CN109212466A - A kind of broadband direction-finding method based on quantum dragonfly mechanism of Evolution - Google Patents
A kind of broadband direction-finding method based on quantum dragonfly mechanism of Evolution Download PDFInfo
- Publication number
- CN109212466A CN109212466A CN201811017243.0A CN201811017243A CN109212466A CN 109212466 A CN109212466 A CN 109212466A CN 201811017243 A CN201811017243 A CN 201811017243A CN 109212466 A CN109212466 A CN 109212466A
- Authority
- CN
- China
- Prior art keywords
- quantum
- dragonfly
- broadband
- signal
- vector
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
- G01S3/00—Direction-finders for determining the direction from which infrasonic, sonic, ultrasonic, or electromagnetic waves, or particle emission, not having a directional significance, are being received
- G01S3/02—Direction-finders for determining the direction from which infrasonic, sonic, ultrasonic, or electromagnetic waves, or particle emission, not having a directional significance, are being received using radio waves
- G01S3/14—Systems for determining direction or deviation from predetermined direction
- G01S3/143—Systems for determining direction or deviation from predetermined direction by vectorial combination of signals derived from differently oriented antennae
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
- Y02D30/00—Reducing energy consumption in communication networks
- Y02D30/70—Reducing energy consumption in communication networks in wireless communication networks
Landscapes
- Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Radar, Positioning & Navigation (AREA)
- Remote Sensing (AREA)
- Radar Systems Or Details Thereof (AREA)
- Complex Calculations (AREA)
Abstract
The present invention provides a kind of broadband direction-finding method based on quantum dragonfly mechanism of Evolution, by establishing Sampling for Wide-Band Signal model, initialize quantum dragonfly evolution parameter, calculate the fitness of every quantum dragonfly, to the relevant parameter of quantum dragonfly group the first half more frontier radius and neighborhood quantum dragonfly, the relevant parameter of every quantum dragonfly is updated to later half, calculate the fitness value of all quantum dragonflies position, judge whether to reach maximum number of iterations, if having reached, then global optimum of quantum dragonfly group quantum position is mapped to optimal location, obtain the angle to be estimated of broadband Mutual coupling.The present invention carries out direction finding to broadband signal, reduce operand and operation time, improve convergence rate and convergence precision, realize high-precision direction finding, Mutual coupling can be carried out to coherent source and independent source simultaneously, and there is outstanding noise robustness and the higher estimation probability of success, Measure direction performance is better than the broadband direction-finding method based on particle swarm algorithm.
Description
Technical field
The invention belongs to array signal processing fields, and in particular to a kind of broadband direction finding based on quantum dragonfly mechanism of Evolution
Method.
Background technique
Array signal processing has a wide range of applications in fields such as communication, radar and sonars, and Mutual coupling is battle array
One of the important research direction in column signal field.Broadband signal, which has to carry, to contain much information, and is easy to target echo detection, parameter is estimated
The advantages that meter and feature extraction, and the broadband signals such as Frequency Hopping Signal, spread-spectrum signal and linear FM signal are in a communications system
Application it is also more and more extensive.Therefore, the Mutual coupling for studying broadband signal has more importantly meaning.
Maximum likelihood algorithm is to calculate the covariance matrix and rectangular projection square of signal by handling the data received
Then battle array obtains the objective function of maximum likelihood algorithm by a series of operations.But this objective function is one about entering
The Multidimensional nonlinear function of firing angle degree, solution procedure is complicated, and operand is larger.
Weighted signal subspace fitting algorithm is to acquire letter by carrying out feature decomposition to the covariance matrix for receiving signal
Work song space and noise subspace, then the son using signal subspace and actual signal the steering vector composition for receiving data is empty
Between between fitting acquire objective function.The objective function is also a Multidimensional nonlinear function, and solution procedure is complicated, operand
It is larger.
According to existing document find, Su Chengxiao, Luo Jingqing " electronic information countermeasure techniques " (2014, Vol.29, No.1,
Broadband direction-finding method proposed in " the broadband weak signal direction-finding method based on orthogonal projection transformation " is delivered on pp.16-20, is received
It is lower to hold back precision.Li Xin, Liu Wenhong are sent out on " computer engineering and application " (2013, Vol.49, No.20, pp.227-229)
Method operand proposed in table " the broadband multi-signal direction-finding technique study based on frequency domain correlation and SVM " is larger, convergence essence
Spend lower, performance is poor.
Although the above method yields good result in the direction finding problem of broadband, solving precision is not high, convergence
Energy is poor and operand is larger, it is therefore desirable to design the new broadband direction-finding method of one kind to solve these problems.
Summary of the invention
The object of the present invention is to provide a kind of broadband direction-finding methods based on quantum dragonfly mechanism of Evolution, in wide band loop
To the quick high accuracy direction finding of independent source and coherent under border.
The object of the present invention is achieved like this:
A kind of broadband direction-finding method based on quantum dragonfly mechanism of Evolution, steps are as follows for concrete implementation:
Step 1. establishes Sampling for Wide-Band Signal model；
Step 2. quantum dragonfly mechanism of Evolution parameter initialization；
Step 3. calculates the fitness of every quantum dragonfly, obtains the worst quantum bit in local optimum quantum position drawn game portion
It sets, global optimum's quantum position and global worst quantum position；
Quantum position and amount of the step 4. to quantum dragonfly group the first half more frontier radius and neighborhood quantum dragonfly
Sub- speed, five kinds of behavior vector sum step-length vectors for updating every quantum dragonfly, the quantum rotation angle for updating every quantum dragonfly
Degree and quantum position；
Step 5. updates quantum speed and the quantum position of every quantum dragonfly to quantum dragonfly group later half；
Step 6. calculates the fitness value of all quantum dragonflies position, updates local optimum quantum position drawn game portion most residual quantity
Sub- position, global optimum's quantum position and global worst quantum position；
Step 7. judges whether to reach maximum number of iterations, if not reaching, return step 4 continues；If having reached
It arrives, then global optimum of quantum dragonfly group quantum position is mapped to optimal location, obtains broadband Mutual coupling to be estimated
Angle.
The process that Sampling for Wide-Band Signal model is established described in step 1 is, under Gaussian noise environment, there is P far field
Broadband signal is respectively with deflection θ1,θ2,…,θPIt is incident on the aerial array of space, which is made of M array element,
Array element spacing is d, and wavelength λ, the bandwidth of incoming signal is B.Using first array element as reference array element, then m-th of array element is connect
The signal received is expressed as
Wherein, sp(t) indicate that incident direction is θpBroadband signal, np(t) additive noise in m-th of array element is indicated,
am,pIndicate that p-th of information source is different to the space loss of each different sensors and is embodied in the signal strength in m-th of array element,Indicate that p-th of information source reaches the time delay in m-th of array element；By observing time ToIt is divided into K subsegment, every time is
Td, i.e.,Then observation data are carried outThe discrete Fourier transform of point, obtains the irrelevant narrowband frequency domain of K group
Component, subsegment TdRatioing signal and noise correlation time are longer, then the data after discrete Fourier transform are incoherent, obtain
Wide band model is
It is z respectivelym(t)、sp(t)、nm(t) it is in k-th time subsegment, frequency
When Fourier coefficient,It is the guiding matrix that size is M × P, when P direction is different, matrix is full rank,The referred to as steering vector of matrix
Wherein p=1,2 ..., P；Under the conditions of signal and noise are incoherent, the data that processing array received arrives, every
One Frequency pointThe covariance matrix that sensor array frequency domain sample data are acquired at place is
Acquiring orthogonal intersection cast shadow matrix using the data received is
The angle estimation value acquired according to maximum likelihood equations is
Feature decomposition is carried out to covariance matrix, obtains signal subspaceAnd noise subspaceThen
Space according to the space of signal subspace and array manifold is the same space, acquires weighted signal subspace fitting
The angle estimation value of equation is
Wherein tr indicates to seek the mark of matrix, and weight matrix meets
For frequencyCorresponding noise power,For frequencyCorresponding signal covariance matrix feature point
After solution by big eigenvalue cluster at diagonal matrix.It is designed in conjunction with maximum likelihood equations and weighted signal subspace fitting equation a kind of new
Broadband direction-finding method, maximum likelihood equations and weighted signal subspace fitting equation are combined with different weights,
Obtained angle estimation value is
Wherein w1+w2=1, w1And w2For the weight factor between [0,1].
Quantum dragonfly mechanism of Evolution parameter initialization detailed process described in step 2 is that quantum dragonfly population size is
Maximum number of iterations is G, and search space dimension is P, radius of neighbourhood r, and step-length vector is
WhereinThe weight factor of five kinds of behaviors of quantum dragonfly group is respectively
And the weight factor of step-length vector is w3, the quantum bit of i-th quantum dragonfly is set to
The speed of i-th quantum dragonfly is
WhereinT is the number of iterations, just
Begin season t=1.
The detailed process that every quantum dragonfly fitness is calculated described in step 3 is, in the direction finding of broadband, i-th quantum
The pth of the quantum position of dragonfly ties up the formula being mapped in the direction finding of broadband
Wherein Amax=90 ° are maximum angle, Amin=-90 ° are minimum angles, p=1,2 ..., P；Calculate i-th quantum
The fitness value of dragonfly position, fitness function are
Determine that i-th quantum dragonfly local optimum quantum bit is set to
Locally worst quantum bit is set to i-th quantum dragonfly
And global optimum's quantum bit is set to
Global worst quantum bit is set to
Wherein global optimum's quantum bit is set to food source quantum position, and global worst quantum bit is set to natural enemy quantum position.
The specific steps of step 4 are as follows:
Step 4.1. updates the quantum position vector sum quantum velocity vector of the radius of neighbourhood and neighborhood quantum dragonfly, every amount
Sub- dragonfly is in the center of circle for the circle that radius is r, when the Euclidean distance between two quantum dragonflies is less than the radius of neighbourhood, then it is assumed that
The two is adjacent, otherwise the two is non-conterminous.The radius of neighbourhood is linearly increasing with the increase of the number of iterations, until entire quantum dragonfly
Group is all adjacent, and the more new formula of the radius of neighbourhood is
rt=(Amax-Amin)/4+(Amax-Amin)×t×2/G
The q of i-th quantum dragonfly is only set to adjacent to the quantum bit of quantum dragonfly
WhereinQ is the neighbouring quantum dragonfly sum of i-th quantum dragonfly, update the
The speed of the q of i quantum dragonfly only neighbouring quantum dragonflies is
Step 4.2. updates five kinds of behavior vector sum step-length vectors of quantum dragonfly group, i-th quantum dragonfly collision avoidance row
It is for the more new formula of vector
WhereinThe more new formula of alignment behavior vector is
The more new formula of cohesion behavior vector is
The more new formula of foraging behavior vector is
The more new formula for keeping away enemy's behavior vector is
Update each weight factorWith inertia weight w3, the step-length vector of i-th quantum dragonfly
More new formula is
Step 4.3. updates the quantum rotation door rotation angle degree and quantum position vector of every quantum dragonfly, when quantum dragonfly
When i-th quantum dragonfly in group has neighbouring quantum dragonfly, the pth dimension of quantum rotation door rotation angle degree is WhereinFor the pth dimension of i-th quantum dragonfly step-length vector, its quantum position is more
New formula is
When i-th quantum dragonfly in quantum dragonfly group is without neighbouring quantum dragonfly, which is flown with Le ' vy
Row mode is flown around search space, and the pth dimension of i-th quantum dragonfly quantum rotation door rotation angle degree is
Its quantum location update formula is
Le ' vy function calculation formula is
Wherein r1, r2It is the random number in [0,1], Γ (1+ η) is Gamma function, and its calculation formula is Γ (1+ η)=η！,
η is a constant.
The detailed process of step 5 is the quantum velocity vector and quantum position vector for updating quantum dragonfly, i-th quantum dragonfly
The quantum speed regulation p of dragonfly ties up more new formula
WhereinThe quantum position of i-th quantum dragonfly, pth are tieed up more
New formula is
Wherein w4It is specific gravity shared by previous generation quantum speed, w5And w6Be respectively local optimum quantum position and it is global most
The weight factor of excellent quantum position, c1And c2It is the constant being randomly generated between [0,1].
The detailed process of step 6 is the fitness value for calculating all quantum dragonflies position, if the adaptation of i-th quantum dragonfly
Angle value is greater than the fitness value being saved, then substitutes the fitness value originally saved with the fitness value of i-th quantum dragonfly,
And the local optimum quantum position originally saved is substituted with the quantum position of i-th quantum dragonfly；Find out quantum dragonfly group
Maximum adaptation angle value, it is suitable with current maximum if current maximum adaptation angle value is greater than the maximum adaptation angle value originally saved
Angle value is answered to substitute the maximum adaptation angle value that originally saved, and with the quantum position of that maximum quantum dragonfly of current fitness value
As global optimum's quantum position.
The beneficial effects of the present invention are: the present invention devises quantum dragonfly mechanism of Evolution and carries out direction finding to broadband signal,
Reduce operand and operation time, improve convergence rate and convergence precision, realizes high-precision direction finding；Designed by the present invention
Broadband direction-finding method can carry out Mutual coupling to coherent source and independent source simultaneously, and with outstanding noise robustness and
The higher estimation probability of success；Measure direction performance of the invention is better than the broadband direction-finding method based on particle swarm algorithm.
Detailed description of the invention
Fig. 1 is the broadband direction-finding method flow chart based on quantum dragonfly mechanism of Evolution.
Fig. 2 is independent source root-mean-square error and Between Signal To Noise Ratio curve.
Fig. 3 is coherent source root-mean-square error and Between Signal To Noise Ratio curve.
Specific embodiment
The present invention will be further described with reference to the accompanying drawing:
Embodiment 1
The present invention is combined with the broadband direction-finding method of maximum likelihood equations and weighted signal subspace fitting equation, and
It devises quantum dragonfly mechanism of Evolution to solve objective function, a kind of broadband direction finding side based on quantum dragonfly mechanism of Evolution
Method, steps are as follows for concrete implementation:
Step 1. establishes Sampling for Wide-Band Signal model；
Step 2. quantum dragonfly mechanism of Evolution parameter initialization；
Step 3. calculates the fitness of every quantum dragonfly, obtains the worst quantum bit in local optimum quantum position drawn game portion
It sets, global optimum's quantum position and global worst quantum position；
Quantum position and amount of the step 4. to quantum dragonfly group the first half more frontier radius and neighborhood quantum dragonfly
Sub- speed, five kinds of behavior vector sum step-length vectors for updating every quantum dragonfly, the quantum rotation angle for updating every quantum dragonfly
Degree and quantum position；
Step 5. updates quantum speed and the quantum position of every quantum dragonfly to quantum dragonfly group later half；
Step 6. calculates the fitness value of all quantum dragonflies position, updates local optimum quantum position drawn game portion most residual quantity
Sub- position, global optimum's quantum position and global worst quantum position；
Step 7. judges whether to reach maximum number of iterations, if not reaching, return step 4 continues；If having reached
It arrives, then global optimum of quantum dragonfly group quantum position is mapped to optimal location, obtains broadband Mutual coupling to be estimated
Angle.
In attached drawing, the broadband direction-finding method based on quantum dragonfly mechanism of Evolution is abbreviated as QDA, particle swarm algorithm will be based on
Broadband direction-finding method be abbreviated as PSO.
The process that Sampling for Wide-Band Signal model is established described in step 1 is, under Gaussian noise environment, there is P far field
Broadband signal is respectively with deflection θ1,θ2,…,θPIt is incident on the aerial array of space, which is made of M array element,
Array element spacing is d, and wavelength λ, the bandwidth of incoming signal is B.Using first array element as reference array element, then m-th of array element is connect
The signal received can be expressed as
Wherein, sp(t) indicate that incident direction is θpBroadband signal, np(t) additive noise in m-th of array element is indicated,
am,pIndicate that p-th of information source is different to the space loss of each different sensors and is embodied in the signal strength in m-th of array element,Indicate that p-th of information source reaches the time delay in m-th of array element；By observing time ToIt is divided into K subsegment, every time is
Td, i.e.,Then observation data are carried outThe discrete Fourier transform of point, obtains the irrelevant narrowband frequency domain of K group
Component, subsegment TdRatioing signal and noise correlation time are longer, then the data after discrete Fourier transform are incoherent, obtain
Wide band model is
It is z respectivelym(t)、sp(t)、nm(t) it is in k-th time subsegment, frequency
When Fourier coefficient,It is the guiding matrix that size is M × P, when P direction is different, matrix is full rank,The referred to as steering vector of matrix
Wherein p=1,2 ..., P；Under the conditions of signal and noise are incoherent, the data that processing array received arrives, every
One Frequency pointThe covariance matrix that sensor array frequency domain sample data are acquired at place is
Acquiring orthogonal intersection cast shadow matrix using the data received is
The angle estimation value acquired according to maximum likelihood equations is
Feature decomposition is carried out to covariance matrix, obtains signal subspaceAnd noise subspaceThen
Space according to the space of signal subspace and array manifold is the same space, acquires weighted signal subspace fitting
The angle estimation value of equation is
Wherein tr indicates to seek the mark of matrix, and weight matrix meets
For frequencyCorresponding noise power,For frequencyCorresponding signal covariance matrix feature point
After solution by big eigenvalue cluster at diagonal matrix.It is designed in conjunction with maximum likelihood equations and weighted signal subspace fitting equation a kind of new
Broadband direction-finding method, maximum likelihood equations and weighted signal subspace fitting equation are combined with different weights,
Obtained angle estimation value is
Wherein w1+w2=1, w1And w2For the weight factor between [0,1].
Quantum dragonfly mechanism of Evolution parameter initialization detailed process described in step 2 is that quantum dragonfly population size is
Maximum number of iterations is G, and search space dimension is P, radius of neighbourhood r, and step-length vector is
WhereinThe weight factor of five kinds of behaviors of quantum dragonfly group is respectively
And the weight factor of step-length vector is w3, the quantum bit of i-th quantum dragonfly is set to
The speed of i-th quantum dragonfly is
WhereinT is the number of iterations, just
Begin season t=1.
The detailed process that every quantum dragonfly fitness is calculated described in step 3 is, in the direction finding of broadband, i-th quantum
The pth of the quantum position of dragonfly ties up the formula being mapped in the direction finding of broadband
Wherein Amax=90 ° are maximum angle, Amin=-90 ° are minimum angles, p=1,2 ..., P；Calculate i-th quantum
The fitness value of dragonfly position, fitness function are
Determine that i-th quantum dragonfly local optimum quantum bit is set to
Locally worst quantum bit is set to i-th quantum dragonfly
And global optimum's quantum bit is set to
Global worst quantum bit is set to
Wherein global optimum's quantum bit is set to food source quantum position, and global worst quantum bit is set to natural enemy quantum position.
The specific steps of step 4 are as follows:
Step 4.1. updates the quantum position vector sum quantum velocity vector of the radius of neighbourhood and neighborhood quantum dragonfly, every amount
Sub- dragonfly is in the center of circle for the circle that radius is r, when the Euclidean distance between two quantum dragonflies is less than the radius of neighbourhood, then it is assumed that
The two is adjacent, otherwise the two is non-conterminous.The radius of neighbourhood is linearly increasing with the increase of the number of iterations, until entire quantum dragonfly
Group is all adjacent, and the more new formula of the radius of neighbourhood is
rt=(Amax-Amin)/4+(Amax-Amin)×t×2/G
The q of i-th quantum dragonfly is only set to adjacent to the quantum bit of quantum dragonfly
WhereinQ is the neighbouring quantum dragonfly sum of i-th quantum dragonfly, update the
The speed of the q of i quantum dragonfly only neighbouring quantum dragonflies is
Step 4.2. updates five kinds of behavior vector sum step-length vectors of quantum dragonfly group, i-th quantum dragonfly collision avoidance row
It is for the more new formula of vector
WhereinThe more new formula of alignment behavior vector is
The more new formula of cohesion behavior vector is
The more new formula of foraging behavior vector is
The more new formula for keeping away enemy's behavior vector is
Update each weight factorWith inertia weight w3, the step-length vector of i-th quantum dragonfly
More new formula is
Step 4.3. updates the quantum rotation door rotation angle degree and quantum position vector of every quantum dragonfly, when quantum dragonfly
When i-th quantum dragonfly in group has neighbouring quantum dragonfly, the pth dimension of quantum rotation door rotation angle degree is WhereinFor the pth dimension of i-th quantum dragonfly step-length vector, its quantum position is more
New formula is
When i-th quantum dragonfly in quantum dragonfly group is without neighbouring quantum dragonfly, which is flown with Le ' vy
Row mode is flown around search space, and the pth dimension of i-th quantum dragonfly quantum rotation door rotation angle degree is
Its quantum location update formula is
Le ' vy function calculation formula is
Wherein r1, r2It is the random number in [0,1], Γ (1+ η) is Gamma function, and its calculation formula is Γ (1+ η)=η！,
η is a constant.
The detailed process of step 5 is the quantum velocity vector and quantum position vector for updating quantum dragonfly, i-th quantum dragonfly
The quantum speed regulation p of dragonfly ties up more new formula
WhereinThe quantum position of i-th quantum dragonfly, pth are tieed up more
New formula is
Wherein w4It is specific gravity shared by previous generation quantum speed, w5And w6Be respectively local optimum quantum position and it is global most
The weight factor of excellent quantum position, c1And c2It is the constant being randomly generated between [0,1].
The detailed process of step 6 is the fitness value for calculating all quantum dragonflies position, if the adaptation of i-th quantum dragonfly
Angle value is greater than the fitness value being saved, then substitutes the fitness value originally saved with the fitness value of i-th quantum dragonfly,
And the local optimum quantum position originally saved is substituted with the quantum position of i-th quantum dragonfly；Find out quantum dragonfly group
Maximum adaptation angle value, it is suitable with current maximum if current maximum adaptation angle value is greater than the maximum adaptation angle value originally saved
Angle value is answered to substitute the maximum adaptation angle value that originally saved, and with the quantum position of that maximum quantum dragonfly of current fitness value
As global optimum's quantum position.
Simulation calculation is carried out to the invention, model design parameter is provided that
Broadband far-field signal, low-limit frequency 80Hz, highest frequency 180Hz, aerial array are even linear array, between array element
Away from for half-wavelength, antenna number 8, signal velocity 1500m/s, number of snapshots 1024, information source number 2, signal incidence angle
Degree is respectively 20 °, 10 °, and incoming signal uses linear FM signal, and noise is Gaussian noise.
The parameter setting of broadband direction-finding method based on quantum dragonfly algorithm is as follows: quantum dragonfly population scale
The number of iterations G=60, initial step length vector ξ=0, initial neighborhood radius r=1.5, collision avoidance behavior weight s=0.2 are aligned behavior
Weight a=0.2, cohesion behavior weight c=0.3, food source weight f=1, natural enemy weight z=1 and step-length vector weight w3
η=1.5 in=0.8, Le ' vy function.
The parameter setting of broadband direction-finding method based on particle swarm algorithm: weight factor w4=1, w5=2, w6=2.
Fig. 2 and Fig. 3 is the root-mean-square error and Between Signal To Noise Ratio curve of independent source and coherent source respectively.It can from analogous diagram
To find out, either independent source or coherent source, the root-mean-square error of the broadband direction finding based on quantum dragonfly algorithm will be less than
The root-mean-square error of particle swarm algorithm.
Embodiment 2
The present invention relates to a kind of broadband direction-finding methods based on quantum dragonfly mechanism of Evolution, belong to array signal processing neck
Domain.
Array signal processing has a wide range of applications in fields such as communication, radar and sonars, and Mutual coupling is battle array
One of the important research direction in column signal field.Broadband signal, which has to carry, to contain much information, and is easy to target echo detection, parameter is estimated
The advantages that meter and feature extraction, and the broadband signals such as Frequency Hopping Signal, spread-spectrum signal and linear FM signal are in a communications system
Application it is also more and more extensive.Therefore, the Mutual coupling for studying broadband signal has more importantly meaning.
Maximum likelihood algorithm is to calculate the covariance matrix and rectangular projection square of signal by handling the data received
Then battle array obtains the objective function of maximum likelihood algorithm by a series of operations.But this objective function is one about entering
The Multidimensional nonlinear function of firing angle degree, solution procedure is complicated, and operand is larger.
Weighted signal subspace fitting algorithm is to acquire letter by carrying out feature decomposition to the covariance matrix for receiving signal
Work song space and noise subspace, then the son using signal subspace and actual signal the steering vector composition for receiving data is empty
Between between fitting acquire objective function.The objective function is also a Multidimensional nonlinear function, and solution procedure is complicated, operand
It is larger.
According to existing document find, Su Chengxiao, Luo Jingqing " electronic information countermeasure techniques " (2014, Vol.29, No.1,
Broadband direction-finding method proposed in " the broadband weak signal direction-finding method based on orthogonal projection transformation " is delivered on pp.16-20, is received
It is lower to hold back precision.Li Xin, Liu Wenhong are sent out on " computer engineering and application " (2013, Vol.49, No.20, pp.227-229)
Method operand proposed in table " the broadband multi-signal direction-finding technique study based on frequency domain correlation and SVM " is larger, convergence essence
Spend lower, performance is poor.
Although the above method yields good result in the direction finding problem of broadband, solving precision is not high, convergence
Energy is poor and operand is larger, it is therefore desirable to design the new broadband direction-finding method of one kind to solve these problems.
The invention proposes the broadband direction finding sides of a kind of combination weighted signal subspace fitting equation and maximum likelihood equations
Then method devises quantum dragonfly mechanism of Evolution and solves to the objective function of this method.By simulation result it is found that the party
Method realizes the quick high accuracy direction finding under broadband environment to independent source and coherent.
The present invention is achieved by the following technical solution, and is mainly comprised the steps that
Step 1: under Gaussian noise environment, there is the broadband signal in P far field respectively with deflection θ1,θ2,…,θPIt is incident
Onto space aerial array, which is made of M array element, and array element spacing is d, wavelength λ, the bandwidth of incoming signal
For B.Using first array element as reference array element, then signal received by m-th of array element can be expressed asWherein, sp(t) indicate that incident direction is θpBroadband signal,
np(t) additive noise in m-th of array element, a are indicatedm,pIndicate that p-th of information source is different to the space loss of each different sensors
And it is embodied in the signal strength in m-th of array element,Indicate that p-th of information source reaches the time delay in m-th of array element.
By observing time ToIt is divided into K subsegment, every time is Td, i.e.,Then observation data are carried outPoint
Discrete Fourier transform obtains the irrelevant narrowband frequency domain components of K group, subsegment TdRatioing signal and noise correlation time compared with
Long, then the data after discrete Fourier transform are incoherent, so that it may which obtaining wide band model isIn formula It is z respectivelym(t)、sp(t)、nm(t) it is in k-th time subsegment, frequencyWhen Fourier coefficient.It is the guiding matrix that size is M × P, when P
When direction is different, matrix is full rank；
The referred to as steering vector of matrix.
Under the conditions of signal and noise are incoherent, the data that processing array received arrives, in each Frequency pointIt asks at place
The covariance matrixes of sensor array frequency domain sample data isUsing connecing
The data received acquire orthogonal intersection cast shadow matrixAccording to maximum likelihood side
The angle estimation value that journey acquires isFeature decomposition is carried out to covariance matrix, is obtained
To signal subspaceAnd noise subspaceThen according to the space of signal subspace and array manifold
At space be the same space, the angle estimation value for acquiring weighted signal subspace fitting equation isWherein tr indicates to seek the mark of matrix, and weight matrix meets For frequencyCorresponding noise power,For frequency
After corresponding signal covariance matrix feature decomposition by big eigenvalue cluster at diagonal matrix.In conjunction with maximum likelihood equations and weighting letter
Work song spatial fit equation designs a kind of new broadband direction-finding method, by maximum likelihood equations and weighted signal subspace fitting side
Journey is combined with different weights, and obtained angle estimation value isw1And w2
For the weight factor between [0,1].
Step 2: quantum dragonfly mechanism of Evolution parameter initialization: quantum dragonfly population size isMaximum number of iterations is
G, search space dimension are P, radius of neighbourhood r, and step-length vector isQuantum dragonfly
The weight factor of five kinds of behaviors of group is respectivelyAnd the weight factor of step-length vector is w3, i-th
The quantum bit of quantum dragonfly is set toThe speed of i-th quantum dragonfly isIts
InT is the number of iterations, initial season t=1.
Step 3: the fitness of all quantum dragonflies position is calculated.In the direction finding of broadband, the quantum of i-th quantum dragonfly
The pth of position ties up the formula being mapped in the direction finding of broadbandWherein Amax=90 ° is most
Wide-angle, Amin=-90 ° are minimum angles, p=1,2 ..., P.The fitness value of i-th quantum dragonfly position is calculated, is adapted to
Spending function isDetermine i-th amount
Sub- dragonfly local optimum quantum bit is set toLocally worst quantum bit is set to i-th quantum dragonflyAnd global optimum's quantum bit is set toWith global worst quantum position
ForWherein global optimum's quantum bit is set to food source quantum position, global worst quantum position
For natural enemy quantum position.
Step 4: the first half of quantum dragonfly group proceeds as follows:
1) the quantum position vector sum quantum velocity vector of the radius of neighbourhood and neighborhood quantum dragonfly is updated.Every quantum dragonfly
The center of circle in the circle that radius is r, when the Euclidean distance between two quantum dragonflies is less than the radius of neighbourhood, then it is assumed that the two phase
Neighbour, on the contrary the two is non-conterminous.The radius of neighbourhood is linearly increasing with the increase of the number of iterations, until entire quantum dragonfly group is whole
Adjacent, the more new formula of the radius of neighbourhood is rt=(Amax-Amin)/4+(Amax-Amin)×t×2/G；The q of i-th quantum dragonfly is only
The quantum bit of neighbouring quantum dragonfly is set to
Q is the neighbouring quantum dragonfly sum of i-th quantum dragonfly, updates the q of i-th quantum dragonfly only adjacent to the speed of quantum dragonfly
For
2) five kinds of behavior vector sum step-length vectors of quantum dragonfly group are updated.I-th quantum dragonfly collision avoidance behavior vector
More new formula beThe more new formula of alignment behavior vector is
The more new formula of cohesion behavior vector isThe more new formula of foraging behavior vector isIt keeps away
The more new formula of enemy's behavior vector isUpdate each weight factorAnd inertia weight
w3.The step-length vector more new formula of i-th quantum dragonfly is
3) the quantum rotation door rotation angle degree and quantum position vector of every quantum dragonfly are updated.When in quantum dragonfly group
I-th quantum dragonfly when having neighbouring quantum dragonfly, the pth dimension of quantum rotation door rotation angle degree is WhereinIt is tieed up for the pth of i-th quantum dragonfly step-length vector, it
Quantum location update formula isI-th in quantum dragonfly group
When quantum dragonfly is without neighbouring quantum dragonfly, which is flown with Le ' vy offline mode around search space, i-th quantum
The pth of dragonfly quantum rotation door rotation angle degree is tieed upIts quantum location update formula isLe ' vy function calculation formula is Wherein r1, r2It is the random number in [0,1], Γ (1+ η) is Gamma function, is calculated public
Formula is Γ (1+ η)=η！, η is a constant.
Step 5: the later half of quantum dragonfly group proceeds as follows:
Update the quantum velocity vector and quantum position vector of quantum dragonfly.The quantum speed regulation p dimension of i-th quantum dragonfly is more
New formula is
The quantum position pth of i-th quantum dragonfly ties up more new formulaIts
Middle w4It is specific gravity shared by previous generation quantum speed, w5And w6It is local optimum quantum position and global optimum's quantum position respectively
Weight factor, c1And c2It is the constant being randomly generated between [0,1].
Step 6: calculating the fitness value of all quantum dragonflies position, if the fitness value of i-th quantum dragonfly is greater than
The fitness value being saved then substitutes the fitness value originally saved with the fitness value of i-th quantum dragonfly, and with i-th
The quantum position of quantum dragonfly substitutes the local optimum quantum position originally saved；Find out the maximum adaptation degree of quantum dragonfly group
Value is substituted if current maximum adaptation angle value is greater than the maximum adaptation angle value originally saved with current maximum adaptation angle value
The maximum adaptation angle value originally saved, and use the quantum position of that maximum quantum dragonfly of current fitness value as it is global most
Excellent quantum position.
Step 7: judging whether to reach maximum number of iterations, if not reaching, return step four continues；If
Reach, then global optimum of quantum dragonfly group quantum position is mapped to optimal location, just obtains broadband Mutual coupling and is wanted
The angle of estimation.
Compared with prior art, the invention has the following advantages that
(1) present invention devises quantum dragonfly mechanism of Evolution and carries out direction finding to broadband signal, reduces operand, improves
Convergence rate and convergence precision.
(2) the broadband direction-finding method designed by the present invention can carry out Mutual coupling to coherent source and independent source simultaneously,
And there is outstanding noise robustness and the higher estimation probability of success.
(3) simulation result shows that Measure direction performance of the invention is better than the broadband direction-finding method based on particle swarm algorithm.
Fig. 1 is the broadband direction-finding method flow chart based on quantum dragonfly mechanism of Evolution.The present invention is combined with maximum likelihood
The broadband direction-finding method of equation and weighted signal subspace fitting equation, and quantum dragonfly mechanism of Evolution is devised to target letter
Number is solved.It is of the present invention solve the problems, such as used by protocol step it is as follows:
Step 1: under Gaussian noise environment, there is the broadband signal in P far field respectively with deflection θ1,θ2,…,θPIt is incident
Onto space aerial array, which is made of M array element, and array element spacing is d, wavelength λ, the bandwidth of incoming signal
For B.Using first array element as reference array element, then signal received by m-th of array element can be expressed asWherein, sp(t) indicate that incident direction is θpBroadband signal,
np(t) additive noise in m-th of array element, a are indicatedm,pIndicate that p-th of information source is different to the space loss of each different sensors
And it is embodied in the signal strength in m-th of array element,Indicate that p-th of information source reaches the time delay in m-th of array element.
By observing time ToIt is divided into K subsegment, every time is Td, i.e.,Then observation data are carried outPoint
Discrete Fourier transform obtains the irrelevant narrowband frequency domain components of K group, subsegment TdRatioing signal and noise correlation time compared with
Long, then the data after discrete Fourier transform are incoherent, so that it may which obtaining wide band model isIn formula It is z respectivelym(t)、sp(t)、nm(t) it is in k-th time subsegment, frequencyWhen Fourier coefficient.It is the guiding matrix that size is M × P, when P
When direction is different, matrix is full rank；
The referred to as steering vector of matrix.
Under the conditions of signal and noise are incoherent, the data that processing array received arrives, in each Frequency pointIt asks at place
The covariance matrixes of sensor array frequency domain sample data isUsing connecing
The data received acquire orthogonal intersection cast shadow matrixAccording to maximum likelihood side
The angle estimation value that journey acquires isFeature decomposition is carried out to covariance matrix, is obtained
To signal subspaceAnd noise subspaceThen according to the space of signal subspace and array manifold
At space be the same space, the angle estimation value for acquiring weighted signal subspace fitting equation isWherein tr indicates to seek the mark of matrix, and weight matrix meets For frequencyCorresponding noise power,For frequency
After corresponding signal covariance matrix feature decomposition by big eigenvalue cluster at diagonal matrix.In conjunction with maximum likelihood equations and weighting letter
Work song spatial fit equation designs a kind of new broadband direction-finding method, by maximum likelihood equations and weighted signal subspace fitting side
Journey is combined with different weights, and obtained angle estimation value isw1And w2
For the weight factor between [0,1].
Step 2: quantum dragonfly mechanism of Evolution parameter initialization: quantum dragonfly population size isMaximum number of iterations
For G, search space dimension is P, radius of neighbourhood r, and step-length vector isQuantum dragonfly
The weight factor of five kinds of behaviors of dragonfly group is respectivelyAnd the weight factor of step-length vector is w3, i-th
The quantum bit of quantum dragonfly is set toThe speed of i-th quantum dragonfly is
WhereinT is the number of iterations, initial season t=1.
Step 3: the fitness of all quantum dragonflies position is calculated.In the direction finding of broadband, the quantum of i-th quantum dragonfly
The pth of position ties up the formula being mapped in the direction finding of broadbandWherein Amax=90 ° is most
Wide-angle, Amin=-90 ° are minimum angles, p=1,2 ..., P.The fitness value of i-th quantum dragonfly position is calculated, is adapted to
Spending function isDetermine i-th amount
Sub- dragonfly local optimum quantum bit is set toLocally worst quantum bit is set to i-th quantum dragonflyAnd global optimum's quantum bit is set toWith global worst quantum position
ForWherein global optimum's quantum bit is set to food source quantum position, global worst quantum position
For natural enemy quantum position.
Step 4: the first half of quantum dragonfly group proceeds as follows:
1) the quantum position vector sum quantum velocity vector of the radius of neighbourhood and neighborhood quantum dragonfly is updated.Every quantum dragonfly
The center of circle in the circle that radius is r, when the Euclidean distance between two quantum dragonflies is less than the radius of neighbourhood, then it is assumed that the two phase
Neighbour, on the contrary the two is non-conterminous.The radius of neighbourhood is linearly increasing with the increase of the number of iterations, until entire quantum dragonfly group is whole
Adjacent, the more new formula of the radius of neighbourhood is rt=(Amax-Amin)/4+(Amax-Amin)×t×2/G；The q of i-th quantum dragonfly is only
The quantum bit of neighbouring quantum dragonfly is set to
Q is the neighbouring quantum dragonfly sum of i-th quantum dragonfly, updates the q of i-th quantum dragonfly only adjacent to the speed of quantum dragonfly
For
2) five kinds of behavior vector sum step-length vectors of quantum dragonfly group are updated.I-th quantum dragonfly collision avoidance behavior vector
More new formula isThe more new formula of alignment behavior vector is
The more new formula of cohesion behavior vector isThe more new formula of foraging behavior vector isKeep away enemy
The more new formula of behavior vector isUpdate each weight factorWith inertia weight w3。
The step-length vector more new formula of i-th quantum dragonfly is
3) the quantum rotation door rotation angle degree and quantum position vector of every quantum dragonfly are updated.When in quantum dragonfly group
When i-th quantum dragonfly has neighbouring quantum dragonfly, the pth dimension of quantum rotation door rotation angle degree is WhereinFor the pth dimension of i-th quantum dragonfly step-length vector, its quantum location update formula isWhen i-th quantum dragonfly in quantum dragonfly group is without neighbouring amount
When sub- dragonfly, which is flown with Le ' vy offline mode around search space, i-th quantum dragonfly Quantum rotating gate rotation
The pth of angle is tieed upIts quantum location update formula isLe ' vy function calculation formula is Wherein r1, r2It is the random number in [0,1], Γ (1+ η) is Gamma function, is calculated public
Formula is Γ (1+ η)=η！, η is a constant.
Step 5: the later half of quantum dragonfly group proceeds as follows:
Update the quantum velocity vector and quantum position vector of quantum dragonfly.The quantum speed regulation p dimension of i-th quantum dragonfly is more
New formula is
The quantum position pth of i-th quantum dragonfly ties up more new formulaIts
Middle w4It is specific gravity shared by previous generation quantum speed, w5And w6It is local optimum quantum position and global optimum's quantum position respectively
Weight factor, c1And c2It is the constant being randomly generated between [0,1].
Step 6: calculating the fitness value of all quantum dragonflies position, if the fitness value of i-th quantum dragonfly is greater than
The fitness value being saved then substitutes the fitness value originally saved with the fitness value of i-th quantum dragonfly, and with i-th
The quantum position of quantum dragonfly substitutes the local optimum quantum position originally saved；Find out the maximum adaptation degree of quantum dragonfly group
Value is substituted if current maximum adaptation angle value is greater than the maximum adaptation angle value originally saved with current maximum adaptation angle value
The maximum adaptation angle value originally saved, and use the quantum position of that maximum quantum dragonfly of current fitness value as it is global most
Excellent quantum position.
Step 7: judging whether to reach maximum number of iterations, if not reaching, return step four continues；If
Reach, then global optimum of quantum dragonfly group quantum position is mapped to optimal location, just obtains broadband Mutual coupling and is wanted
The angle of estimation.
Model design parameter is provided that
Broadband far-field signal, low-limit frequency 80Hz, highest frequency 180Hz, aerial array are even linear array, between array element
Away from for half-wavelength, antenna number 8, signal velocity 1500m/s, number of snapshots 1024, information source number 2, signal incidence angle
Degree is respectively 20 °, 10 °, and incoming signal uses linear FM signal, and noise is Gaussian noise.
The parameter setting of broadband direction-finding method based on quantum dragonfly algorithm is as follows: quantum dragonfly population scale
The number of iterations G=60, initial step length vector ξ=0, initial neighborhood radius r=1.5, collision avoidance behavior weight s=0.2 are aligned behavior
Weight a=0.2, cohesion behavior weight c=0.3, food source weight f=1, natural enemy weight z=1 and step-length vector weight w3
η=1.5 in=0.8, Le ' vy function.
The parameter setting of broadband direction-finding method based on particle swarm algorithm: weight factor w4=1, w5=2, w6=2.
Fig. 2 and Fig. 3 is the root-mean-square error and Between Signal To Noise Ratio curve of independent source and coherent source respectively.It can from analogous diagram
To find out, either independent source or coherent source, the root-mean-square error of the broadband direction finding based on quantum dragonfly algorithm will be less than
The root-mean-square error of particle swarm algorithm.
The present invention carries out Mutual coupling to broadband signal using quantum dragonfly mechanism of Evolution, solves conventional method and deposits
Solution procedure it is complicated, the disadvantages of operand is big, and convergence precision is not high.Steps of the method are establish broadband signal mathematics to adopt
Original mold type；Initialize quantum dragonfly population；The fitness for calculating every quantum dragonfly obtains local optimum quantum position drawn game portion
Worst quantum position, global optimum's quantum position and global worst quantum position；Update the radius of neighbourhood and neighborhood quantum dragonfly
Quantum position and quantum speed, update the behavior vector sum step-length vector of every quantum dragonfly, update every quantum dragonfly
Quantum rotation door rotation angle degree and quantum position；Update quantum rotation angle and the quantum position of every quantum dragonfly；It calculates every
The fitness of quantum dragonfly updates the worst quantum position in local optimum quantum position drawn game portion and global optimum's quantum position
With global worst quantum position；Judge whether to reach maximum number of iterations；Output global optimum's quantum position is simultaneously mapped to broadband
In direction finding.The present invention carries out direction finding to broadband signal with quantum dragonfly mechanism of Evolution, reduces operand and operation time, obtains
Higher convergence precision and faster convergence rate, some problems present in the traditional direction-finding method of effective solution are real
Existing high-precision direction finding.
Claims (7)
1. a kind of broadband direction-finding method based on quantum dragonfly mechanism of Evolution, which is characterized in that steps are as follows for concrete implementation:
Step 1. establishes Sampling for Wide-Band Signal model；
Step 2. quantum dragonfly mechanism of Evolution parameter initialization；
Step 3. calculates the fitness of every quantum dragonfly, obtains the worst quantum position in local optimum quantum position drawn game portion, entirely
The optimal quantum position of office and global worst quantum position；
Step 4. is to the quantum position of quantum dragonfly group the first half more frontier radius and neighborhood quantum dragonfly and quantum speed
Spend, update five kinds of behavior vector sum step-length vectors of every quantum dragonfly, update every quantum dragonfly quantum rotation angle and
Quantum position；
Step 5. updates quantum speed and the quantum position of every quantum dragonfly to quantum dragonfly group later half；
Step 6. calculates the fitness value of all quantum dragonflies position, updates the worst quantum bit in local optimum quantum position drawn game portion
It sets, global optimum's quantum position and global worst quantum position；
Step 7. judges whether to reach maximum number of iterations, if not reaching, return step 4 continues；If having reached,
Quantum dragonfly group global optimum's quantum position is mapped to optimal location, obtains the angle to be estimated of broadband Mutual coupling
Degree.
2. a kind of broadband direction-finding method based on quantum dragonfly mechanism of Evolution according to claim 1, it is characterised in that: step
The process that Sampling for Wide-Band Signal model is established described in rapid 1 is, under Gaussian noise environment, there is the broadband signal point in P far field
Not with deflection θ1,θ2,…,θPIt is incident on the aerial array of space, which is made of M array element, and array element spacing is
D, wavelength λ, the bandwidth of incoming signal are B.Using first array element as reference array element, then signal received by m-th of array element
It is expressed as
Wherein, sp(t) indicate that incident direction is θpBroadband signal, np(t) additive noise in m-th of array element, a are indicatedm,pTable
The space loss for showing p-th of information source to each different sensors is different and is embodied in the signal strength in m-th of array element,It indicates
P-th of information source reaches the time delay in m-th of array element；By observing time ToIt is divided into K subsegment, every time is Td, i.e.,Then observation data are carried outThe discrete Fourier transform of point, obtains the irrelevant narrowband frequency domain components of K group,
Subsegment TdRatioing signal and noise correlation time are longer, then the data after discrete Fourier transform are incoherent, obtain broadband
Model is
It is z respectivelym(t)、sp(t)、nm(t) it is in k-th time subsegment, frequencyWhen Fu
In leaf system number,It is the guiding matrix that size is M × P, when P direction is different, matrix is full rank,Claim
For the steering vector of matrix
Wherein p=1,2 ..., P；Under the conditions of signal and noise are incoherent, the data that processing array received arrives, at each
Frequency pointThe covariance matrix that sensor array frequency domain sample data are acquired at place is
Acquiring orthogonal intersection cast shadow matrix using the data received is
The angle estimation value acquired according to maximum likelihood equations is
Feature decomposition is carried out to covariance matrix, obtains signal subspaceAnd noise subspaceThen according to letter
The space in work song space and the space of array manifold are the same space, acquire weighted signal subspace fitting equation
Angle estimation value is
Wherein tr indicates to seek the mark of matrix, and weight matrix meets
For frequencyCorresponding noise power,For frequencyAfter corresponding signal covariance matrix feature decomposition
By big eigenvalue cluster at diagonal matrix.A kind of new width is designed in conjunction with maximum likelihood equations and weighted signal subspace fitting equation
Band direction-finding method, maximum likelihood equations and weighted signal subspace fitting equation are combined with different weights, obtained
Angle estimation value be
Wherein w1+w2=1, w1And w2For the weight factor between [0,1].
3. a kind of broadband direction-finding method based on quantum dragonfly mechanism of Evolution according to claim 1, it is characterised in that: step
Quantum dragonfly mechanism of Evolution parameter initialization detailed process described in rapid 2 is that quantum dragonfly population size isGreatest iteration time
Number is G, and search space dimension is P, radius of neighbourhood r, and step-length vector is
WhereinThe weight factor of five kinds of behaviors of quantum dragonfly group is respectivelyAnd
The weight factor of step-length vector is w3, the quantum bit of i-th quantum dragonfly is set to
The speed of i-th quantum dragonfly is
WhereinT is the number of iterations, when initial
Enable t=1.
4. a kind of broadband direction-finding method based on quantum dragonfly mechanism of Evolution according to claim 1, it is characterised in that: step
The detailed process that every quantum dragonfly fitness is calculated described in rapid 3 is, in the direction finding of broadband, the quantum of i-th quantum dragonfly
The pth of position ties up the formula being mapped in the direction finding of broadband
Wherein Amax=90 ° are maximum angle, Amin=-90 ° are minimum angles, p=1,2 ..., P；Calculate i-th quantum dragonfly
The fitness value of position, fitness function are
Determine that i-th quantum dragonfly local optimum quantum bit is set to
Locally worst quantum bit is set to i-th quantum dragonfly
And global optimum's quantum bit is set to
Global worst quantum bit is set to
Wherein global optimum's quantum bit is set to food source quantum position, and global worst quantum bit is set to natural enemy quantum position.
5. a kind of broadband direction-finding method based on quantum dragonfly mechanism of Evolution according to claim 1, which is characterized in that step
Rapid 4 specific steps are as follows:
Step 4.1. updates the quantum position vector sum quantum velocity vector of the radius of neighbourhood and neighborhood quantum dragonfly, every quantum dragonfly
Dragonfly is in the center of circle for the circle that radius is r, when the Euclidean distance between two quantum dragonflies is less than the radius of neighbourhood, then it is assumed that the two
It is adjacent, on the contrary the two is non-conterminous.The radius of neighbourhood is linearly increasing with the increase of the number of iterations, until entire quantum dragonfly group
All adjacent, the more new formula of the radius of neighbourhood is
rt=(Amax-Amin)/4+(Amax-Amin)×t×2/G
The q of i-th quantum dragonfly is only set to adjacent to the quantum bit of quantum dragonfly
WhereinQ is the neighbouring quantum dragonfly sum of i-th quantum dragonfly, updates i-th
The speed of the q of quantum dragonfly only neighbouring quantum dragonflies is
Step 4.2. update quantum dragonfly group five kinds of behavior vector sum step-length vectors, i-th quantum dragonfly collision avoidance behavior to
Amount more new formula be
WhereinThe more new formula of alignment behavior vector is
The more new formula of cohesion behavior vector is
The more new formula of foraging behavior vector is
The more new formula for keeping away enemy's behavior vector is
Update each weight factorWith inertia weight w3, the step-length vector of i-th quantum dragonfly updates public
Formula is
Step 4.3. updates the quantum rotation door rotation angle degree and quantum position vector of every quantum dragonfly, when quantum dragonfly group
In i-th quantum dragonfly when having neighbouring quantum dragonfly, the pth dimension of quantum rotation door rotation angle degree is P=1,2 ..., P, whereinFor the pth dimension of i-th quantum dragonfly step-length vector, its quantum position is more
New formula is
When i-th quantum dragonfly in quantum dragonfly group is without neighbouring quantum dragonfly, the quantum dragonfly is with Le ' vy flight mould
Formula is flown around search space, and the pth dimension of i-th quantum dragonfly quantum rotation door rotation angle degree isIt
Quantum location update formula be
Le ' vy function calculation formula is
Wherein r1, r2It is the random number in [0,1], Γ (1+ η) is Gamma function, and its calculation formula is Γ (1+ η)=η！, η is
One constant.
6. a kind of broadband direction-finding method based on quantum dragonfly mechanism of Evolution according to claim 1, it is characterised in that: step
Rapid 5 detailed process is the quantum velocity vector and quantum position vector for updating quantum dragonfly, the quantum speed of i-th quantum dragonfly
Degree pth ties up more new formula
WhereinP=1,2 ..., P, the quantum position of i-th quantum dragonfly, pth tie up more new formula
For
Wherein w4It is specific gravity shared by previous generation quantum speed, w5And w6It is local optimum quantum position and global optimum's amount respectively
The weight factor of sub- position, c1And c2It is the constant being randomly generated between [0,1].
7. a kind of broadband direction-finding method based on quantum dragonfly mechanism of Evolution according to claim 1, it is characterised in that: step
Rapid 6 detailed process is the fitness value for calculating all quantum dragonflies position, if the fitness value of i-th quantum dragonfly is greater than
The fitness value being saved then substitutes the fitness value originally saved with the fitness value of i-th quantum dragonfly, and with i-th
The quantum position of quantum dragonfly substitutes the local optimum quantum position originally saved；Find out the maximum adaptation degree of quantum dragonfly group
Value is substituted if current maximum adaptation angle value is greater than the maximum adaptation angle value originally saved with current maximum adaptation angle value
The maximum adaptation angle value originally saved, and use the quantum position of that maximum quantum dragonfly of current fitness value as it is global most
Excellent quantum position.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811017243.0A CN109212466B (en) | 2018-09-01 | 2018-09-01 | Quantum dragonfly evolution mechanism-based broadband direction finding method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811017243.0A CN109212466B (en) | 2018-09-01 | 2018-09-01 | Quantum dragonfly evolution mechanism-based broadband direction finding method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109212466A true CN109212466A (en) | 2019-01-15 |
| CN109212466B CN109212466B (en) | 2022-09-02 |
Family
ID=64986501
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811017243.0A Active CN109212466B (en) | 2018-09-01 | 2018-09-01 | Quantum dragonfly evolution mechanism-based broadband direction finding method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109212466B (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109669156A (en) * | 2019-02-21 | 2019-04-23 | 哈尔滨工程大学 | Circle battle array model space dynamic direction-finding method under impact noise based on quantum emperor butterfly |
| CN112014789A (en) * | 2020-08-14 | 2020-12-01 | 哈尔滨工程大学 | Composite weighted time-frequency direction-finding method based on quantum dot dog mechanism |
| CN114355768A (en) * | 2020-10-13 | 2022-04-15 | 天津工业大学 | Continuous annealing process robust optimization method based on improved multi-target dragonfly algorithm |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2011252917A (en) * | 2011-07-20 | 2011-12-15 | Japan Atomic Energy Agency | Method for arranging marker in on-site operation support system |
| WO2015048653A1 (en) * | 2013-09-27 | 2015-04-02 | The Johns Hopkins University | Methods of identifying, assessing, preventing, and treating cardiac disorders using gsk-3b and phosphorylation targets thereof |
| CN106249244A (en) * | 2016-07-07 | 2016-12-21 | 哈尔滨工程大学 | A kind of acoustic vector circle battle array wide-band coherent signal source direction estimation method |
| CN107294769A (en) * | 2017-05-16 | 2017-10-24 | 南京邮电大学 | A kind of Agility cloud service management system and its control method based on 5G networks |
| CN107703751A (en) * | 2017-10-13 | 2018-02-16 | 河南工程学院 | PID controller optimization method based on dragonfly algorithm |
-
2018
- 2018-09-01 CN CN201811017243.0A patent/CN109212466B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2011252917A (en) * | 2011-07-20 | 2011-12-15 | Japan Atomic Energy Agency | Method for arranging marker in on-site operation support system |
| WO2015048653A1 (en) * | 2013-09-27 | 2015-04-02 | The Johns Hopkins University | Methods of identifying, assessing, preventing, and treating cardiac disorders using gsk-3b and phosphorylation targets thereof |
| CN106249244A (en) * | 2016-07-07 | 2016-12-21 | 哈尔滨工程大学 | A kind of acoustic vector circle battle array wide-band coherent signal source direction estimation method |
| CN107294769A (en) * | 2017-05-16 | 2017-10-24 | 南京邮电大学 | A kind of Agility cloud service management system and its control method based on 5G networks |
| CN107703751A (en) * | 2017-10-13 | 2018-02-16 | 河南工程学院 | PID controller optimization method based on dragonfly algorithm |
Non-Patent Citations (3)
| Title |
|---|
| WANG MINGLI等: ""Au-Decorated Dragonfly Wing Bioscaffold Arrays as Flexible Surface-Enhanced Raman Scattering (SERS) Substrate for Simultaneous Determination of Pesticide Residues"", 《NANOMATERIALS》 * |
| 张讲社等: "一类模拟生物复眼几何结构的数据处理方法――方向量子化表示的理论", 《计算机学报》 * |
| 陈娟等: "一种用于宽带信号测向的改进TCT算法", 《应用科技》 * |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109669156A (en) * | 2019-02-21 | 2019-04-23 | 哈尔滨工程大学 | Circle battle array model space dynamic direction-finding method under impact noise based on quantum emperor butterfly |
| CN109669156B (en) * | 2019-02-21 | 2021-05-11 | 哈尔滨工程大学 | Quantum emperor butterfly-based circular array mode space dynamic direction finding method under impact noise |
| CN112014789A (en) * | 2020-08-14 | 2020-12-01 | 哈尔滨工程大学 | Composite weighted time-frequency direction-finding method based on quantum dot dog mechanism |
| CN112014789B (en) * | 2020-08-14 | 2022-07-15 | 哈尔滨工程大学 | Composite weighted time-frequency direction-finding method based on quantum dot dog mechanism |
| CN114355768A (en) * | 2020-10-13 | 2022-04-15 | 天津工业大学 | Continuous annealing process robust optimization method based on improved multi-target dragonfly algorithm |
| CN114355768B (en) * | 2020-10-13 | 2023-12-15 | 天津工业大学 | Continuous annealing process robust optimization method based on improved multi-target dragonfly algorithm |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109212466B (en) | 2022-09-02 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN104537249B (en) | Direction of arrival angle method of estimation based on management loading | |
| CN103902826B (en) | Method for tracking multiple moving targets under impact noise environment | |
| CN103901395B (en) | Coherent signal direction of arrival Dynamic Tracking under a kind of impulsive noise environment | |
| CN109752710B (en) | Rapid target angle estimation method based on sparse Bayesian learning | |
| CN111398902B (en) | Passive distance and direction measuring method for right-angled triangle ternary hydrophone array | |
| CN109212466A (en) | A kind of broadband direction-finding method based on quantum dragonfly mechanism of Evolution | |
| CN108872971B (en) | Target positioning method and device based on motion single array | |
| CN108663654B (en) | 360-degree all-dimensional dynamic direction finding method based on continuous quantum pigeon group | |
| CN108845325A (en) | Towed linear-array sonar submatrix error misfits estimation method | |
| CN109239646B (en) | Two-dimensional dynamic direction finding method for continuous quantum water evaporation in impact noise environment | |
| CN105445718A (en) | DOA estimation method for distributed multi-carrier-warship BVR (beyond visual range) radar based on array reconfiguration | |
| CN104793177B (en) | Microphone array direction-finding method based on least square method | |
| CN109669156A (en) | Circle battle array model space dynamic direction-finding method under impact noise based on quantum emperor butterfly | |
| CN101644760A (en) | Rapid and robust method for detecting information source number suitable for high-resolution array | |
| CN109782218A (en) | A kind of non-circular signal DOA estimation method of relevant distribution based on double parallel antenna array | |
| CN106501801A (en) | A kind of bistatic MIMO radar tracking based on chaos Symbiotic evolution on multiple populations | |
| CN109932698A (en) | The low elevation estimate method of metre wave radar based on terrain information | |
| CN109597048A (en) | Metre wave radar DOA estimation method based on two-dimensional convolution neural network | |
| CN108020811A (en) | The 1 dimension uniform linear array direction-finding method based on target source phase shift differential technique | |
| CN108445445B (en) | Distributed passive positioning method and device | |
| CN109358313A (en) | A kind of broadband direction-finding method based on quantum electrified system search mechanism of Evolution | |
| CN109917330A (en) | A kind of angle-of- arrival estimation method there are based on sparse orthogonal matching pursuit theory when phase error | |
| CN113255171A (en) | Direction finding error correction method and device based on transfer learning | |
| Zhang et al. | ML iterative superresolution approach for real-beam radar | |
| CN110412501A (en) | Based on radar signal direction-finding method, device and the computer storage medium improved under polar coordinate representation method model |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |