CN-122017721-A - Small snapshot direction-finding method and system for unmanned aerial vehicle group to offshore target

CN122017721ACN 122017721 ACN122017721 ACN 122017721ACN-122017721-A

Abstract

The invention discloses a small snapshot direction-finding method and a small snapshot direction-finding system for an offshore target by an unmanned aerial vehicle group, relates to the field of array signal processing, and aims to solve the problems that the direction-finding precision and robustness of the existing minimum redundant array are insufficient and are easily affected by model errors under the condition of small snapshot. The method comprises the technical key points of establishing a small snapshot receiving signal model of the unmanned aerial vehicle group for the marine target, establishing a maximum likelihood estimation objective function, initializing a vernonia group, constructing an adaptability function based on the established maximum likelihood estimation objective function, calculating an adaptability value of the vernonia group and determining a local optimal position and a global optimal position, updating the vernonia group, calculating the newly generated position adaptability of each vernonia individual, updating the local optimal position and the global optimal position, screening elite individuals, and carrying out iterative updating on belief space, and finally obtaining the optimal solution of the marine target incoming wave direction.

Inventors

GAO HONGYUAN
CHU YIFAN
ZHU QINGLIN
WANG JIAYI
YUAN SHANLIANG
WANG YUFENG
LI BIAO
GU Yu

Assignees

哈尔滨工程大学

Dates

Publication Date: 20260512
Application Date: 20260130

Claims (8)

1. The small snapshot direction finding method for the offshore targets by the unmanned aerial vehicle group is characterized by comprising the following steps of: The method comprises the steps of establishing a small snapshot receiving signal model of an unmanned aerial vehicle group minimum redundant array for an offshore target, constructing a fractional low-order covariance matrix by utilizing small snapshot data received by the minimum redundant array formed by a plurality of unmanned aerial vehicle groups, and establishing a maximum likelihood estimating target function; Step two: setting the types of the vernonia group as basic vernonia group reverse vernonia and vernonia cultura, initializing a vernonia group and setting related parameters; step three, constructing an adaptability function based on the established maximum likelihood estimation objective function, calculating the adaptability value of the vernonia group and determining a local optimal position and a global optimal position; step four, respectively updating three types of populations of basic vernonia, reverse vernonia and cultural vernonia; Calculating the new generation position fitness of each vernonia individual, updating the local optimal position and the global optimal position, screening elite individuals with better fitness in the group through an acceptance function, and iteratively updating belief space according to a cultural evolution mechanism; Step six, judging whether the preset maximum iteration number is reached, if the preset maximum iteration number is not reached, making the iteration number And returning to the fourth step, otherwise, outputting the global optimal position of the vernonia group as the optimal solution of the sea target incoming wave direction.
2. The method of claim 1, wherein step one includes the process of: constructing a quilt A signal model is received by small shots of the offshore targets of the minimum redundant array of unmanned aerial vehicles formed by array elements, each unmanned aerial vehicle is provided with 1 array element, The unmanned aerial vehicle forms a special array group, and the array element position set is At half wavelength In units of the number of units of the formula, Represent the first The relative positions of the array elements relative to the reference array elements satisfy the following conditions , Array element position difference set , Is the minimum spacing between two array elements, When there is From the sea surface by means of a far-field narrowband signal In all directions When the laser beam enters the minimum redundant array of the unmanned aerial vehicle group, the array is at the first position The sampling data received in the secondary snapshot is Wherein Representation of The dimension array receives an amount of data, Representation of The vector of the dimensional space signal, Representation of The vector of the vickers noise is set, Representation of A vector matrix with a corresponding direction of arrival angle Is the steering vector of the incident signal of (a) , Under the condition of small snapshot, only a limited time sampling point can be obtained, and the snapshot number is recorded as And meets the conditions , The sub-snapshot receiving matrix is expressed as Wherein Representing a matrix of source signals and, Representing an impulse noise matrix; by means of The small snapshot data received by the minimum redundant array formed by the unmanned aerial vehicle group is constructed into a fractional low-order covariance matrix, which is expressed as , Is that The dimension matrix is used to determine the dimensions of the matrix, , First, the Line 1 Column elements are , , , Normalized to infinity Second snapshot data The dimensions of the dimensions, Is a low-order matrix parameter; wherein , ; Weighting the infinity norm of the minimum redundant array by a fractional lower order matrix Infinite norm weighted score low-order matrix extended to virtual uniform array , Wherein , , To expand the number of array elements of the virtually uniform array, , , Representing the mathematical expectation of the solution, , The guiding matrix of the virtual uniform array is The direction of arrival angle is Is the virtual steering vector of the incident signal of (a) ; Obtaining a corresponding maximum likelihood estimation objective function according to the infinity norm weighted score low-order matrix and the guide vector matrix , , The matrix tracing function is represented by a matrix, Representing the steering matrix.
3. The method of claim 2, wherein step two comprises the following process: Setting the size of the vernonia group as In the following Generating three types of vernonia basic groups, reverse vernonia reverse groups and culture vernonia basic groups in the dimension search space, the number is respectively 、 And And (2) and First, a third step The vernonia has Only the root of vernonia is provided, The maximum iteration number is First, a third step At the time of iteration, the first The first vernonia of the group Position of vernonia ganensis only Its speed is Generating an initial belief space and initializing standard knowledge according to rules of a cultural algorithm by using a culture vernonia group to define Represent the first The lower bound of the dimension-specific knowledge, Represent the first The upper bound of the dimension-wise canonical knowledge, Represent the first Substitute for the first Lower bound of dimension specification knowledge The corresponding evaluation value is used for evaluating the quality of the product, Represent the first Substitute for the first Upper bound of dimension specification knowledge The corresponding evaluation value is used for evaluating the quality of the product, And The corresponding boundary values of the variable definition fields are taken respectively, And All the initialization at the time of the first generation are 。
4. A method according to claim 3, characterized in that step three comprises the following procedure: First, the At the time of iteration, the first The first vernonia of the group Position of vernonia ganensis only The fitness function of (2) is And determining each individual up to the first based on the fitness value Substitute local optimum position The whole population is up to the first The substitute is that the global optimal position is 。
5. The method of claim 4, wherein step four comprises the following process: in the basic vernonia, the first Only the vernonia updates the speed and the position according to the basic operator, and the update equation is as follows: Wherein the method comprises the steps of , Representing the map and compass factors, Represents a chaotic random number generated by a chaotic equation, Updating according to the following rules: Wherein the method comprises the steps of Is a uniform random number between 0 and 1; In the reverse vernonia gandersonii group, the first The position of the vernonia barcoo is updated by reverse jump through jump probability Determining whether to perform population jump on the updated individuals, wherein the updating formula is as follows: , In the middle of Is a uniform random number between 0 and 1, , ; The culture vernonia group generates a new vernonia individual according to the influence function, the influence function of the position variable, the change step length and the advancing direction is regulated together according to the standard knowledge of a culture mechanism and the local optimal position, and the update formula is as follows: Wherein the method comprises the steps of Represents the vernonia cultura scaling factor, Representing random numbers satisfying a standard normal distribution.
6. The method of claim 5, wherein step five comprises the following steps: First, the At the time of iteration, the first The first vernonia of the group Position of vernonia ganensis only The fitness function of (2) is And determining each individual up to the first based on the fitness value Instead of locally optimal position, the whole population is determined until the first The generation is the global optimal position; Then, selecting elite individuals with better fitness in the group as knowledge sources through the acceptance function, wherein the position information of the elite individuals is used for updating a standard knowledge base in belief space, and the local optimal positions of vernonia affecting the lower bound of the standard function and the upper bound of the standard function are respectively And ; The specific update equation of the related parameters of the canonical knowledge is as follows: 。
7. A small snapshot direction-finding system of an unmanned aerial vehicle group on an offshore target, characterized in that the system is provided with a program module corresponding to the steps of the method according to any one of claims 1-6, and the steps in the small snapshot direction-finding method of the unmanned aerial vehicle group on the offshore target are executed in operation.
8. A computer readable storage medium, characterized in that the computer readable storage medium stores a computer program configured to implement the steps in the small snapshot direction finding method of a group of unmanned aerial vehicles on an offshore target as claimed in any one of claims 1 to 6 when invoked by a processor.

Description

Small snapshot direction-finding method and system for unmanned aerial vehicle group to offshore target Technical Field The invention relates to the technical field of array signal processing, in particular to a small snapshot direction-finding method and system for an unmanned aerial vehicle group to an offshore target. Background Direction finding, also called direction of arrival (Direction of Arrival, DOA) estimation, is an important component in the field of array signal processing. The traditional high-resolution direction finding method generally utilizes second order statistics to carry out modeling analysis, and has good estimation performance under Gaussian noise. The noise characteristics in the real environment are complex and changeable, and the traditional direction finding method is easy to have the problems of performance degradation and even failure. Most of the existing direction finding methods are built on a Uniform Linear Array (ULA) model, the degree of freedom of the existing direction finding methods is limited by the number of physical array elements, and effective direction finding is difficult to achieve when the number of information sources exceeds the number of array elements. The Minimum Redundant Array (MRA) is used as a special non-uniform linear array structure, and is characterized in that the kirchhoff difference set of the array has a completely continuous expansion characteristic, and a larger equivalent aperture and a higher degree of freedom can be obtained by optimizing the array element positions and using fewer physical array elements. However, the non-uniform characteristic of the method can introduce higher side lobes and grating lobes, so that the spatial spectrum estimation is blurred, and the complex mutual coupling effect among array elements can cause the non-ideal effect when the traditional algorithm is directly applied. Especially under the condition of small snapshot, the estimation error of the sample covariance matrix is further amplified, resulting in the deterioration of the direction finding performance. According to the search findings of the prior art documents, zhang Lijiang et al discuss the DOA estimation of the minimum redundancy linear array in the ' firepower and command control ' (2012,37 (10): 10-13) ', the minimum redundancy linear array direction finding performance adopting the MUSIC algorithm is discussed, the array aperture is effectively expanded by reducing the array redundancy, but the method is verified in an ideal Gaussian noise environment, the influence of the actual noise environment is not considered, the simulation snapshot number is higher, the performance of the algorithm under the small snapshot condition cannot be embodied, zhang Xiuqing et al propose a new method in the ' joint correction robust beam forming algorithm based on the small snapshot scene ' in the ' radio engineering (2024,54 (8): 1900-1907) ', the beam forming robustness under the small snapshot condition is improved by jointly executing the covariance matrix reconstruction and the guide vector optimization, the algorithm utilizes an uncertainty set to complete the solution of an interference guide vector, the covariance matrix reconstruction is realized by means of the power integral, and the method is still dependent on the second order of the covariance matrix reconstruction, and the special structure of the small redundancy array is not considered. In summary, although the existing direction finding methods all achieve a certain effect, the method cannot achieve high-precision direction finding of the unmanned aerial vehicle group on the sea target under the minimum redundant array architecture, the small snapshot condition and the multi-coherent source environment. Disclosure of Invention The invention aims to solve the technical problems that: the existing minimum redundant array has the problems of insufficient direction-finding precision and robustness under the condition of small snapshot and easiness in being influenced by model errors. The invention adopts the technical scheme for solving the technical problems: the invention provides a small snapshot direction-finding method of an unmanned aerial vehicle group on an offshore target, which comprises the following steps: The method comprises the steps of establishing a small snapshot receiving signal model of an unmanned aerial vehicle group minimum redundant array for an offshore target, constructing a fractional low-order covariance matrix by utilizing small snapshot data received by the minimum redundant array formed by a plurality of unmanned aerial vehicle groups, and establishing a maximum likelihood estimating target function; Step two: setting the types of the vernonia group as basic vernonia group reverse vernonia and vernonia cultura, initializing a vernonia group and setting related parameters; step three, constructing an adaptability function based on the established maximum likelihoo