[0006] time the sound waves propagating along each path gamma _k reaches the receivers 1j is of course different because sound velocity distribution in the path gamma _k is the route by. That is, a plurality of arrival time data is obtained from each receiver 1j by one transmission, and a data set of these arrival times (accordingly, a data set of propagation time) is used for analysis. Note that, in practice, the time at which each transmitter 1i transmits a sound wave (for example, every hour or every 30 minutes) is determined, so that the arrival time is obtained, and the propagation time is also obtained. Means.

[0007] Measured transmitter 1i for path Γ _k
And the sound wave propagation time D _k (t) between the receiver 1j is given by the following (1)
Given by the formula.

[0008]

(Equation 1) Here, k is the number of sound paths, s is the length of arc along the acoustic ray path gamma _k. The second term E _k on the right side is a term that represents a change in the sound wave propagation time due to various causes. That is,
Fluctuation A due to clock skew between transmitter 1i and receiver 1j
_k (t), the variation B _k (t) due to the deviation of the transmitter 1i and the receiver 1j from the intended position, and the internal wave having the generation position in the area AR at the peak position (the position of the reception pulse). Peak; means the arrival time if not the wrong peak)
And the sum of the contributions of the various fluctuations to the propagation time D _k (t), such as the influence on the propagation time and the fluctuation ε _k (t) due to the uncertainty of the peak position itself.

If the sound speed field of the ocean changes, the sound wave propagation time changes. When these changes are small, the sound propagation time D _k (t) expressed by the above equation (1) is expressed in a linear form using the standard sound speed c ₀ (x, y, z). be able to. Here, the sound velocity c (x, y, z, t) at time t is represented by a standard sound velocity c ₀ (x, y, z) and a standard sound velocity c ₀ (x , y, z) and the amount of change c ′ (x, y, z, t).

C (x, y, z, t) = c ₀ (x, y, z) + c ′ (x, y, z, t) (2) Standard sound velocity c ₀ (x, y, z) As an example, an average value of sound velocities obtained from observation values of the same season in the past several years can be applied. In addition, a sound velocity obtained by a simulation in which values of various parameters such as the temperature of the target area AR and the position information of the transducer are set can be applied. That is, the standard sound speed c ₀ (x, y, z) is given as a set value.

The variation c '(x, y, z, t) in the equation (2) is sufficiently smaller than the standard sound speed c ₀ (x, y, z). here,
From the equations (1) and (2), the variation D'k of the sound wave propagation time can be expressed by the equation (3) if the first-order term of time is taken.
In equation (3), Γ ⁰ _k is a sound ray path in a standard sound velocity field state c ₀ (hereinafter, referred to as a standard sound ray path), and D ⁰ _k is a propagation time for the standard sound ray path Γ ⁰ _k . (Hereinafter, referred to as a standard propagation time), D ′ _k is a difference (a change amount c ′ (x, y, z, t)) from the standard propagation time D ⁰ _k of the observed propagation times D _k. Component). Note that the standard propagation time D ⁰ _k can be expressed by equation (4).

[0012]

(Equation 2) Since equation (3) is linear, a standard inverse problem analysis method can be used to estimate the sound velocity variation c ′ (x, y, z, t) from the data set {D ′ _k }. . If the sound velocity change c ′ (x, y, z, t) is obtained, the sound velocity change c ′ (x, y, z, t)
And the standard sound speed c ₀ (x, y, z) which is a set value from the outside
From this, the sound velocity (sound velocity distribution) c (x, y, z, t) at the current observation time t can be obtained according to the equation (2). That is,
The state of the ocean at the current observation time t can be captured.

As described above, in order to capture the state of the ocean, sound waves from a plurality of paths reach the receiver 1j, so that the measured sound wave propagation time { D _k (t) } ,
Ray path at standard sound velocity field state c ₀ ｛ Γ
⁰ _k } must be associated with the propagation time { D ⁰ _k } . That is, the standard sound propagation time { D ⁰ _k } (k is 1 to K) shown in FIG.
The measured sound wave propagation time { D _k (t) } shown in (b) does not coincide with the measured sound wave propagation time { D _k (t) } (shown as an arrival time sequence in FIG. 4), and the measured sound wave propagation time { D _k ( t) Since ｝ also includes information such as noise sound waves generated inside the observation area AR, it is necessary to associate (identify) these two types of propagation time information on the same route.

[0014]

However, in the conventional marine acoustic tomography system, the measured sound propagation time { D _k (t) } and the sound ray path {標準⁰ in the standard sound velocity field state c ₀ . Propagation time for _k ｝
The process of associating with { D ⁰ _k } is manually performed by the worker, and the burden on the worker is large.
There has been a problem that the work has a long time and a lot of time is required to obtain a final analysis result.

Since the observation is repeated at a predetermined cycle (for example, every hour), such a problem occurs every time and is a large problem.

[0016]

In order to solve this problem, in the sound ray identification method of the present invention, measurement is performed by transmitting and receiving a single sound wave between a transmitter and a receiver provided in the ocean. Each of a plurality of sound wave propagation time information having different propagation paths is identified as any sound wave propagation time information in a standard plurality of sound wave propagation time information prepared in advance,
Based on the result of this identification, the ocean acoustic tomography system that obtains ocean state information in this measurement by correcting the ocean state information related to the standard multiple sound wave propagation time information described above (1) A first process of associating past sound wave propagation time information which has been performed with measured sound wave propagation time information to be identified from now on by a tracking process, and (2) measurement based on the obtained matching result. The second sound wave propagation time information is identified as any one of a plurality of standard sound wave propagation time information.
And the processing of

[0017]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of a sound ray identification method in a marine acoustic tomography system according to the present invention will be described below in detail with reference to the drawings.

FIG. 5 is a block diagram showing a functional configuration of a main part of a computer system for realizing the sound ray identification method of this embodiment.

In FIG. 5, this computer system has four types of memories in which writing and reading of predetermined data are executed: an observation value data memory 101, a tracking parameter memory 102, and a line parameter memory 1
03, a line data memory 104, an arithmetic processing unit 120 for executing arithmetic processing, and a data input / output unit 110 for exchanging data with the outside of the system, which are mutually connected via a bus 130. I have.

FIG. 6 is a flowchart showing the overall processing of this embodiment, which is executed mainly by the arithmetic processing unit 120.

In the processing of this embodiment, the arithmetic processing unit 12
0, after executing the setting of the initial value (step 201) until the end of the processing is instructed from an internal or external, uptake of observation data (step 202), the tracking processing (step 203), tracking results The output (step 204) is repeatedly executed.

In the following, the definitions of the main terms will be described with reference to FIG. 7 showing a tracking state described later, and then each step processing 201, 202, 203, 2 shown in FIG.
04 will be described in order.

A shot is one of the sound waves from the transmitter 1i.
Times, and the shot number (..., n-2, n-1,
n, n + 1, n + 2,...; serial numbers)
Means that is attached to each set of received data in the receiver 1j corresponding to one transmission of the sound wave from the receiver. Also, the shot time means that a sound wave related to a set of received data is
It is assumed that time T when i makes a call. Here, the time interval between successive shots is constant.

As the reception data, the transmission time T of the sound wave in the transmitter 1i is set to time 0, and the elapsed time (corresponding to the propagation time) from this time and the change in the intensity of the reception signal in the receiver 1j are recorded. It was done. As described above, the sound wave propagates along a plurality of sound ray paths determined by the positional relationship between the transmitter 1i and the receiver 1j and the speed of sound in the sea, so that in tomographic observation, the intensity of the received signal is included in the received signal. Are found in several peaks. These peaks include
It is assumed that a series of numbers is assigned to each set of received data. Here, an example will be described in which, for example, 50 peaks are selected in descending order of intensity from a set of received data peaks and each of the peaks is used as a shot peak. The number on the time axis of each shot in FIG. 7 indicates the peak number of the shot by its time position.

A line is formed by connecting peaks associated between different shots with the nodes as nodes. In FIG. 7, a certain shot n
In some cases, a line with no peak associated with +1 may occur.

Basically, the identification means that one of the standard sound propagation times is replaced by one of the sound propagation times at the peak of the signal strength in the measured reception data between the transmitter and the receiver. It refers to the process of associating one. In this embodiment,
Since the concept of a line is provided, after each of the standard sound wave propagation times is associated with each line, the sound wave propagation at the peak of the signal strength in the measured reception data between the transducers is performed. By associating each time with each line, it is possible to obtain the same result as the result of the identification processing.

The main terms have been described above, but other terms will be clarified in the following description.

(A) Initial Value Setting (Step 201) The arithmetic processing unit 120 uses the data input / output unit 110 to
Initial values of various data and parameters are set in the memories 101, 102, 103, and 104.

In the observation data memory 101, the number of the shot for which the identification process has been executed, the time of this shot, the number of the peak of this shot, and the observation time of these peaks are written. In the line data memory 104, the peak number of the same shot as that in which the data was stored in the observation data memory 101 is written as the number of the first node of each line. In the tracking parameter memory 102,
Threshold value t used in tracking processing described in detail later
_{Write ass} and weighting coefficients α and β. In the line parameter memory 103, a smoothed value and an initial value of a change rate of each line, which are updated for each shot in a tracking process described later in detail, are written.

In this embodiment, once the operator performs the sound ray identification with the standard sound wave propagation time for the measurement data of a certain shot, the sound ray identification processing for the measurement data of the subsequent shots is performed. It is intended to be performed automatically by the computer system. Observation data memory 101
The values initially set in the line data memory 104 are data relating to shots that have been manually subjected to sound ray identification processing by such an operator.

(B) Acquisition of observation data (step 20)
2) Each time a new set of received data (shot) is observed,
The arithmetic processing unit 120 uses the data input / output unit 110 to
The number of this shot, the time of this shot, and a predetermined number included in this shot (for example, the above-mentioned 50)
The peak number and the observation time of each peak are written into the observation data memory 101 as shot information to be subjected to tracking processing (line integration processing as described later).

(C) Tracking processing (step 20)
3) The tracking process is a process of associating each line with any peak in the current shot. FIG. 1 is a flowchart showing details of the tracking processing. Steps 301 to 30 in FIG.
The series of processing of 7 is repeated by the number of lines.

(C-1) The line L currently targeted
For L (one of 1 to 50), a predicted value t _{e, L} of a peak time at which a peak appears in the current shot is obtained (step 301).

In this process, first, the arithmetic processing unit 120
A shot time T ⁿ (n is a shot number) to be tracked from the observation data memory 101 and a shot time T ^m last integrated with the line L from this shot
(M is a shot number, m <n). For example, one of the peaks of the shot at the immediately preceding time is line L
M = n−1 when the integration is performed. Also,
The arithmetic processing unit 120 reads the smoothed value t of the time of the line L to be tracked from the line parameter memory 103.
_{s, L} and the rate of change Δt _L are read out. Here, the smoothed value t _{s, L} and the change rate Δt _L of the line L are calculated in step 30 described later.
6, 307 as appropriate, the former means time information obtained by smoothing the fluctuation in each shot of the line L, and the latter indicates the latest rate of change of the time fluctuation per unit time. Means. From the various data read out as described above, the arithmetic processing unit 120 obtains a predicted value t _{e, L} of the line L at a time when a peak will appear in the current shot according to Expression (5).

[0035] is a _{t e, L = t s,} L + (T n -T m) Δt L ... (5) ( c -2) each observed value of the shot time T ⁿ t _p ⁿ _(p is the peak number, 1 to 50) and
The difference (precisely, the absolute value of the difference) d from the predicted value t _{e, L} of the peak time for the line L obtained in the process of step 301
_p is obtained (step 302).

In this process, first, the arithmetic processing unit 120
Observation data memory 101 reads the time observations of all peaks in the shot tracking processed {t _p ^n}. Note that {x} represents a set having x as an element. Then, the arithmetic processing unit 120 includes a time observations of all peaks readout {t _p ^n}, estimated value t _e of the peak time for the line _L, from the _L, and in accordance with (6), the set of difference {d _{Find p} ｝.

[0037] _{{d p} = {| {} t p n} -t e, L |} ... (6) ( c -3) 50 differences determined in step 302 {d _p}
= {D ₁ , d ₂ ,... D ₅₀ } minimum value _{P P} (P is a peak number from 1 to 50) (step 30)
3). That is, the arithmetic processing unit 120 executes the minimum value selection processing shown in Expression (7). Incidentally, the minimum value selection processing from among the 50 observed peak time t ₁ ⁿ ~t ₅₀ ^n,
Peak time t at which the difference absolute value from the predicted value t _{e, L} is the smallest
_This means that _P ⁿ has been found.

D _P = [{d _p }] _min (7) (C-4) Compare the minimum value D _P obtained in step 303 with the threshold value t _ass (step 304).

In this process, first, the arithmetic processing unit 120
Step 2 from the tracking parameter memory 102
The threshold value t _ass stored in the initial value setting process of 01 is read. Then, the minimum value D _P as shown in equation (8) to determine whether it is less than the threshold value t _ass. This determination is equivalent to determining whether or not the peak with the smallest prediction error (the peak with the difference D _P ) can be integrated at the end of the current line L. I do.

D _P ≦ t _ass (8) In the case where the expression (8) is not satisfied, even the minimum value D _P of the difference (prediction error) is larger than the threshold value t _ass , and is therefore the current target. Since there is no peak suitable for merging with the line L currently being processed among the peaks of the shots, the process returns to step 301 without executing the merging process and returns to the next line (L = L + 1). This is a new processing target.

In the processing of the last line (50), when the result that the expression (8) is not satisfied is obtained, the processing shifts to the output processing of the tracking result in step 204.

[0042] When (C -5) results in satisfying the above-mentioned equation (8) is obtained, the peak number P of the minimum value D _P of (prediction errors) of the difference, the current processing target line L Integration (Step 305).

That is, in the line data memory 104,
Node N _{L at} the n-th shot time T ⁿ of line _L
as the value of ^n, writing the peak number P of the minimum value D _P of (prediction errors) of the difference. It should be _noted that the result of the determination in step 304 that the value is larger than the threshold value t _ass may have been obtained in the past, so that the number of nodes of the line L obtained so far is not necessarily equal to n.

(C-6) When the latest value of the node _NL ⁿ for the line L is obtained, a new smoothed value t _{s, L} is obtained for searching for the next node (step 306).

In this process, first, the arithmetic processing unit 120
Step 2 from the tracking parameter memory 102
The weighting coefficient α stored in the initial value setting process of 01 is read out. Then, the arithmetic processing unit 120 reads the weighting coefficient α (0 ≦ α ≦ 1), the predicted value te _{, L} obtained in step 301 described above, and the line number written in the line data memory 104 in step 305. According to the equation (9) using the observation value t _P ⁿ having P, a new smoothed value t
_{Find s and L} and write them to the line parameter memory 103
Put in .

T _{s, L} = αt _P ⁿ + (1−α) te _{, L} (9) This equation (9) represents a predicted value t _{e, L} reflecting the past information of the line L, The observation value t _P ⁿ which is the current information of the line L
Are weighted and added to obtain a new smoothed value t _{s, L.}

(C-7) In order to search for the next node on the line L, the rate of change Δt _L is updated (step 30).
7).

In this process, first, the arithmetic processing unit 120
Step 2 from the tracking parameter memory 102
The weighting coefficient β stored in the initial value setting process of 01 is read out. Then, the arithmetic processing unit 120 calculates the read weighting coefficient β and the shot time T ^{n of the} current processing target.
When a shot time ^{T m} of the last node integration into the line L in the shot time _{T n} before is performed, and the rate of change Delta] t _{L old} ever, the minimum value _D P of the difference obtained in step 303 ( In the above description, the absolute value is used. However, a value with a sign is used here. Therefore, a new change rate Δt _{L new} is obtained according to the equation (10), and is written in the line parameter memory 103.

[0049] _{Δt L new = Δt L old +} βD P / (T n -T m) ... (10) This equation (10), the minimum value _{D P} of the current and immediately preceding time between nodes of the difference _{(T n} - T ^m ), this time (T ^m )
_(n− T ^m ) is obtained, and this change rate is reflected on the change rate Δt _{L old} up to now with a predetermined weight to obtain a new change rate Δt _Lnew .

As described above, when the integration processing for the line L currently being processed is completed, the process returns to step 301, and the next line (L = L + 1) is set as a new processing target. On the other hand, when the integration processing for the last line (50) is completed, the process proceeds to the output processing of the tracking result in step 204.

(D) Output of Tracking Result (Step 204) As described above, the tracking process for the currently processed shot (shot number n) (Step 20)
3 ; When steps 301 to 307) are completed, the arithmetic processing unit 120 reads out the peak number information integrated into each line for this shot from the line data memory 104, and externally outputs the information via the data input / output unit 110. Output to

For example, the information is presented to the worker by a display device and used for confirmation of the worker. Here, regardless of whether it is accepted or the correction is executed, when the tracking result of the current shot is determined by this check, the peak at the standard propagation time and the current shot Are associated (identified) with the peaks in the observed data. Then, as described in the section of the related art, the fluctuation D′ k of the sound wave propagation time is obtained from the time difference between these peaks, and then the sound velocity change amount c ′ (x, y, z, t) is obtained. , The sound velocity change amount c ′ (x, y, z, t) and the standard sound velocity c which is a set value from the outside.
₀ (x, y, z) and the sound velocity (sound velocity distribution;
The state of the ocean) c (x, y, z, t) is obtained.

When the operator corrects the tracking result, the series of sound ray identification methods shown in FIG. 6 are restarted from the initial value data setting processing in step 201 including the corrected contents. Will be.

As described above, according to the above embodiment, the peak in the past received data already identified as the peak related to the standard sound wave propagation time, the peak in the latest unidentified received data, Are associated with each other by a tracking process to identify peaks in the latest received data that have not been identified as peaks related to the standard sound wave propagation time. It is not necessary to perform the identification process. That is, basically, if the operator executes the identification process only once, the computer system can automatically execute the execution process on the received data in each subsequent shot.

Therefore, the burden on the operator can be reduced, and the identification process for each shot can be performed at a high speed, and the output of the ocean state for each shot can be quickly performed.

Here, even if the result of the automatic identification processing in the computer system is incorrect, the operator partially corrects the presented contents and restarts the identification processing from the initial setting operation, whereby the computer system is re-started. Can continue to be able to execute the sound ray identification processing satisfactorily.

In the above embodiment, the case where the number of peaks in one set of received data is 50 has been described.
The present invention is, of course, not limited to this. For example, if the amount of observation data memory for writing the peak number of each shot and the observation time of these peaks is increased, a marine acoustic tomography system that can handle more peaks can be realized.

In the above embodiment, the peaks in one set of received data are handled collectively. However, the peaks are divided into a plurality of sets according to the direction of arrival at the receiver, and each set is described above. A sound ray identification method may be applied. For example, if beamforming processing is performed on data obtained by arranging receiver sensors in an array, it is possible to realize a system that separates sound waves up and down according to the direction of arrival at the receiver. Can be.

[0059]

As described above, according to the present invention, a plurality of sound wave propagation times having different propagation paths measured by one sound wave transmission and reception between a transmitter and a receiver provided in the ocean are described. Each of the information is identified as any of the sound wave propagation time information in the standard sound wave propagation time information prepared in advance, and based on the identification result, the above-described standard sound wave propagation time information is used. In the marine acoustic tomography system which obtains the marine state information in the current measurement by correcting the marine state information, the past sound wave propagation time information already identified and the measured sound wave propagation time to be identified from now on A first process of associating the information with the tracking process, and, based on the obtained association result, the measured sound wave propagation time information in any of a plurality of standard sound wave propagation time information. Second identifying a sound wave propagation time information
Since it is possible to eliminate the need for the operator to perform the identification process manually every transmission from the transmitter, the burden on the operator can be reduced, and the identification process for each transmission becomes faster. The output of the ocean condition for each transmission can also be quickly performed.

[Brief description of the drawings]

FIG. 1 is a detailed flowchart of a tracking process of a sound ray identification method according to an embodiment.

FIG. 2 is an explanatory diagram of a marine acoustic tomography system (1).
It is.

FIG. 3 is an explanatory diagram of a marine acoustic tomography system (2).
It is.

FIG. 4 is an explanatory diagram of sound ray identification.

FIG. 5 is a block diagram illustrating an execution configuration of a sound ray identification method according to the embodiment.

FIG. 6 is a flowchart of an overall process of a sound ray identification method according to the embodiment.

FIG. 7 is an explanatory diagram showing a tracking state in the sound ray identification method according to the embodiment.

[Explanation of symbols]

101: observation data memory, 102: tracking parameter memory, 103: line parameter memory, 10
4 line data memory 110 data input / output unit 1
20: arithmetic processing unit, 130: bus.

────────────────────────────────────────────────── ─── Continuation of front page (56) References JP-A-8-297135 (JP, A) JP-A-9-190592 (JP, A) JP-A-8-304538 (JP, A) 20506 (JP, B2) (58) Field surveyed (Int. Cl. ⁷ , DB name) G01S ⁷ /52-7/64 G01S 15/00-15/96 G01S 3/80-3/86

Claims (2)

(57) [Claims] 1. A plurality of sound wave propagation time information having different propagation paths measured by one transmission and reception of a sound wave between a transmitter and a receiver provided in the ocean are each prepared in a standard Identified as any one of the sound wave propagation time information in the multiple sound wave propagation time information, based on the identification result,
In the ocean acoustic tomography system that obtains ocean state information in the current measurement by correcting the ocean state information related to the standard plurality of acoustic wave propagation time information described above, the past acoustic wave propagation time information already identified A first process for associating the measured sound wave propagation time information to be identified from now on by a tracking process, and, based on the obtained association result, converting the measured sound wave propagation time information into standard plural sound waves. A second process of identifying any of the sound wave propagation time information in the propagation time information. 2. The method according to claim 1, wherein the first and second processes are performed based on past sound wave propagation time information already identified with the same standard sound wave propagation time information. Processing to obtain a predicted value of the sound wave propagation time information at the time of the present sound wave transmission / reception, and processing of searching for the observed sound wave propagation time information having the smallest difference from the predicted value of the sound wave propagation time information. A process of determining whether the difference between the observed sound wave propagation time information and the predicted value of the sound wave propagation time information is perceived as pertaining to the standard sound wave propagation time information, and And a process of identifying the observed sound propagation time information having the smallest difference from the predicted value of the sound propagation time information as standard sound propagation time information related to the formation of the predicted value. Ocean acoustics according to item 1 Sound line identification method of mode graphics system.