CN104614611B - A kind of online test method of the impaired array element of receiving antenna array - Google Patents

Abstract

The invention belongs to the online test method of the impaired array element of receiving antenna array in radar application, including sets up Test database, it is determined that receiving the autocorrelation matrix of signal, determines virtual array to be measured, the treatment of autocorrelation matrix column vectorization, column vectorThe sequence of element, determines the number of each upper Virtual array in array element position in virtual array to be measured, determines that receiving antenna array whether there is impaired array element and determine impaired element number of array and corresponding array element coordinate bit.The invention has and is not limited by the number size of impaired array element in the scale and aerial array of receiving antenna array, in the case of equal array scale, identical impaired element number of array, compared with support vector cassification method, time spent by the invention on-line checking is only the 1/50 1/4 of support vector cassification method, performance is obviously improved, and detection efficiency and reliability are higher.Thus the present invention has detection method easy, and required detection time is short, the features such as detection efficiency and reliability high.

Description

A kind of online test method of the impaired array element of receiving antenna array

Technical field

The invention belongs to a kind of method for detecting impaired element position in receiving antenna array, particularly one kind is related to The method for detecting its position in the case of there is impaired array element in the array of operation.The method using in impaired array other The information of the signal that intact array element is received, detects the tool of impaired array element exactly under complicated receiving array damage situations Body position.

Background technology

Antenna array signals treatment technology has been widely used for radar, astronomy, earthquake, sonar and radio communication etc. now Field.Aerial array is by being placed on the multiple sensor groups on the difference coordinate bit of space into for reception space signal and right The signal for receiving is processed.But in actual applications, severe natural environment and its shelf-life of sensor hardware The part array element in receiving antenna array can be caused to occur damaging etc. reason, cannot normal received signal, claim these normal The array element of work is impaired array element.With the continuous increase of the aerial array scale in radar application, there is impaired array element can Energy property is consequently increased, and causes to make array performance seriously decline because impaired array element loses many useful informations, therefore detect Impaired array element and to replace impaired array element in time most important to ensureing array performance, environment is fought especially for modern sophisticated electronic So.

The impaired array element detection method that generally uses mainly has a far field beam inversefouriertransform method at present, genetic algorithm, Matrix method, signal source reconstruction method, parameter model, support vector machine classifier method, Bayes's compressed sensing method and singular value point Solution etc..But the above method all has some limitations and defect：Inversefouriertransform method, signal source reconstruction method, ginseng Exponential model method is applied to the less situation of receiving array array element number, and in the case where array scale is larger, hardly results in remote Field wave beam, the hydraulic performance decline of the impaired array element method of detection, limits application of the above method in large scale array.Matrix Method needs to solve the equation about matrix, when solve matrix be ill-condition matrix (in matrix equation coefficient matrix to compared with Small error is very sensitive, so that the disturbance of very little can cause the solution that equation is obtained very big difference occur with potential solution, this The coefficient matrix of sample is referred to as ill-condition matrix) when, the iterative process used by the method occurs the very slow problem of convergence.It is based on The detection method of genetic algorithm is when data volume is larger it is possible that locally optimal solution, causes the impaired array for detecting and reality The impaired array in border is inconsistent (causing flase drop).And support vector cassification method but exist treatment large scale training data and compared with The disadvantages such as larger physical resource (machine internal memory and run time) are taken during high-order matrix number.

The content of the invention

The purpose of the present invention is directed to defect present in background technology, and a kind of receiving antenna array of research and design is damaged battle array The online test method of unit, the detection method is not limited by the scale and impaired array element quantity of reception antenna；To reach The consuming time is shorter under identical array damage situations, and the purpose such as more accurate is detected under equal complex array damage situations.

Solution of the invention is：Array number, the coordinate bit of each array element first according to intact array, examine for setting up The direction vector for surveying sense and its wavelength composition of database carries out relevant treatment, determines covariance matrix, virtual array In the upper Virtual array in each array element position and each array element position maximum number, to set up Test database；Then to array received to be measured To signal data carry out autocorrelation matrix, with determine virtual array to be measured and determine its upper Virtual array in each array element position Number, and then determine that whether receiving antenna array, by the number for undermining impaired array element, each impaired array element is determined finally by the poor method of searching Coordinate bit；The present invention realizes its goal of the invention with this.

Therefore the inventive method includes：

Step 1. sets up Test database：First using first array element on the array left side as the origin of coordinates, and to by complete The direction that the array number of good array, the coordinate bit of each array element, the sense for setting up Test database and its wavelength are constituted Vector carries out relevant treatment, determines covariance matrix, and then determines each array element position and each array element in virtual array by the matrix The number of Virtual array on position, wherein number it is maximum be corresponding Virtual array number in the origin of coordinates, and by intact battle array The maximum of the Virtual array number in the array number of row, the coordinate bit of each array element, virtual array on array element position is stored in processor；

Step 2. determines to receive the autocorrelation matrix of signal：To array received to be measured to signal data carried out by following formula Auto-correlation processing：

In formula：R_xx、R_ssThe autocorrelation matrix of array received signal respectively to be measured and target source signal, E () is covariance matrix Computing, X is array received signal to be measured, []^HIt is conjugate transposition computing, It is the corresponding direction vector of the array to be measured, the azimuth of echo signal and wavelength are respectively θ_sAnd λ, d_iIt is battle array in array to be measured Coordinate, the i=0 of unit, 1 ..., M-1 (M is array number),I is unit matrix,WithRespectively echo signal and The power of noise, footmark s represents signal；

Step 3. determines virtual array to be measured：By step 2 matrix R_xxSuccessively by the battle array in each element after column vectorization treatment The set expression of the difference of first coordinate bit is D={ v_k, v in formula_kIt is the either element in set D, by each element v_kCorresponding battle array The array composition difference array that unit is constituted；By different element composition set D in set D_dis, composition set D_disMethod For：It is defined by first element in set D, the element and above either element of all appearance afterwards are equivalent then to be removed, successively group Into by different element V_FThe set D of composition_dis, set D_disIn different element corresponding to array element constitute it is to be measured Virtual array；

Wherein：K is each element v in set D_kCoordinate (position), span in reference axis are [- M+1, M-1], and F is Set D_disMiddle each element V_FCoordinate (position), span in reference axis are [- M+1, M-1]；

The treatment of step 4. autocorrelation matrix column vectorization：To step 2 gained autocorrelation matrix R_xxMerge by row, i.e., from the 1st Arrange m column to cascade in order, constitute a M²The column vector y of × 1 dimension, obtains：

Then by different element composition column vector in column vector y

[R in formula_xx]_p,qIt is autocorrelation matrix R_xxThe value of middle pth row q row, the span of p and q be 1,2 ..., M, []^TRepresent transposition computing；

Step 5. column vectorThe sequence of element：By step 4 gained column vectorSort as follows：From column vector y First element start, hereafter, all element and above either element equivalences occurred in column vector y then removes so that The column vector being made up of different element is constituted successivelySo that column vectorMiddle each element puts in order and D_disIn each unit Element puts in order corresponding；

Step 6. determines the number of each upper Virtual array in array element position in virtual array to be measured：According to the result in step 4, with Column vectorIn first element be defined, determine successively in column vector y with column vectorIn each equivalent element number, each equivalence Element number correspond respectively to the number of the Virtual array in virtual array to be measured on each array element position, wherein Virtual array number In maximum respective coordinates origin on Virtual array number；

Step 7. determines that receiving antenna array whether there is impaired array element：Will be maximum in step 6 gained Virtual array number Value is compared with the maximum of Virtual array number on array element position in step 1 Test database, if it is equal, if array to be measured It is intact array, exits detection；If less than the maximum of the Virtual array number on array element position in Test database, judging it In the presence of impaired array element, step 8 is transferred to；

Step 8. determines impaired element number of array and corresponding array element coordinate bit：Array element position in the Test database of comparison step 1 On Virtual array number maximum and the difference of step 6 gained Virtual array number maximum, with determine impaired array element Number, and then the coordinate bit of each impaired array element is determined in the following manner：

Step 8.1. determines impaired array element coordinate bit：First by whole array elements (M) group known to array elements coordinate bit Into being all combinations of the different array elements of radix to be damaged element number of array determined by step 8, the seat of each array element in each combination Coordinate bit of the mark with each array element in step 1 Test database is corresponding identical, and each combination is performed in the following manner successively：

A. an optional combination is used as impaired array element group (N number of) to be measured, and will be removed in whole array elements selected to be measured impaired Remaining array element in array element group after all array elements constitutes new current array to be measured, then by the seat of each array element in current array Mark from left to right constitutes row vector successively, and the row vector makees cross correlation process with the row vector after its number of negating；

B. cross correlation process gained matrix in step a is carried out column vector process, column vector h, then this is arranged successively (w) is gathered in different element composition Virtual array position in vector, and method of its set (w) is：With first unit in column vector h Element is defined, and the element and above either element of all appearance afterwards are equivalent then to be removed, to constitute current virtual array；

C. it is defined by first element in column vector h, it is each with Virtual array position set (w) in determination column vector h successively Equivalent element number, each equivalent element number correspond respectively to each upper Virtual array in array element position in current virtual array Number, now by the number and each battle array in virtual array to be measured described in step 6 of each upper Virtual array in array element position in current virtual array The number correspondence of Virtual array is compared on first position, if not quite identical, directly performs step d；If completely the same, by step The array element combination selected in a stores standby as just fixed impaired array element group；

D. and then respectively remaining different array element combination is performed by step a, b, c circulations successively, if remaining each group is held respectively Capable result does not meet and standby condition is stored described in step c, then above-mentioned steps c at the beginning of fixed impaired array element group be reality Impaired array element group, is respectively in step 1 Test database by the coordinate bit of array element in the actual impaired array element group and corresponds to array element Coordinate bit；If being performed once occurring still conforming to storing standby condition described in step c in the result that remaining each group is performed respectively Step 8.2；

Step 8.2：Two groups of array element groups to meeting storage standby condition are detected：The different coordinates of correspondence in finding out two groups Any one array element of position, then by the output situation of the actual test array element, if without output, the array element where the array element Combination is actual impaired array element group；If there is output, another array element group is actual impaired array element group；Actual impaired array element group In the coordinate bit of correspondence array element in step 1 Test database is respectively by the coordinate bit of array element.

Determine the upper virtual array in each array element position and each array element position in virtual array described in above-mentioned steps 1 and then by the matrix Unit number, its method is to carry out vectorization treatment to the covariance matrix, it is determined that in vector each identical element number of repetition, Each number of repetition is the number of the corresponding upper Virtual array in array element position.

The present invention has not by the number size limit that array element is damaged in the scale and aerial array of receiving antenna array System, in the case of equal array scale, identical impaired element number of array, the present invention is supported compared with support vector cassification method Vector machine classification takes in the case where 4,5,10 array elements are damaged and is respectively 1.0225s, 1.4072s, 22.1638s, And the inventive method is then respectively 0.0203s, 0.2974s, 4.8251s, the time spent by embodiment of the present invention is only branch The 1/50-1/4 of vector machine classification is held, performance is obviously improved, detection efficiency and reliability are higher.Thus the present invention has detection Method is easy, and required detection time is short, the features such as detection efficiency and reliability high.

Brief description of the drawings：

Fig. 1 is that (black triangle is impaired array element to actual impaired receiving antenna array in figure, and white triangles shape is intact Array element)；

Fig. 2 is each upper corresponding Virtual array number schematic diagram in array element position in the specific embodiment of the invention；In figure：x Axle is the coordinate bit of each upper corresponding Virtual array in array element position in actual array；Y-axis is on each array element position in virtual array The number of Virtual array；

In Fig. 3：A. the impaired array element coordinate bit in the impaired array detected by the specific embodiment of the invention is illustrated Figure；B. it is the coordinate bit schematic diagram of impaired array element in actual impaired array.

Specific implementation content：

The present embodiment be with the operation wavelength 10cm of aerial array, array element spacing d=5cm be signal wavelength lambda half, As a example by having the M=25 uniform linear array of omnidirectional's array element in 25 times of apertures of half-wavelength (5cm), carrier frequency f=3GHz.Letter Number arrival bearing θ₀It is 0 °, sets horizontal linear as x-axis, each array element is with a distance from the origin of coordinates from closely to being far followed successively byActual impaired array element serial number 7,8,12,13,15,16,17,20,21, 23 }, the actual impaired array schematic diagram is as shown in Figure 2.

Step 1. sets up Test database：First using first array element on the array left side as the origin of coordinates.Array is not having Have in the case of being damaged, determine that the collection that the difference of any two array element coordinate bit in array is constituted is combined into：

D_int={ v₀=0d=0, v₁=1d=5 ..., v₂₄=24d=120, v_-1=-1d=-5, v₀=0d=0 ..., v₂₃ =23d=115, v_-1=-1d=-5 ..., v₂₂=22d=110 ..., v₀=-24d=-120, v₀=-23d=-115 ..., v₀ =0d=0 }

Then, it is determined that the number of Virtual array is gone up in each array element of virtual array position under present case, process is as follows：

A. each array element coordinate bit is in determining virtual array

{v_-24=-24d=-120, v_-23=-23d=-115 ..., v_-2=-2d=-10,

v_-1=-1d=-5, v₀=0d=0, v₁=1d=5 ..., v₂₄=24d=120 }

B. determine the number of each upper Virtual array in array element position in virtual array, obtain

{1,2,3,…,24,25,24,…,3,2,1}

C. by array number 25, array element spacing d=5cm, the maximum of the Virtual array number in virtual array on array element position 25 are stored in processor；

Step 2. determines to receive the autocorrelation matrix of signal：To array received to be measured to signal data carry out auto-correlation fortune Calculation is processed, and determines autocorrelation matrix R_xxFor：

Step 3. determines virtual array to be measured：By step 2 matrix R_xxSuccessively by the battle array in each element after column vectorization treatment The set expression of the difference of first coordinate bit is

D={ v₀=0d=0, v₁=1d=5 ..., v₂₄=24d=120, v_-1=-1d=-5, v₀=0d=0 ..., v₂₃= 23d=115,

v_-1=-1d=-5 ..., v₂₂=22d=110 ..., v_-24=-24d=-120, v_-23=-23d=-115 ..., v₀ =0d=0 }

Be defined by first element in set D, it is all after the element that occurs it is equivalent with above either element then remove, according to It is secondary to constitute the set D being made up of different element_dis

D_dis=0,1d=5,2d=10 ..., 24d=120, -1d=-5, -2d=-10 ..., -24d=-120 }

Set D_disIn different element corresponding to array element constitute current virtual array；

Step 4. autocorrelation matrix column vector：To gained autocorrelation matrix R in step 2_xxCarry out column vectorization and obtain one The column vector of individual 625 × 1 dimension

Y=[5.0618, -1.7223+0.7969j ..., -2.2027-0.2719j, 0.3464+1.0272j ...,

-0.9640-1.3453j,…,-1.4074-1.9790j,…,-0.0648-0.2305j,5.0618]^T

Step 5. column vectorThe sequence of element：It is hereafter, all in vectorial y since first element in column vector y The element of appearance is equivalent with above either element then to be removed, constitutes the vector being made up of different element successively

To cause column vectorMiddle each element puts in order and D_disIn each element put in order it is corresponding；

Step 6. determines the number of each upper Virtual array in array element position in virtual array to be measured：According to the result in step 4, press Column vectorIn first element be defined, determine successively in column vector y with column vectorIn each equivalent element numberEnter And determine the number of Virtual array in virtual array to be measured on each array element positionAnd Virtual array number maximumForIn c-th element, by calculate determine

In y withThe 1st elementEquivalent element number is 15, is obtained

In y withThe 2nd elementEquivalent element number is 1, is obtained

In y withThe 23rd elementEquivalent element number is 1, is obtained

In y withThe 48th elementEquivalent element number is 1, is obtained

In y withThe 49th elementEquivalent element number is 1, is obtained That is with column vector in determination column vector yIn each equivalent element number

And the maximum of the Virtual array number in virtual array on array element position

Step 7. determines that receiving antenna array whether there is impaired array element：Step 6 gained Virtual array number maximum 15 Less than the Virtual array number maximum 25 on array element position in step 1 Test database, determine that current array has impaired array element, Perform step 8；

Step 8. determines impaired element number of array and corresponding array element coordinate bit：By in the Test database of comparison step 1 The difference of Virtual array number maximum 25 and step 6 gained Virtual array number maximum 15 determines the number of impaired array element to be measured It is 10 (as N=25-15=10), and then determines the coordinate bit of each impaired array element in the following manner：

Step 8.1. determines impaired array element coordinate bit：First, by whole 25 array element groups known to array elements coordinate bit Into all different array elements combination on the basis of 10 impaired element number of array determined by step 8, the array element coordinate bit in each combination Respectively：

{{d₀,d₁₁,d₁₂,d₁₃,…,d₂₂,d₂₃,d₂₄},{d₀,d₁,d₁₂,d₁₃,…,d₂₂,d₂₃,d₂₄},...,{d₀,d₁,d₂, d₃,…,d₁₃,d₁₄,d₁₅}}

Each combination in each array element coordinate bit with the coordinate bit (d of each array element in step 1 Test database₀,d₁,…, d₂₄({ 0,5 ..., 120 })) it is corresponding identical, each combination is performed according to following way of example：

A. first, an optional combination, the coordinate bit d of each array element in its combination₁,d₂,d₃,d₄,d₅,d₆,d₇,d₈,d₉,d₁₀ ({ 5,10,15,20,25,30,35,40,45,50 }) will remove selected impaired array element group as impaired array element group in whole array elements In remaining array element after all array elements constitute new current array to be measured, the coordinate bit of remaining good array element is respectively：d₀,d₁₁, d₁₂,d₁₃,…,d₂₄({0,55,60,75,…,120}).Then by the coordinate bit of each array element in current array from left to right successively Composition row vector [d₀,d₁₁,d₁₂,d₁₃,…,d₂₄] ([0,55,60,75 ..., 120]), by the row after the vector and its number of negating Vector-[d₀,d₁₁,d₁₂,d₁₃,…,d₂₄] ([- 0, -55, -60, -75 ..., -120]) make cross correlation process；Obtain matrix：

B. to cross correlation process gained matrix R in step a_ddColumn vector is carried out to process, obtain column vector：H=[h₀=0d= 0,h₁₁=11d=55 ..., h₀=0d=0 ..., h_-1=-1d=-5, h₀=0d=0],

Different element in the column vector is constituted into the set of Virtual array position successively successively again：

W=0d=0,11d=55,12d=60 ..., 24d=120, -11d=-55 ..., -23d=-115, -24d =-120 }；

C. column vector is pressed：H=[h₀=0d=0, h₁₁=11d=55 ..., h₀=0d=0 ..., h_-1=-1d=-5, h₀= 0d=0] in first element (h₀=0d=0) it is defined, column vector h=[h are determined successively₀=0d=0, h₁₁=11d=55 ..., h₀=0d=0 ..., h_-1=-1d=-5, h₀=0d=0] in Virtual array position set (w={ 0d=0,11d=55,12d= 60 ..., 24d=120, -11d=-55 ..., -23d=-115, -24d=-120 }) in each equivalent element number 15,4, 3 ..., 1,1 ..., 1,1 }, each equivalent element number corresponds respectively to the Virtual array on each array element position in current virtual array Number；With the Virtual array number unit number on each array element position in virtual array to be measured described in step 6It is inconsistent, hold Row step d；

D. and then respectively remaining different array element combination is performed by step a, b, c circulations successively；

Combined when m groups array element therein is gone to,

A. will be positioned at d₀,d₁,…,d₂₄Removed in the linear array of all 25 array elements composition on ({ 0,5 ..., 120 }) 10 are located at d respectively₇,d₈,d₁₂,d₁₃,d₁₅,d₁₆,d₁₇,d₂₀,d₂₁,d₂₃(35,40,60,75,80,75,100,105,115) on Array element, i.e. 10 array elements are impaired array element, and the coordinate bit of remaining array element is respectively：d₀,d₁,d₂,d₃,…,d₂₄(0,5,10, 15,…,120).Then the coordinate bit of each array element in current array is from left to right constituted into row vector [d successively₀,d₁,d₂, d₃,…,d₂₄] ([0,5,10,15 ..., 120]), by the row vector-[d after the vector and its number of negating₀,d₁,d₂,d₃,…,d₂₄] ([- 0, -5, -10, -15 ..., -120]) make cross correlation process；Obtain matrix：

B. to cross correlation process gained matrix R in step a_ddColumn vector is carried out to process, obtain column vector ({ h₀=0d=0, h₁=1d=5 ..., h₀=0d=0 ..., h_-2=-2d=-10, h₀=0d=0 }), then successively will be different in the column vector Element constitutes the set of Virtual array position successively：

W=0d=0,1d=5,2d=10 ..., 24d=120, -1d=-5 ..., -23d=-115, -24d=- 120 },

C. column vector ({ h is pressed₀=0d=0, h₁=1d=5 ..., h₀=0d=0 ..., h_-2=-2d=-10, h₀=0d= 0 } first element (h in)₀=0d=0) it is defined, column vector ({ h is determined successively₀=0d=0, h₁=1d=5 ..., h₀=0d= 0,…,h_-2=-2d=-10, h₀=0d=0 }) in Virtual array position set (w=0d=0,1d=5,2d=10 ..., 24d=120, -1d=-5 ..., -23d=-115, -24d=-120) in each equivalent element number 15,9,7 ..., 1, 1 ..., 1,1 }, each equivalent element number corresponds respectively to the number of the Virtual array on each array element position in current virtual array, Now with each array element position in virtual array to be measured described in step 6 on Virtual array number unit numberUnanimously, then incite somebody to action this Array element in the array element combination selected in secondary combination step a stores standby as just fixed impaired array element；

Search all possible array element combination thoroughly successively, finally occur two groups and meet the condition of storing for future use,

Wherein first group of array element combines coordinate bit：

d₇,d₈,d₁₂,d₁₃,d₁₅,d₁₆,d₁₇,d₂₀,d₂₁,d₂₃({35,40,60,65,75,80,85,100,105,115})

Second group of array element combines coordinate bit：

d₁,d₃,d₄,d₇,d₈,d₉,d₁₁,d₁₂,d₁₆,d₁₇({5,15,20,35,40,45,55,60,80,85})

Perform step 8.2；

Step 8.2：Two groups of array elements combination to meeting storage standby condition is detected：Find out step 8.1 two groups are received Damaging different in array element coordinate bit has d₁,d₃,d₄,d₉,d₁₁,d₂₀,d₂₁,d₂₃({ 5,15,20,45,55,100,105,115 }), this D is located in second group of array element combination of place's test₁The output situation of array element on ({ 5 }), it is found that the array element has output data, therefore really Fixed first group of array element combination is actual impaired array element, is respectively sat by array element in the impaired array element group of 10 reality of impaired array element Mark is the coordinate bit of correspondence array element in step 1 Test database：{d₇,d₈,d₁₂,d₁₃,d₁₅,d₁₆,d₁₇,d₂₀,d₂₁,d₂₃, Namely { 35,40,60,65,75,80,85,100,105,115 }.

According to the signal that impaired array received is arrived, the corresponding virtual array of impaired array is determined using method proposed by the present invention Row；Void in Virtual array number virtual array corresponding with intact array in virtual array on each array element position on each array element position Intend element number of array comparison diagram as shown in Figure 2；The impaired array element coordinate bit detected by method proposed by the present invention with actually receive The schematic diagram for damaging impaired array element coordinate bit in array is as shown in Figure 3.

Present embodiment is being all the uniform linear array of array number 25, actual impaired battle array by Matlab simulation runs In the case that first number is equal, the present invention compares with the support vector cassification method detection consuming time in background technology：

	4 array elements are damaged	5 array elements are damaged	10 array elements are damaged
				Consuming time (/s) of the present invention	0.0203s	0.2974s	4.8251s
Support vector cassification method (/s)	1.0225s	1.4072s	22.1638s

Be can be seen that from above-mentioned list, under the conditions of being damaged in identical homogenous linear battle array and on an equal basis, the time that the present invention expends With the time-consuming contrast of support vector cassification method in the background technology for generally using, it can be seen that embodiment of the present invention The spent time is only the 1/50-1/4 of support vector cassification method, and performance is obviously improved, and detection efficiency and reliability are higher.

Claims (2)

1. a kind of receiving antenna array is damaged the online test method of array element, including：

Step 1. sets up Test database：First using first array element on the array left side as the origin of coordinates, and to by intact battle array The direction vector that the array number of row, the coordinate bit of each array element, the sense for setting up Test database and its wavelength are constituted Relevant treatment is carried out, covariance matrix is determined, and then is determined in virtual array on each array element position and each array element position by the matrix The number of Virtual array, wherein number it is maximum be corresponding Virtual array number in the origin of coordinates, and by intact array The maximum of the Virtual array number in array number, the coordinate bit of each array element, virtual array on array element position is stored in processor；

Step 2. determines to receive the autocorrelation matrix of signal：To array received to be measured to signal data carried out from phase by following formula Pass is processed：

$\begin{array}{c}{R}_{xx}=E\left({\mathrm{XX}}^H\right)=b\left({\mathrm{\θ}}_s\right)R_{ss}{b}^H\left({\mathrm{\θ}}_s\right)+{\mathrm{\σ}}_n^{2}I={\mathrm{\σ}}_s^{2}b\left({\mathrm{\θ}}_s\right)b^H\left({\mathrm{\θ}}_s\right)+{\mathrm{\σ}}_n^{2}I\\ ={\mathrm{\σ}}_s^2\left[\begin{array}{c}{e}^{-j2{\mathrm{\πd}}_0{\mathrm{sin\θ}}_s/\mathrm{\λ}}\\ e^{-j2{\mathrm{\πd}}_1{\mathrm{sin\θ}}_s/\mathrm{\λ}}\\ .\\ .\\ .\\ e^{-j2{\mathrm{\πd}}_{M-1}{\mathrm{sin\θ}}_s/\mathrm{\λ}}\end{array}\right]\·\left[\begin{array}{cccc}{e}^{j2{\mathrm{\πd}}_0{\mathrm{sin\θ}}_s/\mathrm{\λ}}& e^{j2{\mathrm{\πd}}_1{\mathrm{sin\θ}}_s/\mathrm{\λ}}& \mathrm{...}& e^{j2{\mathrm{\πd}}_{M-1}{\mathrm{sin\θ}}_s/\mathrm{\λ}}\end{array}\right]+{\mathrm{\σ}}_n^{2}I\\ ={\mathrm{\σ}}_s^2\left[\begin{array}{cccc}{e}^{j2\mathrm{\π}(d_0-d_0){\mathrm{sin\θ}}_s/\mathrm{\λ}}& e^{j2\mathrm{\π}(d_1-d_0){\mathrm{sin\θ}}_s/\mathrm{\λ}}& \mathrm{...}& e^{j2\mathrm{\π}(d_{M-1}-d_0){\mathrm{sin\θ}}_s/\mathrm{\λ}}\\ e^{j2\mathrm{\π}(d_0-d_1){\mathrm{sin\θ}}_s/\mathrm{\λ}}& e^{j2\mathrm{\π}(d_1-d_1){\mathrm{sin\θ}}_s/\mathrm{\λ}}& \mathrm{...}& e^{j2\mathrm{\π}(d_{M-1}-d_1){\mathrm{sin\θ}}_s/\mathrm{\λ}}\\ .& .& & .\\ .& .& \mathrm{...}& .\\ .& .& & .\\ e^{j2\mathrm{\π}(d_0-d_{M-1}){\mathrm{sin\θ}}_s/\mathrm{\λ}}& e^{j2\mathrm{\π}(d_1-d_{M-1}){\mathrm{sin\θ}}_s/\mathrm{\λ}}& \mathrm{...}& e^{j2\mathrm{\π}(d_{M-1}-d_{M-1}){\mathrm{sin\θ}}_s/\mathrm{\λ}}\end{array}\right]+{\mathrm{\σ}}_n^{2}I\end{array}$

In formula：R_xx、R_ssThe autocorrelation matrix of array received signal respectively to be measured and target source signal, E () is transported for covariance matrix Calculate, X is array received signal to be measured, []^HIt is conjugate transposition computing, It is the corresponding direction vector of the array to be measured, the azimuth of echo signal and wavelength are respectively θ_sAnd λ, d_iIt is battle array in array to be measured Unit coordinate, i=0,1 ..., M-1, M be array number,I is unit matrix,WithRespectively echo signal and The power of noise；

Step 3. determines virtual array to be measured：By step 2 matrix R_xxThe array element in each element is sat successively after column vectorization treatment The set expression of the difference of mark is D={ v_k, v in formula_kIt is the either element in set D, by each element v_kCorresponding array element structure Into array constitute difference array；By different element composition set D in set D_dis, composition set D_disMethod be：Press First element is defined in set D, and the element and above either element of all appearance afterwards are equivalent then to be removed, and is constituted successively by each Differ element V_FThe set D of composition_dis, set D_disIn different element corresponding to array element constitute virtual array to be measured Row；

Wherein：K is each element v in set D_kCoordinate bit, span in reference axis are [- M+1, M-1], and F is set D_dis Middle each element V_FCoordinate bit, span in reference axis are [- M+1, M-1]；

The treatment of step 4. autocorrelation matrix column vectorization：To step 2 gained autocorrelation matrix R_xxBy row merge, i.e., from the 1st row to M column is cascaded in order, constitutes a M²The column vector y of × 1 dimension, obtains：

$\begin{array}{c}y={\[{\[R_{xx}\]}_{1,1},{\[R_{xx}\]}_{2,1},\mathrm{...},{\[R_{xx}\]}_{M,1},{\[R_{xx}\]}_{2,1},\mathrm{...},{\[R_{xx}\]}_{M,2},\mathrm{...},{\[R_{xx}\]}_{p,q},\mathrm{...},{\[R_{xx}\]}_{M,M}\]}^T\\ ={\mathrm{\σ}}_s^2{\[1,e^{j2\mathrm{\π}(d_0-d_1){\mathrm{sin\θ}}_s/\mathrm{\λ}},\mathrm{...},e^{j2\mathrm{\π}(d_0-d_{M-1}){\mathrm{sin\θ}}_s/\mathrm{\λ}},\mathrm{...},e^{j2\mathrm{\π}(d_{M-1}-d_0){\mathrm{sin\θ}}_s/\mathrm{\λ}},\mathrm{...},e^{j2\mathrm{\π}(d_{M-1}-d_{M-1}){\mathrm{sin\θ}}_s/\mathrm{\λ}}\]}^T\\ ={\mathrm{\σ}}_n^2{\[1,0,\mathrm{...},0,0,1,\mathrm{...},0,\mathrm{...},1,\mathrm{...},0,1\]}^T\end{array}$

Then by different element composition column vector in column vector y

[R in formula_xx]_p,qIt is autocorrelation matrix R_xxThe value of middle pth row q row, the span of p and q is 1,2 ..., M, [ ]^TRepresent transposition computing；

Step 5. column vectorThe sequence of element：By step 4 gained column vectorSort as follows：From in column vector y One element starts, and hereafter, all elements occurred in column vector y are equivalent with above either element then to be removed, so that successively The column vector that composition is made up of different elementSo that column vectorMiddle each element puts in order and D_disIn each element row Row order is corresponding；

Step 6. determines the number of each upper Virtual array in array element position in virtual array to be measured：According to the result in step 4, with arrange to AmountIn first element be defined, determine successively in column vector y with column vectorIn each equivalent element number, each equivalent unit Plain number corresponds respectively to the number of the Virtual array on each array element position in virtual array to be measured, wherein in Virtual array number The number of the Virtual array on maximum respective coordinates origin；

Step 7. determines that receiving antenna array whether there is impaired array element：By maximum in step 6 gained Virtual array number with The maximum of Virtual array number in step 1 Test database on array element position is compared, if it is equal, if array to be measured be Good array, exits detection；If less than the maximum of the Virtual array number on array element position in Test database, judging that it is present Impaired array element, is transferred to step 8；

Step 8. determines impaired element number of array and corresponding array element coordinate bit：In the Test database of comparison step 1 on array element position The maximum of Virtual array number and the difference of step 6 gained Virtual array number maximum, to determine the number of impaired array element, enter And the coordinate bit of each impaired array element is determined in the following manner：

Step 8.1. determines impaired array element coordinate bit：First by whole array element compositions known to array elements coordinate bit with step 8 Identified impaired element number of array for radix different array elements all combinations, each combination in each array element coordinate bit with step The coordinate bit correspondence of each array element is identical in rapid 1 Test database, and each combination is performed in the following manner successively：

A. an optional combination is used as impaired array element group to be measured, and institute in selected impaired array element group to be measured will be removed in whole array elements New current array to be measured is constituted by remaining array element after array element, then by the coordinate bit of each array element in current array from a left side to The right side constitutes row vector successively, and the row vector makees cross correlation process with the row vector after its number of negating；

B. column vector is carried out to cross correlation process gained matrix in step a to process, obtain column vector h, then successively by the column vector In the composition Virtual array position set of different element, method of its set is：First in column vector h element is defined, all The element for occurring afterwards is equivalent with above either element then to be removed, to constitute current virtual array；

C. it is defined by first element in column vector h, each equivalent unit in gathering with Virtual array position in determination column vector h successively Plain number, each equivalent element number corresponds respectively to the number of each upper Virtual array in array element position in current virtual array, now By the number of each upper Virtual array in array element position in current virtual array and each upper void in array element position in virtual array to be measured described in step 6 The number correspondence of matroid unit is compared, if not quite identical, directly performs step d；If completely the same, will be selected in step a Fixed array element combination stores standby as just fixed impaired array element group；

D. and then respectively remaining different array element combination is performed by step a, b, c circulations successively, if what remaining each group was performed respectively Result does not meet and standby condition is stored described in step c, then impaired array element group fixed at the beginning of above-mentioned steps c institutes is actual being damaged Array element group, is respectively the coordinate of correspondence array element in step 1 Test database by the coordinate bit of array element in the actual impaired array element group Position；If performing step once occurring still conforming to storing standby condition described in step c in the result that remaining each group is performed respectively 8.2；

Step 8.2：Two groups of array element groups to meeting storage standby condition are detected：Find out the different coordinate bits of correspondence in two groups Any one array element, then by the output situation of the actual test array element, if without output, the array element combination where the array element As actual impaired array element group；If there is output, another array element group is actual impaired array element group；It is each in actual impaired array element group The coordinate bit of correspondence array element in step 1 Test database is by the coordinate bit of array element.

2. the online test method of the impaired array element of the receiving antenna array as described in claim 1, it is characterised in that institute in step 1 State and then the number of the upper Virtual array in each array element position and each array element position in virtual array is determined by the matrix, its method is to this Covariance matrix carries out vectorization treatment, it is determined that in vector each identical element number of repetition, each number of repetition is corresponding battle array The number of Virtual array on first position.