CN112016628A

CN112016628A - Space debris impact damage interpretation method based on dynamic multi-target prediction

Info

Publication number: CN112016628A
Application number: CN202010920401.4A
Authority: CN
Inventors: 黄雪刚; 谭旭彤; 殷春; 石安华; 李晶; 罗锦阳; 兰胜威
Original assignee: Ultra High Speed Aerodynamics Institute China Aerodynamics Research and Development Center
Current assignee: Ultra High Speed Aerodynamics Institute China Aerodynamics Research and Development Center
Priority date: 2020-09-04
Filing date: 2020-09-04
Publication date: 2020-12-01
Anticipated expiration: 2040-09-04
Also published as: CN112016628B

Abstract

The invention discloses a space debris impact damage interpretation method based on dynamic multi-target prediction, which comprises the following steps: representing the acquired infrared thermal image sequence of the spacecraft impact damage test piece by using a three-dimensional matrix; selecting a pixel point corresponding to the transient thermal response with the minimum rate of rise from the three-dimensional matrix; determining a block size based on the minimum rate of rise transient thermal response; partitioning and determining the searching step length in the block; selecting transient thermal response in a long-step manner by blocks; classifying the selected transient thermal response by adopting an unsupervised clustering algorithm; realizing dynamic prediction and multi-objective optimization based on an improved CLIQUE clustering algorithm, and selecting a representative of each type of transient thermal response to form a matrix; judging the actual defect category number based on the spearman correlation coefficient, and obtaining a two-dimensional image; and performing feature extraction on the two-dimensional image by using the regional convolutional neural network R-CNN to obtain defect features of the spacecraft impact damage test piece. Therefore, the effective interpretation of the characteristics of the tiny damage defects caused by the space debris impact is realized.

Description

Space debris impact damage interpretation method based on dynamic multi-target prediction

Technical Field

The invention belongs to the technical field of space debris impact damage detection and evaluation, and particularly relates to a space debris impact damage interpretation method based on dynamic multi-target prediction.

Background

With the continuous development of human space activities, the increasing space debris poses a great threat to the safety of human space activities and in-orbit spacecrafts, and the space debris problem becomes a real problem which develops and is serious gradually in human space development and practice. Particularly, the tiny space fragments with huge number below centimeter level exist on the earth orbit, which have become potential killers threatening the on-orbit operation of the spacecraft and the astronauts because the tiny space fragments can not be effectively monitored, early-warned and actively avoided, and the space accidents caused by the collision of the tiny space fragments are rare. When various types of spacecrafts are impacted by space debris, how to acquire, analyze and evaluate the impact damage information is very important. Considering that the space debris impact events occur randomly in a dynamic environment, the long-term in-orbit spacecraft is usually subjected to irregular multiple impacts during service, and the impact damage quantity, damage degree, damage position, damage type and the like of the long-term in-orbit spacecraft are unpredictable. Therefore, the method can rapidly and accurately detect, identify and analyze the collision damage of the tiny space debris in an on-orbit manner, can facilitate astronauts and ground workers in space to timely make correct operation and take necessary measures, and provides important technical support for on-orbit evaluation and decision of space debris collision events of the spacecraft, thereby ensuring the safety of the astronauts and the spacecraft and the smooth completion of space missions.

The infrared thermal imaging device and the related technology thereof are widely applied to the field of aerospace and play an important role in the aspects of spacecraft damage detection and evaluation. The damage detection and evaluation technology based on the infrared thermal imaging principle is based on the infrared radiation characteristic, utilizes different structures or different physical thermal radiation characteristics of materials to detect the nonuniformity or the abnormality on the surface and the inside of the material, has the advantages of high speed, non-contact, no pollution, large single detection area, visualized result, wide applicable material types and the like, and is very suitable for carrying out in-situ in-service detection on complex damage caused by space debris impact. Based on the surface temperature field change data (namely an infrared thermal image sequence) of the tested object in the external thermal excitation environment collected by the infrared thermal imaging device, transient thermal response information of different damage areas of the tested object in space and time dimensions can be obtained, and further, the visual detection and evaluation of complex damage defects caused by fragment impact in a broken space can be realized by utilizing a corresponding feature extraction processing algorithm. Meanwhile, the long-term in-orbit spacecraft can be impacted by the tiny space debris for many times, so that the impact damage of the tiny space debris can be monitored in-orbit by analyzing and processing the regularly acquired infrared thermal image sequence data. It can be seen that, for the detection and evaluation of the impact damage of the tiny space debris and the on-track monitoring of the impact damage change process, how to automatically, rapidly and accurately extract and separate the damage characteristic information from the massive infrared thermal image sequence data is crucial. Practice shows that in the process of processing infrared thermal image sequence data, the efficiency and the precision of data analysis and processing can be improved by using a proper multi-objective optimization method, so that rapid and accurate identification and interpretation of space debris impact damage in a dynamic environment are realized.

The name of application in 2018, 11, month and 30 is ' a method for extracting defect characteristics of infrared thermal images based on dynamic multi-objective optimization ' (the publication number is 201811451744.X) and ' the method is based on uniformityIn the Chinese patent application of the evolved multi-target optimization infrared thermal image defect feature extraction method (publication number 201811451866.9), when the environment of a multi-target optimization problem changes, a multi-direction prediction strategy based on prediction is adopted to predict the ideal PS position after each environment change, in the prediction process, the algorithm stores PS capable of fully describing m-1 times and m times of time respectively in m-1 times and m times of external environments,

and

w of the shapes and the diversity of (A) represent a multidirectional prediction set of transient thermal responses constituting an external environment of m-1 and m times

And

in the multi-target environment of m +1 times, multi-directional prediction sets of m-1 times and m times are utilized

And

estimating new circumstances

And point-evolving the trajectory, and generating a new initial population solution near the predicted PS to accelerate the convergence of the multi-objective optimization algorithm under the new environment, so that the operating efficiency of the dynamic multi-objective optimization algorithm is improved. But in

In finding a multi-directional prediction set constituting an m-1 external environment

And in

In finding a multi-directional prediction set constituting an m-times external environment

When W multi-directional prediction set elements are used, it utilizes

Transient thermal response and multi-directional prediction set in (1)

The distance of the initial element in (a),

the transient thermal response and the multi-directional prediction set

The distance of the initial element in the multi-directional prediction set is used as a measurement scale for selecting the newly added multi-directional prediction set element, so that a large amount of repeated distance calculation is needed, the algorithm efficiency is low, a large amount of high-dimensional transient thermal response data exist in the aerospace field, an effective clustering effect is difficult to form on the high-dimensional data by utilizing distance clustering, and meanwhile, a large amount of time is consumed by frequent distance calculation, so that the algorithm speed is low, and the dynamic change multi-target environment is difficult to rapidly respond. On the basis, the invention adopts an improved CLIQUE clustering algorithm combining grids and density, the algorithm is based on the grids, and a high-dimensional data space R is obtainedⁿDividing the grid into a plurality of rectangular unit grids for processing, taking the number of points falling into each grid as a grid density criterion in combination with density, and then connecting the dense grids according to the density so as to perform clustering. The method integrates essences based on density and grid clustering algorithms, and the grids are taken as processing objects in each clustering, so that the clustering speed is greatly improved; the algorithm utilizes subspace clustering to divide the original transient thermal response characteristic space of the data into different characteristic subsets, examines the meaning of clustering division of each data cluster from different subspace angles, and simultaneouslyThe corresponding characteristic subspace is found for each data cluster in the clustering process, so that the subspace of a high-dimensional data space can be automatically identified, mixed types and high-dimensional spatial data in a large database are processed, and high efficiency is achieved; and meanwhile, the method is combined with a clustering method based on density, so that the frequent distance calculation of each point and all other points is avoided, and the method can be used for filtering a 'noise' data object.

Meanwhile, in the aspect of decomposition of the multi-objective optimization problem, a Chebyshev decomposition method is adopted to enable the leading edge of each transient thermal response category to approximate a solution set

The solution in (2) is directed by the weight vector to evolve in a certain direction towards the actual leading edge solution set PF. However, the Chebyshev decomposition method is difficult to obtain a uniform approximate frontal plane solution set when facing a multi-target optimization problem higher than two dimensions

With weight vector λ ═ λ₁，λ₂，λ₃)^TFor example, for the sake of the chebyshev aggregate form itself, the guideline pertains to the weight vector λ ═ (λ ═ λ)₁，λ₂，λ₃)^TThe solution of (a) is evolved in the direction vector of λ' ═ 1/λ₁，1/λ₂，1/λ₃)^TA straight line of (2). Since the evolution direction of the solution is not along the straight line where the weight vector itself is located, the obtained solution is not uniform even if the weight vector is uniform, resulting in obtaining a leading edge approximate solution set of each transient thermal response class

The density degrees of the medium solutions are different, the optimal solution can not be converged in the region where the solution set is sparse on the PF of the actual front edge solution set in the multi-target environment in the field of aerospace, the transient thermal response of the characteristic defect information cannot be accurately found, the conditions of defect detection failure and missed detection are caused, and the region where the solution is too dense actually existsThe diversity of the solution sets is limited, so that the diversity condition of the solution sets in the region can be described only by using a small amount of approximate leading edge solutions, if the same number of uniform weight vectors are still adopted to guide the evolution of the solutions, the transient thermal response representing the same type of defect information is easy to find repeatedly, not only can the waste of resources be caused, but also the overall time of the algorithm is increased because the iteration times required by the convergence of the regions with solution concentration and solution concentration sparsity are different, and the response to the dynamic multi-target environment is slowed down. The invention adopts a boundary crossing method based on penalty terms to decompose the multi-objective optimization problem, starts with the aggregation function, and improves the form of the aggregation function into a solution and a weight vector lambda (lambda is equal to lambda)₁，λ₂，λ₃)^TThe solution evolution direction is limited on the weight vector from the decomposition form, so that the solution of the multi-objective optimization algorithm evolves along the direction of the weight vector, and a uniformly distributed front edge approximate solution set can be obtained when the large number of multi-objective optimization problems higher than two dimensions in the aerospace field are faced

The method has the advantages that the solution of the self optimal solution set sparse area on the actual front edge PF of various transient thermal responses in the evolution process can be evolved to the position of the optimal solution, the detection accuracy is improved, the weight vector required for evolving to the actual front edge optimal solution set dense area is reduced, the resource waste is avoided, and the algorithm speed is improved. And a penalty factor is introduced to balance the convergence and diversity among population solutions and reduce the overall operation time of the algorithm, thereby dealing with the complex multi-objective optimization problem environment of the spacecraft.

In addition, in the original patent, small-size row and column block segmentation is carried out on the basis of the maximum point of the integral temperature aiming at the infrared thermal image data block, so that the number of the data blocks needing to be processed is large, the requirement that defect detection needs to be carried out on a target frequently and in real time in the aerospace process is met, the calculated amount of small-size blocks is large, the processing efficiency is low, and the small-size blocks are difficult to be processedThe defects are timely and rapidly discovered and detected after being generated, meanwhile, excessive data block segmentation can cause search omission and false removal of transient thermal response data representing the defect part, and the defect detection precision is influenced. The invention improves the block operation part aiming at the infrared thermal image data block, firstly finds out the part representing the background area in the three-dimensional data block based on the transient thermal response rising rate, divides the data block into the data blocks with larger size according to the size of the background area, and adopts different search step lengths to search data in different data blocks, thereby effectively avoiding the excessive redundancy removal phenomenon caused by small-size division blocks based on the maximum temperature value, and improving the data processing speed and the detection accuracy. According to the leading edge optimal solution set in the original patent

When various transient thermal response representatives are selected, a random selection mode is adopted, so that the detection precision and accuracy of the algorithm are not high. The invention proposes approximating a solution set from a leading edge

Selecting a representation of the transient thermal response of class i_i′The specific scheme of REP based on weighted membership avoids uncertainty caused by random selection. In the original patent, when the total number of defect types is set, the fixed defect type number is adopted, and then thermal data corresponding to each type of defect is searched from infrared data for detection, so that the error identification and the detection omission of the defect types are caused. The invention provides a defect type number judging method based on a spearman correlation coefficient, which avoids the phenomena of false detection and missing detection caused by fixed defect type number detection and improves the detection accuracy. The index for representing the dynamic multi-target environment change degree in the original patent adopts a simple arithmetic mean value, so that the index is too sensitive to extreme function value change. The invention improves the judgment index formula of the intensity degree of each environmental change in the dynamic multi-target environment, avoids the overlarge influence of the extreme function value on the judgment value after the environmental change, and improves the response speed to the dynamic environmental change.

Disclosure of Invention

An object of the present invention is to solve at least the above problems and/or disadvantages and to provide at least the advantages described hereinafter.

To achieve these objects and other advantages in accordance with the purpose of the invention, there is provided a space debris impact damage interpretation method based on dynamic multi-target prediction, comprising the steps of:

step one, an infrared thermal image sequence of a spacecraft impact damage test piece acquired by an infrared thermal imager is represented by a three-dimensional matrix S, wherein elements S (i, j, t) represent pixel values of an ith row and a jth column of a t-frame thermal image of the thermal image sequence;

step two, selecting a pixel point S (i) corresponding to the transient thermal response with the minimum rate of rise from all transient thermal responses of the three-dimensional matrix S_zz，j_zzT) in which i_zz、j_zzRespectively representing the row number of the row where the pixel point corresponding to the transient thermal response with the minimum rate of rise is located and the column number of the column where the pixel point is located;

step three, determining the size of the block based on the transient thermal response of the minimum rising rate;

fourthly, partitioning according to the size of the partitions and determining the searching step length in the blocks;

step five, selecting transient thermal response in a long-step manner by blocks;

step six, classifying the selected transient thermal response by adopting an unsupervised clustering algorithm;

seventhly, realizing dynamic prediction based on an improved CLIQUE clustering algorithm, performing multi-objective optimization by using a boundary crossing method based on penalty items, and selecting a representative of each type of transient thermal response to form a matrix Y;

judging the actual defect category number based on the Spireman correlation coefficient, and obtaining a two-dimensional image according to the actual defect category number matrix change;

and step nine, performing feature extraction on the two-dimensional image by using the regional convolutional neural network R-CNN to obtain defect features of the spacecraft impact damage test piece.

Preferably, the step three of determining the block size based on the minimum rate of rise transient thermal response specifically comprises:

setting a block row threshold K _ THV_rBlock column threshold value K _ THV_cSequentially calculating other temperature points S (i) of the column where the pixel point corresponding to the transient thermal response with the minimum rate of rise is positioned_N，j_zzT) the pixel point S (i) corresponding to the transient thermal response with the minimum rate of rise_zz，j_zzCorrelation of t)

Wherein i_NExpressing the Nth point closest to the pixel point corresponding to the transient thermal response with the minimum rate of rise; always find the first

Less than the block row threshold K _ THV_rCounting the number N of the pixel points;

sequentially calculating other temperature points S (i) of the row where the pixel point corresponding to the transient thermal response with the minimum rate of rise is positioned_zz，j_MT) the pixel point S (i) corresponding to the transient thermal response with the minimum rate of rise_zz，j_zzCorrelation of t)

Wherein j is_MExpressing the Mth point closest to the pixel point corresponding to the transient thermal response with the minimum rate of rise; looking up to the first

Less than blocking column threshold K _ THV_cAnd (4) counting the number M of the pixel points.

Preferably, the step four of partitioning according to the partition size and determining the intra-block search step size specifically includes:

sequentially decomposing the three-dimensional matrix into K sub three-dimensional matrix blocks with the size of NxM according to the number M, N of pixel points obtained based on the block row and column thresholds_kS(i_n，j_mT), where k denotes the kth sub-three-dimensional matrix block, i_n、j_mAnd t respectively represent the ith of the kth sub three-dimensional matrix block_nLine, j_mThe pixel values of the column and the T-th frame, N is 1,2, …, N, M is 1,2, …, M, T is 1,2, … T, and T is the total number of the three-dimensional matrix S frames;

at kth sub-three-dimensional data block_kS(i_n，j_mAnd t), searching around by taking the central point of the sub three-dimensional data block as the center of a circle, finding the central point in the sub three-dimensional data block, and recording the central point as the central point_kS(_ki_N/2，_kj_M/2，_kt) of the first and second groups, wherein,_ki_N/2，kj_M/2，kt respectively represents the number of rows, the number of columns and the number of frames of the maximum pixel point in the kth sub three-dimensional data block;

setting an intra block inner row threshold R _ THV of a kth sub three-dimensional data block_kSequentially calculating the temperature points of the frame and the line where the central point is located in the sub three-dimensional data block_kS(_ki_n′，kj_M/2，kt) and the center point in the sub three-dimensional data block_kS(_ki_N/2，kj_M/2，kCorrelation of t)

Wherein_ki_n′The nth' pixel point which is expressed in the kth sub three-dimensional data block and is closest to the central point in the sub three-dimensional data block is calculated and counted

Number of temperature points of (D), is recorded as_kRSS, as the intra block row step size of the kth sub-three-dimensional data block;

setting an intra-block column threshold C _ THV for a kth sub-three-dimensional data block_kSequentially calculating the temperature points of the frame and the column where the central point is located in the sub three-dimensional data block_kS(_ki_N/2，kj_m′，kt) and the center point in the sub three-dimensional data block_kS(_ki_N/2，kj_M/2，kCorrelation of t)

Wherein_kj_m′Is shown in the kth sub-three-dimensional data block from the center point in the sub-three-dimensional data block_kS(_ki_N/2，kj_M/2，kt) the nearest mth pixel point; calculate and count

Number of temperature points, is_kCSS as the intra-block column step size of the kth sub-three-dimensional data block.

Preferably, the step of selecting the transient thermal response in five blocks and in steps comprises the following specific steps:

step S51, partitioning the three-dimensional matrix S according to M, N pixel values counted in the step three to obtain K sub three-dimensional data blocks with the size of NxMxT, wherein the K sub three-dimensional data blocks are obtained_kS(i_n，j_mAnd t) represents the ith in the kth sub-three-dimensional data block_nLine, j_mTransient thermal response of the column pixels, wherein T is 1,2, …, and T is the total number of S frames of the original three-dimensional matrix;

step S52, for pixel points in each sub three-dimensional data block_kS(i_n，j_mT), setting a threshold DD, initializing a set number g to 1, and initializing a pixel point position i_n＝1，j _m1, and the maximum value in the block_kS(_ki_zz，kj_zz，kt_zz) Corresponding transient thermal response_kS(_ki_zz，kj_zz，T), T1, 2,. and T, stored in the set x (g); then calculating pixel points in the sub three-dimensional data block_kS(i_n，j_mLocated in i) in t)_nLine, j_mCorrelation Re between the transient thermal response of the column and the set X (g)_i，jAnd judging:

if Re_i,j<DD, g is g +1, and transient thermal response is carried out_kS(i_n,j_mT) as a new feature in the set X (g); otherwise, let i_n＝i_n+_kRSS, continue to calculate the next transient thermal response_kS(i_n,j_mT), T1, 2, T and TAnd the degree of correlation of X (g); if i_nIf > N, then let i_n＝i_n-N， j_m＝j_m+_kCSS, i.e. change to j_m+_kCSS column is calculated if j_mIf the number of the sub-three-dimensional matrixes is more than M, the transient thermal response of the kth sub-three-dimensional matrix is selected, and k is k + 1; n, M denotes the number of rows and columns of the kth sub-three-dimensional data block of size nxmxt, respectively.

Preferably, the method for classifying the selected transient thermal responses by using the unsupervised clustering algorithm in the sixth step is to divide all sets x (g) of all K data blocks selected in the fifth step, that is, the transient thermal responses into L classes by using a fuzzy C-means clustering FCM algorithm, so as to obtain the class to which each transient thermal response belongs, and specifically includes the following steps:

step S61, setting the number of clusters L, setting the number of initial iterations c to 0, and setting a threshold of termination iteration conditions;

step S62, using formula

Calculating a membership matrix U;

wherein i ═ 1,2, …, L, c ∈ L,_n′d_k′＝||x_k′-_i′V||，n′＝i′，j′，_n′d_k′representing the k 'th pixel point and the i' th cluster center_i′Euclidean distance of V, x_k′Representing the coordinates of the kth pixel point; τ is a constant;_i′u_k′the degree of the k 'th pixel point belonging to the i' th class is represented;

step S63, updating the clustering center_i′V

Wherein the content of the first and second substances,

expressing the thermal response value of the k' th pixel point;

step S64: if the iteration times reach the maximum value L or the absolute value of the difference between the clustering centers of the two times is smaller than the maximum value L, finishing the algorithm, outputting a membership matrix U and a clustering center V, and then entering the step S65; otherwise, let c be c +1, return to step S62;

step S65, defuzzifying all pixel points by utilizing membership maximization criterion to obtain the category of each pixel point, namely M_k′＝arg_i′max(_i′u_k′)。

Preferably, the step seven of selecting the representation of each type of transient thermal response based on dynamic multi-target to form the matrix Y specifically includes:

in step S71, when the i '(i' ═ 1., L) -th transient thermal response is represented in the (m + 1) -th external environment, a multi-objective function is defined:

wherein the content of the first and second substances,

a transient thermal response selected for the i' th class transient thermal response in the m +1 th external environment

Is expressed as:

a transient thermal response selected for the i' th class transient thermal response

The calculated Euclidean distance between L-1 classes

The components are renumbered and the components are,

expressed as:

for transient thermal response

The pixel value at time t i.e. the temperature value,

the pixel value of the ith' type transient thermal response cluster center at the t-th moment, namely the temperature value,

the pixel value of the j' th class transient thermal response clustering center at the t th moment is a temperature value;

s72, the m-1 th and m-th environments respectively obtain multiple-target function approximate leading edge solution sets

And

the corresponding population transient thermal response solution sets are respectively

And

the number of which is respectively

And

after the environment changes, according to the history information of the m-1 and m-th environments, the transient thermal response of the initialized population of the approximate leading edge solution set under the m +1 environment is calculated in a pre-measuring way, and the steps are as follows:

step 721,

Is from

Randomly selecting N in the solution set_ETransient thermal response

A constructed set of transient thermal responses, N' ═ 1,2_ECalculating

And (3) concentrating the number W representing the transient thermal response, and obtaining a multidirectional prediction set in the (m + 1) th environment:

wherein, W₁And W₂Respectively a lower limit value and an upper limit value of W, and W₁＝L+1，W₂＝3L，

Is an evaluation value of the degree of environmental change at the m-th time, obtained by the following formula:

wherein the content of the first and second substances,

is from

Randomly selecting N in the solution set_ETransient thermal response

A constructed set of transient thermal responses, N' ═ 1,2_E；

Step S722, selecting W +1 representative transient thermal responses, the specific method includes:

step A, representing a PS multi-direction prediction set formed by transient thermal response, which consists of two parts:

firstly, the

The center of the transient thermal response is solved and recorded

Wherein the content of the first and second substances,

as a solution set

The nth transient thermal response;

secondly, W generations which are obtained based on a fully adaptive spectral clustering algorithm and can fully describe the shape and the diversity of the current PSSet of transient thermal responses of meter

Step B, using CLIQUE clustering algorithm to collect solution

Clustering into cluster sets of transient thermal responses in

The method comprises the following specific steps:

step B1, according to the preset division interval parameter, dividing the data into a plurality of sections

Each dimension attribute of the whole T-dimension data space of the transient thermal response in (1) is divided into rectangular area grids which are equally spaced and do not overlap with each other, wherein d represents the dimension of the transient thermal response data; storing the division condition information of each Grid unit through a Hash table Hash _ Grid, taking the number of the Grid unit as the key of the Hash table, storing a self-defined Grid Class data type Class _ Grid as the value of the Hash table, wherein one Grid Class _ Grid comprises 4 variables of count, classID, isChecked and points count, and recording the number of transient thermal response data falling into the Grid unit; the classID is the serial number of the class cluster to which the grid unit belongs, and the default is 0; the isChecked represents the traversed state of the grid unit, the value is 0 or 1, 1 represents that the inspection is finished, and the default value is 0; points is a stringbuffer object for storing coordinate information in the temperature subspace of all transient thermal response data points in the grid;

b2, setting a density threshold value theta, traversing all Grid units in Hash _ Grid of the Hash table, judging whether a variable count in the Grid Class _ Grid of each Grid meets the condition that the count is more than or equal to the theta, if so, marking the Grid as a dense Grid unit and adding the dense Grid unit into a candidate dense Grid unit set Bin; if not, marking the grid as a sparse grid unit;

b3, traversing all the sparse grid cells, judging whether the sparse grid cells are adjacent to the dense grid cells, and if the sparse grid cells are adjacent to the dense grid cells, turning to the step B4 to execute a boundary correction method; if not, go to step B5 to execute the sliding grid method;

step B4, executing the boundary correction method to the sparse grid unit adjacent to the dense grid unit;

recording the sparse grid unit as u and the adjacent dense grid unit as u', further dividing each dimension of the sparse grid unit u into two equal parts to obtain 2^TSub-grid cells, note

Statistics fall into each sub-unit_iu_sub(i＝1，2，…，2^T) And determining for each subunit the number of transient thermal response data points in:

wherein dens (_iu_sub) Indicating the density of the ith sub-cell grid, i.e. falling into a sub-cell_iu_subThe number of transient thermal responses; theta is a density threshold; merging all the sub-unit grids meeting the conditions in the sparse Grid u into a dense Grid u', and updating Hash _ Grid information of a Hash table;

performing boundary correction on each sparse grid adjacent to the dense grid, and turning to the step B6 after the boundary correction is finished;

step B5, performing a sliding grid algorithm on the sparse grid cells that do not have contiguous dense grid cells; by 2^TPerforming sliding grid operation on a group of sparse grid units; a set of these sparse grid cells without a neighboring dense grid is recorded as

Wherein u is_i，i＝1，2，…，2^TRespectively representing sparse grid cells, firstly calculating the center point coordinate C of each sparse grid_i，i＝1，2，…2^TWherein each center point coordinate is

D-dimensional coordinates representing a center point of the ith sparse grid, wherein each one-dimensional coordinate is:

wherein

And

respectively representing the upper limit and the lower limit of a coordinate range on the d-dimension of the ith sparse grid;

computing a common vertex C of a sparse grid set_M，

D-dimensional coordinates representing common vertices of the sparse grid set, wherein each dimension coordinate is:

wherein

And

respectively represent 2 nd^TThe upper limit of the coordinate range of the d-th dimension of each grid and the lower limit of the coordinate range of the d-th dimension of the 1 st grid;

computing vector coordinates for each sparse grid

Expressing the vector coordinates of the ith grid, and calculating the sliding direction of the sliding gridVector quantity

Wherein, count (u)_i) And count (u)_j) Respectively representing the number of transient thermal response data in the ith sparse grid and the jth sparse grid;

then common vertex C of sparse grid set_MDividing a mesh u 'with the same size as the original mesh into a central point again, and carrying out vector transformation on the mesh u' along the sliding direction

Is slid in the direction of (1) to obtain a slid mesh denoted by u'_SSliding distance of

Counting grid u 'after falling into sliding'_SThe number of transient thermal response data in (2) is judged whether:

dens(u′_S)≥θ

wherein, dens (u'_S) Denotes mesh u 'after sliding'_SIs in mesh u 'after sliding'_SIf the condition is satisfied, the new grid u'_SMarking as dense grids, adding the dense grids into a candidate dense Grid set Bin, and adding the information into a Hash _ Grid of a Hash table;

inspecting all other sparse grid unit sets, and turning to step B6 after finishing the inspection;

step B6, randomly selecting a grid unit in a candidate dense unit grid set Bin, performing horizontal and vertical expansion by taking the grid as the center according to the greedy algorithm retrieval principle, and ensuring that all adjacent and connected dense grid units in the horizontal and vertical directions of the grid are covered to form a maximum connected rectangular area andthe label is C₁Mixing C with₁Setting the classID in the Hash table Hash _ Grid of all grids to be 1 indicates that C is₁All grid cells are clustered into a first class, and C is₁The isshecked variable of all grids in (1); then randomly selecting a dense grid with the isChecked position of 0 in the candidate dense cell grid set Bin, and continuously performing horizontal and vertical expansion to form a maximum connected rectangular area C₂While C is₂Cannot contain a dense grid with the isshecked variable bit already at 1, C₂Class IDs of all grids are set to 2, which means C₂All grid cells are clustered into a second class, and C is then added₂The position of an isChecked variable of all grids is also 1; repeating all the operations until all the grid cells in the candidate dense cell grid set Bin are traversed, namely the isChecked variable bits of all the grid cells in Bin are all 1;

step B7, storing the clustering information result to the set

Step C, calculating the clustering center of each category in the clustering result:

wherein

Is as follows

Individual clustering result

The k-th of (a) represents a transient thermal response,

the total number of transient thermal responses contained in the h clustering result is obtained;

counting all transient thermal response points still in the sparse grid after the clustering step B so as to

Representing the ith transient thermal response always located in the sparse grid, calculating the distance between the ith transient thermal response and the center of each cluster which is clustered at present:

wherein the content of the first and second substances,

for the cluster center of the h-th cluster, then find the cluster center closest to each transient thermal response point still in the sparse grid

Then the transient thermal response is attributed to the category to obtain an updated clustering result

Based on updated clustering results

Selecting from each class one representative transient thermal response that adequately describes the current PS shape and diversity

Wherein the content of the first and second substances,

is as follows

Individual clustering result

Are clustered in the center, so as to obtain

A representative transient thermal response that adequately describes the current PS shape and diversity; judgment of

And the size of W, if

Then a new representative transient thermal response is required

Logging in sets as representative of transient thermal responses

In (1),

obtained by the following formula

Wherein the content of the first and second substances,

is as follows

Individual clustering result

I.e. finding all clustered results

Medium transient thermal response

Transient thermal response with maximum distance from all cluster centers; if it is

Then is at

Randomly selecting W transient thermal responses to form a multi-directional prediction set;

step S723, PS multi-directional prediction set according to m-1 th environment and m-th environment

And

wherein the content of the first and second substances,

obtained by the method of step S721, step S722, W' is

Collecting a number representing the transient thermal response;

calculating a predicted direction

Wherein the content of the first and second substances,

is PS multidirectional prediction set

Neutralization of

The nearest transient thermal response has the sequence number h';

in step S724, when the iteration number g' is 0, the number of transient thermal responses of the initialized population of the approximate leading edge solution set in the m +1 th environment is N_pWherein, in the step (A),

the transient thermal response of the initial population is randomly generated within a value range,

the transient thermal response of the initial population is obtained by predicting according to the following formula:

wherein h is_nFor transient thermal response

The cluster result

The serial number of (a) is included,

is a obedience mean value of 0 and a variance of

Normally distributed random number, variance of

The calculation formula of (2) is as follows:

according to the method, the transient thermal response of the initialized population of the approximate leading edge solution set in the (m + 1) th environment is obtained according to the historical information in the previous environment, so that a guide direction is provided for population evolution, and the multi-target optimization algorithm is helped to make quick response to new changes;

step S73, initializing relevant parameters

The number of initialization iterations g' is 0, and a set of evenly distributed weight vectors

Wherein:

initializing reference points

Is a function of

A corresponding reference point;

maximum number of iterations g'_max；

The evolution speed for initializing each population transient thermal response is

Global optimal and local optimal satisfaction of population transient thermal response

Step S74, use

Constructing dynamic objective function fitness value of transient thermal response of each population under boundary crossing method based on penalty term

Wherein

Is a pre-set parameter of the process,

and

the setting of (a) is as follows:

step S75, where N is 1_P: updating speed according to particle swarm algorithm

And population transient thermal response

Comparison according to a Multi-objective optimization Algorithm

Update global optimum

Local optimization

And a reference point

From

Middle reserve branch and match

Removing all quilt

Dominant solution vector if

None of the vectors in (1) dominates

Will be provided with

Adding into

N is N +1, N is less than or equal to N_PThen g '═ g' + 1;

step S76, evolution termination judgment: if g 'is less than or equal to g'_maxThen step S75 is repeated if g '> g'_maxObtaining the final leading edge approximate solution set of the i' th class temperature transient thermal response

Step S77, approximate solution set from leading edge based on weighting membership degree scheme

Selecting a representation of class i' transient thermal response_i′REP；

Calculating a leading edge approximate solution set according to the following formula

The affiliation of the kth solution to the ith objective function:

wherein, F_lFor the value of the l-th objective function,

maximum and minimum values of the corresponding objective function, respectively;

setting a weight λ for an objective function_l(L ═ 1,2, …, L), calculating membership weighted value of leading edge approximate solution set, taking leading edge solution corresponding to maximum value as representative of i' th class transient thermal response_i′REP, formula as follows:

wherein the content of the first and second substances,

approximate solution set for leading edge

The number of solution sets contained, L is the number of objective functions,

is a membership function value;

all transient thermal response representatives of class L are placed in columns forming a matrix Y of T × L, one column being the pixel values at time T, i.e. the temperature values.

Preferably, the step eight of distinguishing the actual defect type number based on the spearman correlation coefficient, and obtaining the two-dimensional image f (x, y) according to the change of the actual defect type number matrix includes:

to be provided with_i′REP and_j′REP, (i', j ═ 1,2, …, L) represents any two transient thermal response representatives, which will_i′REP and_j′temperature element value of each time corresponding to REP_i′REP_t，(t＝12, …, T) and_j′REP_t(T ═ 1,2, …, T) to their descending ranking among the temperature values at the respective transient thermal response vector population instants, the element with the highest temperature value being converted to 1, the element with the lowest temperature value being converted to T, the remaining temperature value elements being converted in size order to the ranking and denoted Ra (r: (r) ((r)), (r) (r))_i′REP_t) And Ra (_j′REP_t) According to the formula:

calculating the difference Da of element values at corresponding moments between the two transient thermal response representatives, and finally according to the formula:

calculating two transient thermal response representations_i′REP and_j′spearman correlation coefficient between REP;

setting a correlation threshold value theta, comparing the Spireman correlation coefficients between every two, and keeping a transient thermal response representative with minimum correlation if Rs (R) ((R))_i′REP_t，j′REP_t) If the value is less than theta, keeping the i-th and j-th class transient thermal response representatives, otherwise, removing one class of transient thermal response representatives to obtain an L' -class transient thermal response representative; all the transient thermal response representatives of the L ' class are arranged in columns to form a matrix Y ' of T multiplied by L ', and one column is the pixel value at T moments, namely the temperature value;

starting each frame in the three-dimensional matrix S from a first column, connecting a next column to the tail of a previous column to form a new column, obtaining T-column pixel values corresponding to the T frame, sequentially placing the T-column pixel values according to time sequence to form an I multiplied by J row and T-column two-dimensional image matrix O, and performing linear transformation on the two-dimensional matrix O by using a matrix Y, namely:

a two-dimensional image matrix R is obtained, wherein,

is an L 'x T matrix, is a pseudo-inverse of the matrix Y', O^TA transpose matrix of the two-dimensional image matrix O, wherein an obtained two-dimensional image matrix R is L' rows and I multiplied by J columns;

and sequentially intercepting each row of the two-dimensional image matrix R according to J columns, and sequentially placing the intercepted J columns according to the rows to form an I multiplied by J two-dimensional image, so that L 'rows obtain L' I multiplied by J two-dimensional images, wherein the images all contain defect areas, and in order to facilitate defect contour extraction, a two-dimensional image with the largest pixel value difference between the defect area and the non-defect area, namely a two-dimensional image with the largest temperature value difference is selected and recorded as f (x, y).

Preferably, in the ninth step, the two-dimensional image f (x, y) obtained in the eighth step is subjected to defect detection and defect region segmentation by using a regional convolutional neural network R-CNN, so as to realize location identification and feature extraction of the defect part.

The invention at least comprises the following beneficial effects:

1. the method adopts a multi-objective optimization method to realize the comprehensive consideration of the difference and the similarity, accurately describes the defect outline, makes up for some defects of the traditional method on dimension reduction processing, and has more representativeness than the defect characteristic extracted by an algorithm only based on the difference;

2. the method adopts a multi-direction prediction strategy, combines a CLIQUE clustering algorithm to quickly select and introduce a plurality of shapes which represent transient thermal response and properly describe PS (pareto set), and records the distribution condition of the PS in each environment so as to predict the new position of the PS. After the environment is changed, the new position of the PS is predicted by using the representative transient thermal responses of the previous two environments, and a plurality of new initial population transient thermal responses are generated at the new position, so that the response to the environment change is accelerated. Based on the CLIQUE clustering algorithm, on the premise of ensuring the speed of the algorithm, the uncertainty caused by manual input of a threshold value is avoided, the error caused by division of a grid with a fixed length is reduced, the accuracy of the clustering algorithm is improved, and the rapidity and the effectiveness of the PS new position prediction are realized.

3. The invention adopts a boundary crossing method based on punishment items to carry out multi-target problem decomposition. When the multi-target optimization problem with the number of targets exceeding two dimensions is processed, the optimal solution distribution obtained by the method is more uniform compared with that obtained by a Chebyshev method, and when the high-dimensional multi-target optimization problem is processed, the boundary crossing method based on the penalty term is obviously superior to the Chebyshev method and is more suitable for the requirements of the space debris complex impact damage detection and evaluation aspect of the spacecraft. Meanwhile, due to the introduction of the penalty term, the balance between the convergence and the diversity of the optimal solution obtained by the evolutionary algorithm can be freely selected so as to meet different requirements of a dynamic multi-objective optimization problem in a dynamic environment.

Additional advantages, objects, and features of the invention will be set forth in part in the description which follows and in part will become apparent to those having ordinary skill in the art upon examination of the following or may be learned from practice of the invention.

Description of the drawings:

FIG. 1 is a flow chart of an embodiment of the method for interpreting space debris impact damage based on dynamic multi-objective prediction according to the present invention;

FIG. 2 is a flow chart of a modified infrared thermal image data large-size block variable step size transient thermal response search;

FIG. 3 is a flow chart of an improved CLIQUE clustering algorithm incorporating meshes and densities;

FIG. 4 is a graph of results of classifying selected transient thermal responses using fuzzy C-means clustering;

FIG. 5 is a scatter plot of the leading edge approximate solution of defect 1 based on the boundary crossing method of penalty terms and a transient thermal response representation of defect 1 selected based on a weighted membership scheme;

FIG. 6 is a scatter plot of the leading edge approximate solution for defect 2 based on the boundary crossing method of penalty terms and a transient thermal response representation for defect 2 selected based on a weighted membership scheme;

FIG. 7 is a plot of the leading edge approximate solution scatter of the background region based on the boundary crossing method of penalty terms and a representation of the transient thermal response of the background region selected based on a weighted membership scheme;

FIG. 8 is a graph of transient thermal response at defect 1 temperature point;

FIG. 9 is a graph of transient thermal response at defect 2 temperature points;

FIG. 10 is a graph of transient thermal response for background zone temperature points;

FIG. 11 is a graph of transient thermal response for the corresponding defect 1 temperature point selected based on the present invention;

FIG. 12 is a graph of transient thermal response for corresponding defect 2 temperature points selected based on the present invention;

FIG. 13 is a graph of transient thermal response for corresponding background area temperature points selected in accordance with the present invention;

FIG. 14 is a defect 1 feature map extracted based on the present invention;

fig. 15 is a defect 2 feature map extracted based on the present invention.

The specific implementation mode is as follows:

the present invention is further described in detail below with reference to the attached drawings so that those skilled in the art can implement the invention by referring to the description.

It will be understood that terms such as "having," "including," and "comprising," as used herein, do not preclude the presence or addition of one or more other elements or groups thereof.

As shown in fig. 1-3: the invention discloses a spacecraft infrared thermal image extraction method based on dynamic multi-objective optimization, which comprises the following steps of:

The invention relates to an infrared thermal image defect characteristic extraction method based on dynamic multi-target optimization, which selects transient thermal response of pixel points by changing step length of a thermal image sequence, classifies by adopting FCM (fuzzy C-means) to obtain the category of the transient thermal response of each pixel point, considers the pixel value (temperature value) similarity of each category pixel point and the same category pixel point, simultaneously considers the difference between the pixel point (temperature point) and different category pixel points (temperature points) to construct a corresponding multi-target function, simultaneously, after each environment changes, adopts a CLIQUE (closed-loop fuzzy algorithm) combining grid-based and density-based improvement through a prediction mechanism, improves the operation speed of the clustering algorithm, can more quickly obtain population PS (packet switched) information, improves the speed and efficiency of population prediction in a new environment, and can popularize and apply higher dimensional data, meanwhile, self-adaptive grid division is carried out on each dimension data by utilizing the relative entropy, and the uncertainty of the manual input parameters of the user is eliminated by providing a calculation formula of the density threshold value, so that the accuracy and the effectiveness of the clustering algorithm are ensured, and more accurate and faster population tracking can be further realized in a dynamic environment. The method provides a guiding direction for population evolution, helps a multi-objective optimization algorithm to make quick response to new changes, enables optimal solution distribution obtained by the evolution algorithm to be more uniform through the multi-objective optimization algorithm of a boundary crossing method based on punishment items, and simultaneously flexibly determines diversity and convergence balance among the optimal solutions according to different multi-objective optimization problem environments more conveniently, so that dimension reduction results of thermal image sequences are obtained more accurately, and finally, a pulse coupling neural network is used for feature extraction, so that defect features of the infrared thermal images are extracted. Through the steps, the accurate selection of the corresponding (temperature point) representing the transient heat is realized, the accuracy of defect feature extraction is ensured, and the calculation consumption for obtaining each category of information representing the corresponding transient heat in a dynamic environment is reduced.

In the above technical solution, the step three of determining the size of the block based on the minimum rate of rise transient thermal response specifically includes:

In the above technical solution, the four specific steps of partitioning according to the size of the partition and determining the search step size in the block include:

at kth sub-three-dimensional data block_kS(i_n，j_mAnd t), searching around by taking the central point of the sub three-dimensional data block as the center of a circle, finding the central point in the sub three-dimensional data block, and recording the central point as the central point_kS(_ki_N/2，kj_M/2，kt) of the first and second groups, wherein,_ki_N/2，kj_M/2，kt respectively represents the number of rows, the number of columns and the number of frames of the maximum pixel point in the kth sub three-dimensional data block;

Wherein_ki_n′Expressed in the k-th sub-three-dimensional data block, sub-three-dimensionallyThe n' th pixel point with the nearest central point in the data block is calculated and counted

In the above technical solution, the step of selecting the transient thermal response in five blocks and in steps includes:

step S52, for pixel points in each sub three-dimensional data block_kS(i_n，j_mT), setting a threshold DD, initializing a set number g to 1, and initializing a pixel point position i_n＝1，j _m1, and the maximum value in the block _kS(_ki_zz，kj_zz，kt_zz) Corresponding transient thermal response_kS(_ki_zz，kj_zz，T), T1, 2,. and T, stored in the set x (g); then calculating pixel points in the sub three-dimensional data block_kS(i_n，j_mLocated in i) in t)_nLine, j_mCorrelation Re between the transient thermal response of the column and the set X (g)_i，jAnd judging:

if Re_i,j<DD, g is g +1, and transient thermal response is carried out_kS(i_n,j_mT) as a new feature in the set X (g); otherwise, let i_n＝i_n+_kRSS, continue to calculate the next transient thermal response_kS(i_n,j_mT), T ═ 1,2,. and the degree of correlation of T with the set x (g); if i_nIf > N, then let i_n＝i_n-N， j_m＝j_m+_kCSS, i.e. change to j_m+_kCSS column is calculated if j_mIf the number of the sub-three-dimensional matrixes is more than M, the transient thermal response of the kth sub-three-dimensional matrix is selected, and k is k + 1; n, M denotes the number of rows and columns of the kth sub-three-dimensional data block of size nxmxt, respectively.

In the above technical solution, the method for classifying the selected transient thermal responses by using the unsupervised clustering algorithm in the sixth step is to divide all sets x (g) of all K data blocks selected in the fifth step, that is, the transient thermal responses into L classes by using the fuzzy C-means clustering FCM algorithm, to obtain the class to which each transient thermal response belongs, and specifically includes the following steps:

step S62, using formula

Calculating a membership matrix U;

step S63, updating the clustering center_i′V

Wherein the content of the first and second substances,

expressing the thermal response value of the k' th pixel point;

In the above technical solution, the step seven of selecting a representative composition matrix Y of each type of transient thermal response based on dynamic multi-objective specifically includes:

wherein the content of the first and second substances,

Is expressed as:

The calculated Euclidean distance between L-1 classes

The components are renumbered and the components are,

expressed as:

for transient thermal response

The pixel value at time t i.e. the temperature value,

clustering the pixel value at the t-th time for the j' -th transient thermal responseI.e. a temperature value;

And

And

the number of which is respectively

And

step 721,

Is from

Randomly selecting N in the solution set_ETransient thermal response

A constructed set of transient thermal responses, N' ═ 1,2_ECalculating

wherein the content of the first and second substances,

is from

Randomly selecting N in the solution set_ETransient thermal response

A constructed set of transient thermal responses, N' ═ 1,2_E；

firstly, the

The center of the transient thermal response is solved and recorded

Wherein the content of the first and second substances,

as a solution set

The nth transient thermal response;

secondly, W representative transient thermal response sets capable of fully describing the shape and diversity of the current PS and obtained based on a fully adaptive spectral clustering algorithm

Step B, using CLIQUE clustering algorithm to collect solution

Clustering into cluster sets of transient thermal responses in

The method comprises the following specific steps:

Each dimension attribute of the whole T-dimension data space of the transient thermal response in (1) is divided into rectangular area grids which are equally spaced and do not overlap with each other, wherein d represents the dimension of the transient thermal response data; storing the division condition information of each Grid unit through a Hash table Hash _ Grid, taking the number of the Grid unit as the key of the Hash table, storing a self-defined Grid Class data type Class _ Grid as the value of the Hash table, wherein one Grid Class _ Grid comprises 4 variables of count, classID, isChecked and points count, and recordingThe number of transient thermal response data falling within the grid cell; the classID is the serial number of the class cluster to which the grid unit belongs, and the default is 0; the isChecked represents the traversed state of the grid unit, the value is 0 or 1, 1 represents that the inspection is finished, and the default value is 0; points is a stringbuffer object for storing coordinate information in the temperature subspace of all transient thermal response data points in the grid;

wherein dens (_iu_sub) Indicating the density of the ith sub-cell grid, i.e. falling into a sub-cell_iu_subThe number of transient thermal responses; theta is a density threshold; thinning out theMerging all the sub-unit grids meeting the conditions in the Grid u into a dense Grid u', and updating Hash _ Grid information of the Hash table;

wherein

And

computing a common vertex C of a sparse grid set_M，

wherein

And

computing vector coordinates for each sparse grid

Expressing the vector coordinates of the ith grid, and calculating the sliding direction vector of the sliding grid

dens(u′_S)≥θ

step B6, randomly selecting a grid unit in a candidate dense unit grid set Bin, performing horizontal and vertical expansion by taking the grid as the center according to the greedy algorithm retrieval principle, ensuring that all adjacent and connected dense grid units in the horizontal and vertical directions of the grid are covered, forming a maximum connected rectangular area and marking as C₁Mixing C with₁Setting the classID in the Hash table Hash _ Grid of all grids to be 1 indicates that C is₁All grid cells are clustered into a first class, and C is₁The isshecked variable of all grids in (1); then randomly selecting a dense grid with the isChecked position of 0 in the candidate dense cell grid set Bin, and continuously performing horizontal and vertical expansion to form a maximum connected rectangular area C₂While C is₂Cannot contain a dense grid with the isshecked variable bit already at 1, C₂Class IDs of all grids are set to 2, which means C₂All grid cells are clustered into a second class, and C is then added₂The position of an isChecked variable of all grids is also 1; repeating all the operations until all the grid cells in the candidate dense cell grid set Bin are traversed, namely the isChecked variable bits of all the grid cells in Bin are all 1;

step B7, storing the clustering information result to the set

wherein

Is as follows

Individual clustering result

The k-th of (a) represents a transient thermal response,

wherein the content of the first and second substances,

Based on updated clustering results

Wherein the content of the first and second substances,

is as follows

Individual clustering result

Are clustered in the center, so as to obtain

And the size of W, if

Then a new representative transient thermal response is required

Logging in sets as representative of transient thermal responses

In (1),

obtained by the following formula

Wherein the content of the first and second substances,

is as follows

Individual clustering result

I.e. finding all clustered results

Medium transient thermal response

Then is at

And

wherein the content of the first and second substances,

obtained by the method of step S721, step S722, W' is

Collecting a number representing the transient thermal response;

calculating a predicted direction

Wherein the content of the first and second substances,

is PS multidirectional prediction set

Neutralization of

The nearest transient thermal response has the sequence number h';

wherein h is_nFor transient thermal response

The cluster result

The serial number of (a) is included,

is a obedience mean value of 0 and a variance of

Normally distributed random number, variance of

The calculation formula of (2) is as follows:

step S73, initializing relevant parameters

Wherein:

initializing reference points

Is a function of

A corresponding reference point;

maximum number of iterations g'_max；

Initializing each populationThe evolution speed of the transient thermal response is

Step S74, use

Wherein

Is a pre-set parameter of the process,

and

the setting of (a) is as follows:

step S75, where N is 1_P: updating speed according to particle swarm algorithm

And population transient thermal response

Comparison according to a Multi-objective optimization Algorithm

Update global optimum

Local optimization

And a reference point

From

Middle reserve branch and match

Removing all quilt

Dominant solution vector if

None of the vectors in (1) dominates

Will be provided with

Adding into

N is N +1, N is less than or equal to N_PThen g '═ g' + 1;

step S76, evolution termination judgment: if g 'is less than or equal to g'_maxThen step S75 is repeated if g '> g'_maxThen the i' th class temperature is obtainedFinal leading edge approximate solution set of transient thermal response

Selecting a representation of class i' transient thermal response_i′REP；

The affiliation of the kth solution to the ith objective function:

wherein, F_lFor the value of the l-th objective function,

wherein the content of the first and second substances,

approximate solution set for leading edge

The number of solution sets contained, L is the number of objective functions,

is a membership function value;

In the above technical solution, the step eight of distinguishing the actual defect type number based on the spearman correlation coefficient, and obtaining the two-dimensional image f (x, y) according to the change of the actual defect type number matrix specifically includes:

to be provided with_i′REP and_j′REP, (i', j ═ 1,2, …, L) represents any two transient thermal response representatives, which will_i′REP and_j′temperature element value of each time corresponding to REP_i′REP_t(T ═ 1,2, …, T) and_j′REP_t(T ═ 1,2, …, T) to their descending ranking among the temperature values at the respective transient thermal response vector population instants, the element with the highest temperature value being converted to 1, the element with the lowest temperature value being converted to T, the remaining temperature value elements being converted in size order to the ranking and denoted Ra (r: (r) ((r)), (r) (r))_i′REP_t) And Ra (_j′REP_t) According to the formula:

a two-dimensional image matrix R is obtained, wherein,

In the above technical solution, the step nine uses the regional convolutional neural network R-CNN to perform defect detection and defect region segmentation on the two-dimensional image f (x, y) obtained in the step eight, thereby implementing location identification and feature extraction of the defect portion.

Example (b):

in the present example, there are two defects on the test piece, i.e., a surface tearing hollow defect 1 and a surface layer breakdown defect 2 by impact.

The flow chart of the improved large-size block variable-step transient thermal response search of the infrared thermal image data is shown in fig. 2.

A flow chart of the modified CLIQUE clustering algorithm incorporating the mesh and density is shown in fig. 3.

In this example, the result of classifying the selected transient thermal response by using the fuzzy C-means clustering is shown in fig. 4.

Collecting approximate leading edge solution set of defect 1 temperature point, defect 2 temperature point and material temperature point obtained by boundary crossing method based on penalty term₁AP、₂AP and₃and (7) AP. The representative transient thermal response of the defect 1 temperature point, the defect 2 temperature point and the material temperature point which are simultaneously selected by adopting the weighting membership scheme is_AFV₃₂、_BFV₃₄And_CFV₂₈as shown in fig. 5, 6 and 7.

Three known temperature points, namely transient thermal response curves of a material temperature point, a defect 1 temperature point and a defect 2 temperature point are directly extracted from a thermal image sequence of the test piece and are respectively marked as TTR_Background、TTR_De1And TTR_De2As shown in fig. 8, 9 and 10.

By using the method for dynamically selecting the transient thermal response representatives through multi-objective optimization, three transient thermal response representatives are obtained:_AFV₃₂、_BFV₃₄and_CFV₂₈the curves are shown in fig. 11, 12, and 13, and correspond to the defect 1 temperature point, the defect 2 temperature point, and the material temperature point, respectively.

From the thermal response curves, it can be seen that: the temperature peak value of the defect 1 temperature point is higher than that of the background area, the peak value of the defect 2 temperature point is higher than that of the defect 1 temperature point and is greatly higher than that of the background area, and the peak value of the background area temperature point is lowest. Compared with the three characteristics, the temperature point of the defect 2 absorbs most heat, and the temperature point of the defect 1 absorbs more heat.

The correlation between the transient thermal response curve of the present invention and the corresponding transient thermal response curve extracted directly from the thermographic sequence is shown in table 1.

TABLE 1

	Temperature point of itself	Temperature point of defect 1	Temperature point of defect 2
				The invention	0.995	0.997	0.990

From table 1, it can be seen that the transient thermal response curves selected by the method of the present invention have better correlation.

In the present embodiment, the defect features extracted are as shown in fig. 12.

The number of apparatuses and the scale of the process described herein are intended to simplify the description of the present invention. Applications, modifications and variations of the present invention will be apparent to those skilled in the art.

While embodiments of the invention have been described above, it is not limited to the applications set forth in the description and the embodiments, which are fully applicable to various fields of endeavor with which the invention may be practiced, and further modifications may readily be effected therein by those skilled in the art, without departing from the general concept as defined by the claims and their equivalents, which are not limited to the details given herein and the examples shown and described herein.

Claims

1. A space debris impact damage interpretation method based on dynamic multi-target prediction is characterized by comprising the following steps:

step two, selecting a pixel point S (i) corresponding to the transient thermal response with the minimum rate of rise from all transient thermal responses of the three-dimensional matrix S_zz，j_zzT) in which i_zz、j_zzRespectively representing the row number of the pixel point corresponding to the transient thermal response with the minimum rate of rise in the row and the column number of the pixel point corresponding to the minimum rate of rise in the column;

2. The method for interpreting space debris impact damage based on dynamic multi-target prediction as claimed in claim 1, wherein the third step of determining the size of the block based on the transient thermal response with the minimum rate of rise comprises the following specific steps:

Wherein j_MExpressing the Mth point closest to the pixel point corresponding to the transient thermal response with the minimum rate of rise; always find the first

3. The method for judging space debris impact damage based on dynamic multi-target prediction as claimed in claim 1, wherein the four specific steps of blocking according to the size of the block and determining the search step length in the block comprise:

according to the pixel point number M, N obtained based on the block row and column threshold, the three-dimensional matrix is sequentially decomposed into sub three-dimensional matrix blocks with the number of K and the size of NxM_kS(i_n，j_mT), where k denotes the kth sub-three-dimensional matrix block,i_n、j_mand t respectively represent the ith of the kth sub three-dimensional matrix block_nLine, j_mThe pixel values of the column and the T-th frame, N is 1,2, …, N, M is 1,2, …, M, T is 1,2, … T, and T is the total number of the three-dimensional matrix S frames;

at kth sub-three-dimensional data block_kS(i_n，j_mAnd t), searching around by taking the central point of the sub three-dimensional data block as the center of a circle, finding the central point in the sub three-dimensional data block, and recording the central point as the central point_kS(_ki_N/2，_kj_M/2，_kt) of the first and second groups, wherein,_ki_N/2，_kj_M/2，_kt respectively represents the number of rows, the number of columns and the number of frames of the maximum value pixel points in the kth sub three-dimensional data block;

setting an intra block inner row threshold R _ THV of a kth sub three-dimensional data block_kSequentially calculating the temperature points of the frame and the line where the central point is located in the sub three-dimensional data block_kS(_ki_n′，_kj_M/2，_kt) and the center point in the sub three-dimensional data block_kS(_ki_N/2，_kj_M/2，_kCorrelation of t)

Number of temperature points, is_kRSS, as the intra block row step size of the kth sub-three-dimensional data block;

setting an intra-block column threshold C _ THV for a kth sub-three-dimensional data block_kSequentially calculating the temperature points of the frame and the column of the central point in the sub three-dimensional data block_kS(_ki_N/2，_kj_m′，_kt) and the center point in the sub three-dimensional data block_kS(_ki_N/2，_kj_M/2，_kCorrelation of t)

Wherein_kj_m′Is shown in the kth sub-three-dimensional data block from the center point in the sub-three-dimensional data block_kS(_ki_N/2，_kj_M/2，_kt) the nearest mth pixel point; calculate and count

4. The method for judging space debris impact damage based on dynamic multi-target prediction according to claim 1, wherein the concrete step of selecting transient thermal response in five blocks and step sizes in the step comprises the following steps:

step S51, partitioning the three-dimensional matrix S according to M, N pixel values counted in the step three to obtain K sub three-dimensional data blocks with the size of NxMxT, wherein the K sub three-dimensional data blocks are obtained_kS(i_n,j_mAnd t) represents the ith in the kth sub three-dimensional data block_nLine, j_mTransient thermal response of the column pixels, wherein T is 1,2, …, and T is the total number of S frames of the original three-dimensional matrix;

step S52, for pixel points in each sub three-dimensional data block_kS(i_n,j_mT), setting a threshold DD, initializing a set number g to 1, and initializing a pixel point position i_n＝1,j_m1, and the maximum value in the block_kS(_ki_zz,kj_zz,kt_zz) Corresponding transient thermal response_kS(_ki_zz,kj_zzT), T1, 2, T, stored in the set x (g); then calculating pixel points in the sub three-dimensional data block_kS(i_n,j_mLocated in i) in t)_nLine, j_mCorrelation Re between the transient thermal response of the columns and the set X (g)_i,jAnd judging:

if Re_i,j<DD, theng +1, and responding to transient heat_kS(i_n,j_mT) as a new feature stored in the set X (g); otherwise, let i_n＝i_n+_kRSS, continue to calculate the next transient thermal response_kS(i_n,j_mT), T ═ 1,2,. and the degree of correlation of T with the set x (g); if i_nIf > N, then let i_n＝i_n-N，j_m＝j_m+_kCSS, i.e. change to j_m+_kCSS column is calculated if j_mIf the number of the sub three-dimensional matrixes is more than M, the transient thermal response of the kth sub three-dimensional matrix is selected, and k is k + 1; n, M denotes the number of rows and columns of the kth sub-three-dimensional data block of size nxmxt, respectively.

5. The method for judging space debris impact damage based on dynamic multi-target prediction according to claim 1, wherein the sixth step of classifying the selected transient thermal responses by using an unsupervised clustering algorithm is to classify all sets x (g) of all K data blocks selected in the fifth step, namely the transient thermal responses, into L classes by using a fuzzy C-means clustering (FCM) algorithm, so as to obtain the class to which each transient thermal response belongs, and specifically comprises the following steps:

step S61, setting the number of clusters L, setting the number of initial iterations c to 0, and setting a threshold value for terminating the iterations;

step S62, using formula

Calculating a membership matrix U;

wherein i ═ 1,2, …, L, c ∈ L,_n′d_k′＝||x_k′-_i′V||，n′＝i′，j′，_n′d_k′representing the k 'th pixel point and the i' th cluster center_i′Euclidean distance of V, x_k′Representing the coordinates of the kth pixel point; τ is a constant;_i′u_k′expressing the degree of the k 'th pixel point belonging to the i' th class;

step S63, updating the clustering center_i′V

Wherein the content of the first and second substances,

expressing the thermal response value of the k' th pixel point;

6. The method for judging space debris impact damage based on dynamic multi-target prediction as claimed in claim 1, wherein the step seven of selecting a representative composition matrix Y of each type of transient thermal response based on dynamic multi-target includes the specific steps of:

in step S71, in the (m + 1) th external environment, when the i '(i' ═ 1., L) th class transient thermal response is represented, a multi-target function is defined:

wherein the content of the first and second substances,

Is expressed as:

The calculated Euclidean distance between L-1 classes

The components are renumbered and the components are,

expressed as:

for transient thermal response

The pixel value at time t i.e. the temperature value,

the pixel value of the j' th class transient thermal response clustering center at the t-th moment is a temperature value;

And

And

the number of which is respectively

And

after the environment changes, according to the history information of the m-1 th environment and the m-th environment, the transient thermal response of the initialized population of the approximate leading edge solution set under the m +1 th environment is predicted and calculated, and the steps are as follows:

step 721,

Is from

Randomly selecting N in the solution set_ETransient thermal response

A constructed set of transient thermal responses, N' ═ 1,2_ECalculating

wherein the content of the first and second substances,

is from

Randomly selecting N in the solution set_ETransient thermal response

A constructed set of transient thermal responses, N' ═ 1,2_E；

firstly, the

The center of the transient thermal response is solved and recorded

Wherein the content of the first and second substances,

as a solution set

The nth transient thermal response;

secondly, W representative transient thermal response sets capable of fully describing shape and diversity of the current PS and obtained based on a fully adaptive spectral clustering algorithm

Step B, using CLIQUE clustering algorithm to collect solution

Clustering into cluster sets of transient thermal responses in

The method comprises the following specific steps:

Each dimension of the attribute of the whole T-dimension data space of the transient thermal response in (1) is divided into rectangular area grids which are equally spaced and do not overlap with each other, wherein d represents the dimension of the transient thermal response data; hash _ Gr by Hash tableStoring the dividing condition information of each Grid unit, taking the number of each Grid unit as the key of a hash table, storing a self-defined Grid data type Class _ Grid as the value of the hash table, wherein one Grid Class _ Grid comprises 4 variables, namely count, classID, isChecked and points count, and recording the number of transient thermal response data falling into the Grid unit; the classID is the serial number of the class cluster to which the grid unit belongs, and the default is 0; the isChecked represents the traversed state of the grid unit, the value is 0 or 1, 1 represents that the inspection is finished, and the default value is 0; points is a stringbuffer object for storing coordinate information in the temperature subspace of all transient thermal response data points in the grid;

b2, setting a density threshold value theta, traversing all Grid units in Hash _ Grid of the Hash table, judging whether a variable count in the Grid Class _ Grid of each Grid meets the condition that the count is not less than theta, if so, marking the Grid as a dense Grid unit and adding the dense Grid unit into a candidate dense Grid unit set Bin; if not, marking the grid as a sparse grid unit;

wherein dens (_iu_sub) Indicating the density of the ith sub-cell grid, i.e. falling into a sub-cell_iu_subThe number of transient thermal responses of (a); theta is a density threshold; merging all the sub-unit grids meeting the conditions in the sparse Grid u into a dense Grid u', and updating Hash _ Grid information of a Hash table;

step B5, performing a sliding grid algorithm on the sparse grid cells that do not have contiguous dense grid cells; by 2^TPerforming sliding grid operation on a group of sparse grid units; consider a set of these sparse grid cells without a neighboring dense grid as

Wherein u is_i，i＝1，2，…，2^TRespectively representing the sparse grid cells, firstly calculating the center point coordinate C of each sparse grid_i，i＝1，2，…2^TWherein each center point coordinate is

wherein

And

computing a common vertex C of a sparse grid set_M，

D-dimensional coordinates representing a common vertex of the sparse grid set, wherein each one-dimensional coordinate is:

wherein

And

computing vector coordinates for each sparse grid

dens(u′_S)≥θ

step B6, randomly selecting a grid unit in a candidate dense unit grid set Bin, performing horizontal and vertical expansion by taking the grid as the center according to a greedy algorithm retrieval principle, ensuring that all adjacent and connected dense grid units in the horizontal and vertical directions of the grid are covered, forming a maximum connected rectangular area and marking as C₁Mixing C with₁Setting the classID in the Hash table Hash _ Grid of all grids to be 1 indicates that C is₁All grid cells are clustered into a first class, and C is₁The isshecked variable of all grids in (1); then randomly selecting a dense grid with the isChecked position of 0 in the candidate dense cell grid set Bin, and continuously performing horizontal and vertical expansion to form a maximum connected rectangular area C₂While C is₂Cannot contain a dense grid with the isshecked variable bit already at 1, C₂Class ID of all grids in the tree is set to 2, which means C₂All grid cells are clustered into a second class, and C is then added₂The position of an isChecked variable of all grids is also 1; repeating all the operations until all the grid cells in the candidate dense cell grid set Bin are traversed, namely all the grids in BinThe isshecked variable bits of the cells are all 1;

step B7, storing the clustering information result to the set

wherein

Is as follows

Individual clustering result

The k-th of (a) represents the transient thermal response,

wherein the content of the first and second substances,

for the h-th cluster center, then find the cluster center closest to each transient thermal response point still in the sparse grid

Then the transient thermal response is classified into the category to obtain an updated clustering result

Based on updated clustering results

Wherein the content of the first and second substances,

is as follows

Individual clustering result

So as to obtain a cluster center of

And the size of W, if

Then a new representative transient thermal response is required

Logging in sets as representative of transient thermal responses

In (1),

obtained by the following formula

Wherein the content of the first and second substances,

is as follows

Individual clustering result

I.e. finding all clustered results

Medium transient thermal response

Then is at

And

wherein the content of the first and second substances,

obtained by the method of step S721, step S722, W' is

Collectively representing the number of transient thermal responses;

calculating a predicted direction

Wherein the content of the first and second substances,

is PS multidirectional prediction set

Neutralization of

The nearest transient thermal response is given as the serial number h';

step S724, when the iteration number g' is 0, the initialization population transient heat of the approximate leading edge solution set in the m +1 th environmentThe number of responses is N_pWherein, in the step (A),

the transient thermal response of the initial population is randomly generated in a value range,

wherein h is_nFor transient thermal response

The cluster result

The serial number of (a) is included,

is a obedient mean of 0 and variance of

Normally distributed random number, variance of

The calculation formula of (2) is as follows:

according to the method, the transient thermal response of the initialized population of the approximate leading edge solution set in the (m + 1) th environment is obtained according to the historical information in the previous environment, so that a guide direction is provided for population evolution, and the multi-objective optimization algorithm is helped to make quick response to new changes;

step S73, initializing relevant parameters

Wherein:

initializing reference points

Is a function of

A corresponding reference point;

maximum number of iterations g'_max；

Step S74, use

Wherein

Is a pre-set parameter of the process,

and

the setting of (a) is as follows:

step S75, where N is 1, …, N_P: updating speed according to particle swarm algorithm

And population transient thermal response

Comparison according to a Multi-objective optimization Algorithm

Updating global optimum

Local optimization

And a reference point

From

Middle reservation domination

Removing all quilt

Dominant solution vector if

None of the vectors in (1) dominates

Will be provided with

Adding into

N is N +1, N is less than or equal to N_PThen g '═ g' + 1;

Selecting a representation of class i' transient thermal response_i′REP；

Degree of membership of the kth solution to the l-th objective function:

wherein, F_lFor the value of the l-th objective function,

respectively the maximum and minimum values of the corresponding objective function;

wherein the content of the first and second substances,

approximate solution set for leading edge

The number of solution sets contained, L is the number of objective functions,

is a membership function value;

7. The method for judging space debris impact damage based on dynamic multi-target prediction as claimed in claim 1, wherein the step eight is to judge the actual defect category number based on the spearman correlation coefficient, and the specific method for obtaining the two-dimensional image f (x, y) according to the change of the actual defect category number matrix is as follows:

to be provided with_i′REP and_j′REP, (i', j ═ 1,2, …, L) represents any two transient thermal response representatives, which will_i′REP and_j′temperature element value of each time corresponding to REP_i′REP_t(T ═ 1,2, …, T) and_j′REP_t(T ═ 1,2, …, T) is converted into its descending ranking among the temperature values at the respective transient thermal response vector population instants, the element with the highest temperature value is converted into 1, the element with the lowest temperature value is converted into T, and the remaining temperature value elements are converted into rankings in order of magnitude and denoted as Ra: (a: (b) (b))_i′REP_t) And Ra (_j′REP_t) According to the formula:

calculating the difference Da of element values at corresponding moments between the two transient thermal response representatives, and finally according to a formula:

setting a correlation threshold value theta, comparing the Spireman correlation coefficients between every two, and keeping a transient thermal response representative with minimum correlation if Rs (R) ((R))_i′REP_t，_j′REP_t) If the value is less than theta, keeping the i-th and j-th class transient thermal response representatives, otherwise, removing one class of transient thermal response representatives to obtain an L' -class transient thermal response representative; all the transient thermal response representatives of the L ' class are arranged in columns to form a matrix Y ' of T multiplied by L ', and one column is the pixel value at T moments, namely the temperature value;

starting each frame in the three-dimensional matrix S from the first column, willThe latter column is connected with the end of the former column to form a new column, T columns of pixel values corresponding to the T frames are obtained, then the T columns of pixel values are sequentially placed according to time sequence to form an I multiplied by J row and T column two-dimensional image matrix O, and the two-dimensional matrix O is linearly transformed by a matrix Y, namely:

a two-dimensional image matrix R is obtained, wherein,

and sequentially intercepting each row of the two-dimensional image matrix R according to J columns, and sequentially placing the intercepted J columns according to the rows to form an I multiplied by J two-dimensional image, so that L 'rows obtain L' I multiplied by J two-dimensional images, wherein the images all contain defect areas, and in order to facilitate defect contour extraction, a two-dimensional image with the largest difference between pixel values of the defect area and the non-defect area, namely a two-dimensional image with the largest difference between temperature values is selected and recorded as f (x, y).

8. The method for interpreting space debris impact damage based on dynamic multi-target prediction according to claim 7, wherein in the ninth step, the two-dimensional image f (x, y) obtained in the eighth step is subjected to defect detection and defect region segmentation by using a regional convolutional neural network R-CNN, so that the defect part is positioned and identified and the feature is extracted.