JPH0312710B2 - - Google Patents

Description

Detailed Description of the Invention (Technical Field) The present invention forms multi-beams using an array of wave-receiving elements having the same directivity pattern, and the direction from the output of the multi-beams to the signal source position. The present invention relates to a direction cosine estimation method that removes an error in estimating the direction cosine caused by the directivity pattern of the wave receiving element when estimating the cosine.

(Background technology) In sonar, acoustic positioning, and radar,
A multi-beam is formed using an array in which elements are arranged in a straight line (one-dimensional), plane (two-dimensional), or three-dimensionally in space, the maximum point of the discrete output of this multi-beam is found, and the area near the maximum point is calculated. A widely used directional cosine estimation method is to perform interpolation on the multi-beam output in the spatial frequency domain, find the maximum point of the output after interpolation, and then calculate the directional cosine of the signal source position using the spatial frequency corresponding to the maximum point. ing.

By the way, the conventional method is based on the assumption that the directivity of the element is omnidirectional, that is, the receiving sensitivity of the element with respect to the direction is constant. However, the element usually has some degree of directivity, and therefore, when the direction cosine is estimated by the conventional method, there is a drawback that an error occurs due to the directivity of the element.

FIG. 1 is a geometric explanatory diagram when the array is a linear (one-dimensional) array, with 11 ₁ , 11 ₂ ,...,
11 _i , ..., 11 _M are respective wave receiving elements, 12 is a reference axis on which the wave receiving elements are arranged, 13 is a straight line indicating the direction of the signal source, and θ _x is a straight line indicating the direction of the signal source with respect to the reference axis 12 13 direction cosine angle.

FIG. 2 is a functional block diagram of a conventional direction cosine _estimation method assuming that _the directivity pattern of each receiving element is non-directional.
_i , ..., 21 _M are amplifiers, 22 ₁ , 22 ₂ , ...,
22 _i ,..., 22 _M are each bandpass filters, 23
is the beamformer, y ₁ , y ₂ , ..., y _o , y _N is the multi-beam output, 24 is the first maximum point detector, k _n
is the discrete value of the spatial frequency at which the multi-beam output is maximum, 25 is the interpolator, y(k) is the beam output after interpolation, 26 is the second maximum point detector, and k^ is the beam output after the interpolation. y(k) is the maximum spatial frequency, 27 is a converter, cosθ^ _x is a direction cosine estimate corresponding to the k^, and 28 is an output terminal.

Consider the case where the signal source is sufficiently far away compared to the aperture of a linear (one-dimensional) array. The direction cosine angle of the signal source direction 13 is θ _x , and the received signal S _p (t) at the reference position coordinates of the reference axis 12 is S _p (t)=A・e ^j(2 〓 ^ft+ 〓 ⁾ (1 ). However, A is the amplitude, f is the signal frequency, and φ is the initial phase. Then, the input signal S _i (t) to the i-th element 11 _i is θ Considering the arrival time difference between the reference point of the signal wavefront arriving from _{the x} direction and the i-th element 11 i. is given by However, A _i is the amplitude, and ζ _i is the i-th 1st
1 _i position coordinates, c is the propagation velocity. Here, the spatial frequency k [Reference: S.
Haykin, Editor JHJustice NLOwsley JL
Yen ACKak: “Array Signal
Processing”，PRENTICE−HALL SIGNAL
PROCESSING SERIES, P119, 1985].

kΔ=2πcos θ _x /λ (3) where λ is the wavelength of the signal.

The physical meaning of the spatial frequency k is that the phase difference from the reference position coordinate when measured in the direction cosine angle θ _x direction of the signal source direction at the position coordinate ζ _i on the reference axis is kζ _i
(radians). The spatial frequency k
Using the relationship f/c=1/λ and omitting the carrier component and initial phase component, equation (2) is rewritten as follows: S _i (t)=A _i ·e ^−jk 〓 ⁱ (4).

The beamformer 23 has N discrete values k ₁ , k ₂ , ..., k _o , ..., k _N of the spatial frequency k in directional cosine angles θ _x1 , θ _x2 , ..., θ _xo , θ _xN directions. The outputs of the multi-beams y ₁ ,
Output y ₂ ,…, y _o ,…, y _N. In order to obtain the beamformer output y _o , the beamformer 23
Phase correction for the signal from i-th element 11 _i
Perform e ^jkn 〓 ⁱ , k _o = 2πcos θ _xo /λ. Take the sum of all M signals after the phase correction, and calculate the integration time.
Output the power at T ₂ − T ₁ as the following formula y _o .

y _o =｜∫ ^T2 _T1 M/Σ/i=1 S _i (t)・e ^jkn 〓 ⁱ
dt
| ² (5) The first maximum point detector 24 detects y ₁ , y ₂ , ..., _yo ,
Find the maximum value y _n of ..., y _N , and output the spatial frequency k _n at that time.

The interpolator 25 performs an interpolation operation on the multi-beam output near the spatial frequency k _n and outputs a beam output y(k) as a continuous function of the spatial frequency k.

The second maximum point detector 26 detects the maximum value of y(k).
Find max y(k) and output the spatial frequency k^ at that time.

The converter 27 outputs the direction cosine cosθ^ _x =λk^/2π (6) from the above k^ based on equation (3). FIG. 3 shows elements 11 ₁ , 11 ₂ , 11
_i ,...,11 The conventional method shown in Fig. 2 when _M is omnidirectional, that is, the normalized directivity pattern of the element holds b _(k) = 1 for all spatial frequencies k. FIG. When b _(k) = 1, the first half k¨ obtained by the conventional method matches the true value, and therefore the cosθ^ _x also matches the true value.

Figure 4 shows elements 11 ₁ , 11 ₂ ,..., 11 _i ,..., 1
FIG. 1 is an explanatory diagram of the case where the conventional method is applied as is when _1M is not omnidirectional.

Literature “RJUrick: “Principles of
“Underwater Sound”, McGraw-Hill, P.57,
1967, each element 11 ₁ , 11 ₂ , ..., 11 _i , ..., 11 constituting the array
When _M has the same directivity pattern b _(k) , the multi-beam directivity pattern is the product of the multi-beam directivity pattern when each element is omnidirectional and the directivity pattern of the element. Since the output of the beamformer in this case is given, y′ ₁ ,
y′ ₂ ,…, y′ _o ,…, y′ _N are the beamformer outputs y ₁ , y ₂ ,…, y _o , when each element is omnidirectional.
…, y _N , y′ _o = b(k _o )・_yo ; n=1, 2, …, n, …,
It can be expressed as N (7). Here, b(k _o ) is the directivity pattern value at the discrete value k _o of the spatial frequency, that is, the element sensitivity.

Therefore, by interpolating y′ ₁ , y′ ₂ , . . _{. , y′ o ,} . That is, in the conventional direction cosine estimation method, when the elements constituting the array have directivity, an estimation error Δcosθ _x =λΔk/2π (where Δk=k^′−k^) occurs.

(Problem to be solved by the invention) The purpose of the present invention is to eliminate this drawback by multiplying the output of the beamformer by the reciprocal of the element sensitivity to obtain equivalently the same beam output as when the element is omnidirectional, and This method eliminates the spatial frequency estimation error caused by applying the conventional method using elements, and will be explained in detail below.

(Structure and operation of the invention) _FIG . 5 shows an embodiment of the present invention.
1 ₂ ,...,51 _o ,...,51 _N are multipliers, 52
₁ , 52 ₂ , . . . , 52 _o , . . . , 52 _N are registers. Registers 52 ₁ , 52 ₂ ,..., 52 _o ,...,
52 _N is the spatial frequency k ₁ of the directivity pattern of the element,
The reciprocals of the values at k ₂ , ..., k _o , ..., k _N , that is, the reciprocals of the element sensitivity 1/b (k ₁ ), 1/b (k ₂ ), ...,
The values of 1/b(k _o ),..., 1/b(k _N ) are stored as constants, and the multipliers 51 ₁ , 51 ₂ ,..., 51
_o ,...,51 _N is the constant and the beamformer 23
The product of the outputs y′ ₁ , y′ ₂ ,…, y′ _o ,…, y′ _N is y′ ₁ /b
(k ₁ ), y′ ₂ /b(k ₂ ),…, y′ _o /b(k _o ),…, y
′ _N /
b(k _N ) and the first maximum point detector 24
is output to the interpolator 25.

From the above equation (7), the multipliers 51 ₁ , 51 ₂ , ..., 5
The output of 1 _o ,...,51 _N is y′ _o /b(k _o )=b(k _o ) ・y _o /b(k _o )= _yo ; n=1,2,...,N
(8), and therefore the detection of the maximum point k _n by the first maximum point detector 24 and the interpolation operation by the interpolator 25 are independent of the directivity pattern of the element, and the detection of the maximum point k n by the first maximum point detector 24 and the interpolation operation by the interpolator 25 are independent of the directivity pattern of the element. Same beam power as in case y ₁ , y ₂ ,
…, y _o ,…, y _N.

As explained above, in the embodiment of the present invention, in the direction cosine estimation method using elements with the same directivity, by multiplying the beamformer output by the reciprocal of the reception sensitivity, as shown in equation (8), 1st
The detection of the maximum point k _n by the maximum point detector and the interpolation operation by the interpolator will be performed on the same beam output as if the element was omnidirectional, regardless of the directivity pattern of the element. , it is possible to eliminate the estimation error Δk of the spatial frequency that occurs when applying the conventional method, and therefore there is an advantage that the estimation error Δcosθ _x of the direction cosine can be eliminated.

(Effects of the Invention) The present invention has the advantage that even when the elements constituting the array have directivity, it is possible to accurately estimate the direction cosine of the signal source position using the output of the multi-beam. , acoustic positioning, and directional cosine estimation methods in radar.

[Brief explanation of the drawing]

Figure 1 is a geometric explanatory diagram when the array is a linear (one-dimensional) array, and Figure 2 is a conventional direction cosine estimation assuming that the directivity of each receiving element is omnidirectional. A functional block diagram of the method. Figure 3 is an explanatory diagram of the conventional method when each element is omnidirectional, and Figure 4 is an explanatory diagram of the conventional method when each element is not omnidirectional and has the same directivity pattern. FIG. 5 is a functional block diagram of an embodiment of the present invention, which is an explanatory diagram when applied as is. Explanation of symbols (Fig. 2, Fig. 5), 21 ₁ , 21 ₂ ,
..., 21 _i , ..., 21 _M are amplifiers, 22 ₁ , 2
2 ₂ ,..., 22 _i ,..., 22 _M are band pass filters, 23 is a beamformer, y ₁ , y ₂ ,..., y _o ,
..., y _N is the output of the multi-beam, 24 is the first maximum point detector, k _n is the discrete value of the spatial frequency at which the multi-beam output is maximum, 25 is the interpolator, y _(k) is after interpolation beam output, 26 is the second maximum point detector;
k^ is a spatial frequency at which the beam output y _(k) after the interpolation is maximum, 27 is a converter, cosθ^ _x is an estimated direction cosine value corresponding to k^, and 28 is an output terminal. 51 ₁ ,
51 ₂ ,...,51 _o ,...,51 _N is a multiplier, 52 ₁ ,
52 ₂ ,..., 52 _o ,..., 52 _N are registers.

Claims (1)

[Claims] 1 Using an array of wave receiving elements with the same directivity pattern, beams (hereinafter referred to as multi-beams) are formed in directions corresponding to multiple discrete values of spatial frequencies, and signals are extracted from the output of the multi-beams. A directional cosine estimating device for estimating a directional cosine of a source position, comprising: a multiplier for calculating the product of the reciprocal of the element sensitivity in the direction corresponding to the discrete value of the spatial frequency and the multi-beam output in the direction; It has an interpolator that performs interpolation on the output, and a maximum point detector that determines the spatial frequency at which the output of the interpolator is maximum, and calculates the direction cosine corresponding to the spatial frequency as the estimated value of the direction cosine of the signal source position. A direction cosine estimation method characterized by: