JPS63284481A - Passive ranging system - Google Patents

Abstract

PURPOSE:To reduce a calculation quantity without lowering position measurement accuracy by dividing the reception band of noises from a moving target into two, selecting their center frequencies and band widths under specific conditions, and performing arithmetic operations. CONSTITUTION:Radiation noises of the moving target which are received by receivers 11-13 are limited by band-limiting filters 21 and 22 to proper reception bands. Here, the center frequencies fC and f'C and band widths W and W' of the filters 21 and 22 are so selected that an inequality I holds. A signal of a 1st reception band is inputted to the circuit composed of mutual correlator 241 and 242, maximum point detectors 261 and 262, interpolators 281 and 282, and maximum point detectors 301 and 302, and a 1st estimated value of arrival time difference is obtained through the detectors 301 and 302. The mutual correlators 231 and 232 calculates a mutual function from a 2nd reception band within a range of delay time difference centered on the arrival time difference. Then the maximum value detectors 291 and 292 obtains time difference which maximizes the mutual function and a target position calculator 8 calculates the position of the moving target from said value.

Description

DETAILED DESCRIPTION OF THE INVENTION (Industrial Application Field) The present invention uses a plurality of receivers to receive noise emitted from a target moving in seawater, etc. The present invention relates to a passive ranging system that estimates the position of a moving target by estimating the arrival time difference.

(Prior art) Conventionally, this type of system is described in Reference 1 "A, H, Quazi: Effects of
Signal and Noise 5pectral
5lopes on Time Delay Est
imationin Pa5sive Localiz
ation”, U, S, Government VVor
kNUSO, J. and.

Reference 2 “G,O,0arter,”Pa5sive
Ranging Errors due to Rec
eiving Hydrophone Po5iti
on”.

J, Acoust Soc, Am, Vol, 62. N
o, 2.1979. For the simplest case where the number of receivers N = 3, there is a method disclosed by JK, and Figures 3 to 5
As shown in the figure.

In FIG. 3, the radiation noise of the moving target received by the receivers 1□, 1□, 13 is amplified to an appropriate level by the amplifier '1+22+23, and then the band-limiting filter 31
, 3□, 33, the signal 5 is limited to an appropriate reception band.
A cross-correlation function between t(t), 52(t), and 5a(t) is calculated by cross-correlators 41 and 4□. In the example shown in FIG. 1, 81 (t) and 82 (
The cross-correlation function 几1,2(τm) of
By □, the cross-correlation function R of 5a(O and 82 (t)
We are looking for 3,2(τm). Here, τ□ indicates a discrete time difference obtained by discretizing the time difference region τ with a step width Δτ.

The first maximum point detectors 5□, 5□ find the maximum point in the region τ□ of the cross-correlation functions ⇠□, 2(τm), Rs, z(τm), and determine the maximum point? □, □, 2 and? □, 3.
Output as 2.

The interpolator 61+61 outputs the first maximum point detector 51+52? □, 1.2 and ? □211. The cross-correlation functions 几1,2(τ) and 几1,2(τ) in the vicinity of are expressed as the cross-correlation functions R+,x(Tm
) and R3,2(Tm) by interpolation.

The second maximum point detectors 7□, 7□ detect the cross-correlation function 1
, 2(τ) and Rs, z(τ) in the area τ, and determine the maximum point as ? □、2+? Output as 3.2. The target position calculator 8 calculates ↑t, x l ? based on the principle of the calculation formula shown in Reference 2, for example. 312 + said receiver 1□, 1□.

Using the array interval of 13 and the propagation speed of the signal, an estimated value of the moving target position coordinates X is calculated and outputted to the output terminal 9.

FIG. 4 shows a first detailed implementation example of the cross-correlator 4□ in FIG.
is a fixed delay device, 12 is a variable delay device, 13 is a multiplier, 14
is an integrator, and 15 is an output terminal.

4 in Figure 3 as a cross-correlator. Assuming that, the band-limited signal 5t (
O and s z (t) are input. The fixed delay device 1 applies a fixed time delay amount τ to the signal 52(t) inputted from the input terminal 101. The variable delay unit 12 applies a variable time delay amount τ to the signal S□(1) input from the input terminal 102. +τ□
give. The delayed signal S, (t-τ.-τm)
and 82(t-τ.) are multiplied by a multiplier 13 and integrated by an integrator 14, and then the cross-correlation function R,z(Tm) is sent to the output terminal as the output of the cross-correlator 4□. 15.

FIG. 4 shows a second detailed implementation example of the cross-correlator 4□ in FIG. 1, 161, 16 . is each AD converter (AD
O), 77□, 172 are each digital Fourier transformer (
18 is a multiplier, 19 is an inverse digital Fourier transform (IDFT), and 20 is an accumulator.

The AD converter 161.16□ converts the band-limited signal 51
(t) and S 2 (t) are sampled at time intervals T5 and converted into digital signals S(tk) and 5t(tk). The digital Fourier transformer 171 transforms the digital signal into time series signals S, (tk): k=1
,..., digital Fourier transform value X x (
fk): k = 1 +..., K is calculated, and the digital Fourier transformer 17□ calculates 52(tk): k
=1,...K digital Fourier transform value X2(f
k): to=1. ..., calculate K.

The multiplier 18 calculates the Fourier transform values X5(fk) and xz(
A) Product Yt, 2(fk) = xscfk)・XxCf
k): k=x.

..., calculate K, and inverse digital Fourier transformer 19
is the signal YCfk): k=1, . Output to output terminal 15.

(Problems to be Solved by the Invention) In the method shown in FIGS. 3 to 5, the interpolators 68 and 6 in FIG.
In order to estimate the true maximum point of the cross-correlation function by an interpolation operation in step 2, the cross-correlator 4. .

Cross-correlation function R14 (τln) + 几1 calculated by 4□
, 2(τm) because there is a constraint that the step width Δτ of τ□ must be selected so that the interpolation theorem holds.
The center frequency of the band-limiting filter is f. , the bandwidth is W. Therefore, if (1) there is no prior information regarding the position or orientation of the moving target, (2) the arrangement interval of the receivers is long, and the maximum value of the arrival time difference between the receiver signals is large, the cross-correlation function 41,
4□ has the drawback that the number of cross-correlation function points that need to be calculated is maximized and the amount of processing required increases.

The above problems will be explained in detail. Figure 6 is a diagram showing the relationship between the receiver arrangement spacing d1□, d3□ and the arrival time difference of 1□, τ, 2 of the radiation noise from the target when the number of receivers N is 3. p, 20 indicate the moving target, x, y are the orthogonal coordinate system selected such that the origin is placed at the receiver 1□, and the Y axis passes through the receiver 1□, and θa is the coordinate of the target 20 with respect to the Y axis. The direction cosine angle, r, is the distance of the target 20 from the receiver 12.

According to Fig. 6, when the distance r is sufficiently larger than the receiver array spacing d1□, the propagation speed of the signal is C and l and 1,2 to d.
...°″′ey, therefore, when there is no prior information regarding the position or direction of the moving target 20,
□=-! It is necessary to calculate the cross-correlation function assuming that the cross-correlation function exists in any of CC between . The same thing can be said about τ3 and τ2, so if there is no incident information regarding the position or direction of the moving target 20, the points that need to be calculated by the cross-correlators 41 and 42 are as follows.

However, λ- indicates the wavelength of the maximum frequency fc in the receiving frequency band.

Therefore, when do,2 and d3,2 are large compared to the wavelength λ, the amount of processing required increases as M and M2 become large. For example, fc=5.5 kHz, W=
10 kHz, C=1500 m/s (sound speed in seawater)
, d, , 2=d3, 2=2.000 m, then it is of the order of.

On the other hand, as shown in the above-mentioned literature, the accuracy of estimating the position of the target 20 in passive ranging increases as the fc and W become larger and as the receiver array spacing becomes larger. Therefore, the higher the accuracy is sought in passive ranging, the more the number of calculation points increases.

Therefore, this invention has the following advantages: ■ There is no prior information regarding the position or direction of the moving target;
It is necessary to assume that there are targets in all directions, and the number of required calculation points in the cross-correlator increases, which becomes a problem when the spacing between the receiver arrays is larger than the wavelength 'min of the maximum frequency fmax of the reception band. remove the problem of
It is an object of the present invention to provide a wide ranging method that can reduce the increase in the number of calculation points without reducing position estimation accuracy even when the reception bandwidth increases and the spacing between receiver arrays increases. .

(Means for Solving the Problems) This invention receives noise radiated from a moving target using N receivers, and calculates the arrival time difference of the radiated noise at the receivers. Estimated value of t, j (t<Nt j<N, trj); i
In a passive ranging system that calculates IJ and estimates the position of the moving target from f,, j, the reception band in the receiver is divided into a first reception band whose center frequency is fc and whose bandwidth is w', and the center the second with frequency fc and bandwidth W
, and divide f′c and W′ into fIc+W′/
2<fc+W/2, calculate the cross-correlation functions R1s j (τm) and R1,j(τ) between the N signals in the first reception band, and maximize the R15j(τ). Let τ be the first estimated value 71 t J of the arrival time difference τitj of the radiation noise, and in the vicinity of τi,j, the cross-correlation function Ri,j between the N signals of the second reception band
(Hori) and R153(τ), and calculate the R15j(τ)
The second estimated value "isi" of the arrival time difference of the radiation noise is set to τ that maximizes the time difference, and the position of the moving target is calculated using the T 1 and j.

(Function) ■ By selecting the step width of the τ region of the cross-correlation function R'1 z j (τd) as Δτ'≧M>lr, "1
In addition to reducing the number of calculation points for t j (τ), ■ the calculation range of the τ region of the cross-correlation function RLj (τm) is reduced to 7.
By restricting to the vicinity of 1 t j, R1,j (
By reducing the number of calculation points of "m), the total number of calculation points of the cross-correlation function can be reduced.

(Embodiment) FIG. 1 is a functional block diagram showing an embodiment of the present invention.
,! l) PA21 ・'212 and 213 are each band-limiting filter B, 227, 22□, 223 are each band-limiting filter A1231.23□ are each cross-correlator B, 241.24□
is each cross-correlator A, 251.25□ is each first maximum point detector B, 261, 262 is each first maximum point detector A, 2
74, 27□ are each interpolator B, 284, 282 are each interpolator A, 291.29□ are each second maximum point detector B1304
, 302 are each second maximum detector A.

Band-limiting filter B214, 212°21 of this embodiment
3 and the band-limiting filters A221.22□, 223, as shown in FIG. 2, fc +W'/2 < 1c+
Choose something like w/2. That is, band-limiting filter A2
21, 22□.

223 is the signal s,'(t
).

Output s/(ri S3'(ri)
11°21□, 213 is the signal S of the second reception bandwidth W,
(vinegar.

52(t) and 53(t) are output.

The cross-correlators A241, 242 detect signals s,'(t) and s2'(t), s3'(t) and s of the first reception band 1.
1) The cross-correlation functions R1', 2 (1g) and R5', 2 (τX) are calculated using a cross-correlator for 2'(t). In order to make the interpolation theorem hold in the interpolation operations of the interpolators 281 and 28□, the increment width Δτ' of τ' is therefore given as follows: It will be done.

However, h indicates the wavelength C/f she of the maximum frequency f' of the first reception band.

ax The first maximum point detectors A264, 26□ are similar to the first maximum point detectors 51, 52 in FIG.
τ(τ) and R;, the value of τ(τ) where 2(τ4) takes the maximum value,
Find τM, 1, 2 and -23, 2, and interpolator 2B1, 2
g□ is the same as the interpolators 61 and 6□ in FIG.
, 1, 2, and 7', calculate the interpolated values R'1, 2 (τ) and 5.2 (τ) of However, the second maximum point detector A 301, 30□ is similar to the second maximum point detector 71.7□ in FIG.
1,2(τ) and R3,2(τ) take the maximum values, τ values τ1,2 and τ3,2 are determined.

The cross-correlators B234, 23□ calculate the arrival time difference τ1, 2
The cross-correlation functions R1,2(τm) and R5,2Cτm) are calculated within the range of the delay time width T centered on and 73,2. That is, the calculated score M4 of the cross-correlation function. M
2 is approximately T.yo7(7) M2″″−77. The time width T is at least τ1, 2 and τ3.
It is necessary to select a value larger than the estimated error σ (71, 2) and σC; '3.2) of 2 plus the order of the interpolation function used for interpolation calculation.

The first maximum point detector B251, 25□
(τm) and R3,2(τm) are maximum [-tor T. value,
Determine τm, 1.2 and τ, and use the interpolator 27. 272
is m, 3.2
1 In the vicinity of m, 1.2 and τ□, 3.2, R
Interpolated values R1,2 of 4,2(τm) and R5,2(τ□)
(τ) and R3,2(τ), and the second maximum point detector 291,29□ calculates R1,2(τ) and R3,2(
τ value at which τ) takes the maximum value, ? 1, 2 and? 3 and 2 are calculated, and the estimated values are calculated by the target position calculator 8, and the estimated value of the position coordinates of the moving target 20 is calculated and outputted to the output terminal 9 as in FIG.

According to this embodiment, for example, f, = 5.5 kHz
, W =10kHz, d,2=d2. ,=2,00
0m, when C=1,500m/sec, fc'
= 1 kHz, W' = l kHz, then
M,' = M7, p 4 x μ ball = 8 x 1
The order is 03□ (9). Moreover, if the order of T=10
2 is Δτ2Δτ′/1o in this example, but M1=
It is on the order of M2z100 αQ.

Therefore, the total calculation score of the cross-correlation function is approximately 1.
If it is on the order of 6X10' or more, the order of the number of points calculated by the conventional method is from the above formula (3), IOX 10'
Therefore, according to this embodiment, approximately one-sixth the number of calculation points is required.

(Effects of the Invention) As described above in detail, according to the present invention, the reception band is divided into a first reception band whose center frequency is f/ and whose bandwidth is τ', and a first reception band whose center frequency is fc and whose bandwidth is is divided into W's second reception band, and t = fc' + W'/2 is divided into f =
f+W/2 By choosing a smaller value, max
c. First, calculate the cross-correlation function using the signal in the first reception band, and obtain a loose estimate of the arrival time difference of the radiation noise. 1. seek,
next,? Since the cross-correlation function is calculated using the signals of the 21st and K reception bands only in the vicinity of '□, j, and a detailed estimate τl+3 is obtained, there is no prior information about the position or direction of the moving target. Both methods have the advantage that the amount of calculation can be significantly reduced without reducing the accuracy of position estimation.

As shown in the above explanation, the present invention applies when the ratio d/λmi□ of the spacing d between the receiver array and the wavelength λmi□ of the maximum frequency fmslx in the receiving frequency band is large, that is,
It is more effective when it is necessary to realize high-precision passive ranging, for example, fc = 5.5 kHz, W
= 10kHz 1d = 2,000m, fc
' = 1 kHz.

If W' = l kHz, the amount of calculation can be reduced to about one-sixth or more.

[Brief explanation of drawings]

FIG. 1 is a functional block diagram showing an embodiment of the present invention, FIG. 2 is an explanatory diagram of characteristics of band-limiting filters A and B in FIG. 1, FIGS. 3 to 5 are explanatory diagrams of the prior art, The figure is an explanatory diagram of the problem to be solved. 21. 21□, 213 are respective band-limiting filters B; 221.22□, 225 are each band-limiting filter A, 23
1°23° is each cross-correlator B, 241.24□ is each cross-correlator A, 251, 252 is 6th maximum point detector B,
261, 26□ are 6 first maximum point detectors A, 27. . 272 is each interpolator B, 281, 2g□ is each interpolator A,
29, j292 is 6 second maximum point detector B, 30. . 30° is the 6th second maximum point detector A. 51(t), 52(t), s3(su... Received signal band-limited by band-limiting filter B in the receiving band, S1'(t), S2'(t), 83'(su...
・Received signal band-limited by band-limiting filter A in the reception band, τ, τ□, τ... time difference, R(τ), R(τ) * R1', 2(1m)''3.
2 (1g)...Phase 1.2m 3.2
m cross-correlation function, ``i?,?', τ'... cross-correlation function rrl+1.2' mr 3,2
τ for which m, 1 + 2 mr j 2 takes the maximum value. , τ value, R1,2(τ), R3,2(τ), R4',2(τ),
R3', 2 (S... Interpolated value near the maximum value, τ1,21 5.2 '♀1,2.τ3,2... Value of τ at which the interpolated value takes the maximum value. Patent applicant: Oki Electric Kogyo Co., Ltd. 2!E, 7 Concave r\yk6.4'l Pll A) v7
fl'F%'A2 Figure 2-5 Shellfish fishing 5 years ago Figure 6 Procedural amendment (Mutsu) 1. Indication of the case 1988 Patent Application No. 118892 2. Name of the invention Passive Ranging System 3, Relationship with the case of the person making the amendment Patent application Person address (
105) 1-7-12-4 Toranomon, Minato-ku, Tokyo.
Agent address (105) 1-7-1 Toranomon, Minato-ku, Tokyo
No. 2 No. 6, Contents of amendment (1) The "Claims" column of the specification is amended as shown in the attached sheet. (2) On page 5, line 7 of the same book, the phrase ``Figure 4'' is corrected to ``Figure 5 is''. (3) In the fourth line of page 7 of the same book, the phrase ``scores are maximized,'' is amended to ``scores increase,''. (4) Formulas (5) and (6) on page 12 of the same book are corrected as follows. (5) In the 8th line of page 16 of the Ministry, the phrase ``The amount of calculation is'' will be corrected to ``The amount of calculation.'' (6) Delete the words ``receiving band'' in the third and fifth lines of page 17 of the same book. (7) Same book, same page, line 10: r Tmr 1e 2'
At Tm+5+2' □11+2. τm+3+2' is r? mg 1
t2'Tm+5+2'?jn1112'
? Correct it as '17115@2 j. (8) Amend the drawing “Figure 6” as shown in the attached sheet. Claims: Noise 'k radiated from a moving target, received by N receivers, arrival time difference τi of the radiated noise at the receivers,
Find the estimated value τi,j of j mouth≦N, j≦N, i#j),
The estimated value τ3. In a Pudgebrenson system that estimates the position 1+ of the moving target from: a)
dividing the reception band in the receiver into a first reception band with a center frequency of fcl and a bandwidth of W' and a second reception band with a center frequency of fc and a bandwidth of W; Let the center frequency f' and the bandwidth W' be f'+W
/2 (fc+W/2), C) Calculate the cross-correlation function between the N signals in the first reception band, and calculate the time difference τ that maximizes the cross-correlation function as the arrival time difference τ1. ．． d) Calculate the cross-correlation function between the N signals in the second reception band in the vicinity of the first estimated value τ1+J, and maximize the cross-correlation function. The arrival time difference τ11. The second *'i@T is characterized by estimating the position of the target. Tsushi cleansing system.

Claims (1)

[Claims] The noise radiated from a moving target is received by N receivers, and the arrival time difference τ_i_ of the radiated noise at the receivers is
, _j (i≦N, j≦, i≠j) ■_i_, ＿
In a passive ranging system that calculates j and estimates the position of the moving target from the estimated values ■_i_, _j, a) the reception band in the receiver is set to a center frequency f
b) divide the center frequency f_c' and the bandwidth W' into a first reception band with a center frequency f_c and a bandwidth W'; , f_c
'+W'/2<f_c+W/2, c)
The cross-correlation function between the N signals in the first reception band is calculated, and the time difference τ that maximizes the cross-correlation function is determined as the first estimated value of the arrival time difference τ_i_,_j ■_i_,_
j, and d) Calculate the cross-correlation function between the N signals of the second reception band in the vicinity of the first estimated value ■_i_,_j, and calculate the time difference τ that maximizes the cross-correlation function. is a second estimated value of the arrival time difference τ_i_,_j, and e) the position of the moving target is estimated based on the second estimated value.