A linear sliding pulse sorting method and apparatus based on horse herd foraging algorithm

Abstract

This invention discloses a linear slip pulse sorting method and apparatus based on a horse herd foraging algorithm, comprising the following steps: Step S1, obtaining a linear slip pulse sequence containing interference pulses; Step S2, performing a difference operation on the sequence to estimate the first pulse repetition period and the basic slip interval of the linear slip pulse; Step S3, mapping the linear slip pulse sequence onto a plane based on the first pulse repetition period and the basic slip interval; Step S4, randomly selecting two non-coincident points on the plane, establishing a straight line model and calculating the interior point set; Step S5, determining the final interior point set ND on the plane. final Step S6: Based on the final interior point set ND final This invention achieves accurate sorting of linear slip pulses using a herd foraging algorithm. The technical solution employed addresses the problem of poor sorting accuracy caused by pulse loss and interference when sorting linear slip pulses.

Description

Technical Field

[0001] This invention belongs to the field of radar pulse sorting technology, and particularly relates to a linear sliding pulse sorting method and device based on a horse herd foraging algorithm. Background Technology

[0002] Radar signal sorting is a crucial aspect of electronic reconnaissance. Only by separating individual radar pulse sequences from randomly overlapping pulse streams can pulse parameter measurements, operating mode identification, and threat level assessments be performed, providing effective support for subsequent radar jamming allocation. With the continuous development of electronic information technology, radar signal sorting faces increasingly severe challenges. Linear sliding pulses, due to their strong anti-jamming capabilities and low probability of interception, are widely used in new radar systems and concepts. Therefore, accurately sorting linear sliding pulses in complex electromagnetic environments with pulse loss and interference becomes paramount.

[0003] Existing linear sliding pulse sorting methods include sorting methods based on inter-pulse parameters and sorting methods based on deep learning. These methods do not fully consider the influence of electromagnetic environmental factors, resulting in poor sorting accuracy when pulse loss and interference pulses coexist.

[0004] In their paper "Automatic Extraction Algorithm for Linear Slip-Variable Pulses Based on Signal Shaping Law," Meng Xianghao et al. proposed a quadratic modulus-Hough transform-based line extraction algorithm, achieving automatic extraction of linear slip-variable pulse sequences. However, their algorithm employed a traversal loop approach for setting the grid search precision, failing to address the inherent limitations of the Hough transform. If random consistency sampling and Radon transform could be combined with line detection to replace the Hough transform, and a swarm intelligence algorithm could be designed to intelligently optimize its parameters, the accuracy and robustness of the line detection method could be enhanced, effectively improving the sorting success rate of linear slip-variable pulses. Summary of the Invention

[0005] The technical problem to be solved by the present invention is to provide a linear slip pulse sorting method and apparatus based on a horse herd foraging algorithm, so as to solve the problem of poor sorting accuracy caused by pulse loss and pulse interference when sorting linear slip pulses.

[0006] To achieve the above objectives, the present invention adopts the following technical solution:

[0007] A linear sliding pulse sorting method based on a horse herd foraging algorithm includes the following steps:

[0008] Step S1: Obtain a linear slip pulse sequence containing interference pulses;

[0009] Step S2: Perform a difference operation on the sequence to estimate the first pulse repetition period and the basic slip interval of the linear slip pulse;

[0010] Step S3: Based on the first pulse repetition period and the basic slip interval, map the linear slip pulse sequence onto the plane. ;

[0011] Step S4: Randomly select a plane For two non-coincident points, establish a straight line model and calculate the set of interior points;

[0012] Step S5: Determine the plane The final interior set ND final ;

[0013] Step S6: Based on the final interior point set ND final Accurate sorting of linear slip pulses is achieved through a herd foraging algorithm.

[0014] Preferably, step S3 includes:

[0015] The cumulative transform of the first-order difference sequence is calculated based on the first pulse repetition period and the basic slip interval.

[0016] Based on the cumulative transform of the first-order difference sequence, the linear sliding pulse sequence is mapped to the plane through a second-order modulus transform. .

[0017] Preferably, step S5 includes:

[0018] plane The largest interior set ND max The corresponding point is transformed to the Radon plane Ω;

[0019] Transform the local brightness maximum of the Radon plane Ω back to the plane. , to obtain a plane The upper straight line model is used to determine the final interior point set ND. final .

[0020] Preferably, step S6 includes:

[0021] The parameters of linear slip pulse sorting are optimized using a herd foraging algorithm.

[0022] According to the final interior point set ND final By optimizing the parameters, the optimal interior point set ND is obtained. best ;

[0023] ND is removed from linear slip pulse sequences containing interference pulses. best The corresponding pulses in the data are used to achieve accurate sorting of linear sliding pulses.

[0024] The present invention also provides a linear sliding pulse sorting device based on a horse herd foraging algorithm, comprising:

[0025] The acquisition module is used to acquire a linear sliding pulse sequence containing interference pulses;

[0026] The first calculation module is used to perform differential operations on the sequence to estimate the first pulse repetition period and the basic slip interval of the linear slip pulse;

[0027] The processing module maps the linear slip pulse sequence onto a plane based on the first pulse repetition period and the basic slip interval. ;

[0028] The second calculation module is used to randomly select a plane. For two non-coincident points, establish a straight line model and calculate the set of interior points;

[0029] Determine the module, used to determine the plane. The final interior set ND final ;

[0030] The sorting module is used to sort the final interior point set ND. final Accurate sorting of linear slip pulses is achieved through a herd foraging algorithm.

[0031] Preferably, the processing module includes:

[0032] The first calculation unit is used to calculate the cumulative transformation of the first-order difference sequence based on the first pulse repetition period and the basic slip interval.

[0033] The second computational unit is used to map the linear sliding pulse sequence to a plane through a quadratic modulus transformation based on the cumulative transformation of the first-order difference sequence. .

[0034] Preferably, the determining module includes:

[0035] Transformation unit, used to transform a plane The largest interior set ND max The corresponding point is transformed to the Radon plane Ω;

[0036] The defining element is used to transform the local maximum brightness of the Radon plane Ω back to the plane. , to obtain a plane The upper straight line model is used to determine the final interior point set ND. final .

[0037] Preferably, the sorting device includes:

[0038] The third calculation unit is used to optimize the parameters of linear slip pulse sorting using the herd foraging algorithm;

[0039] Optimization unit, used to optimize based on the final interior set ND final By optimizing the parameters, the optimal interior point set ND is obtained. best ;

[0040] The rejection unit is used to reject ND from a linear slip pulse sequence containing interference pulses. best The corresponding pulses in the data are used to achieve accurate sorting of linear sliding pulses.

[0041] The beneficial effects of this invention are as follows:

[0042] The purpose of this invention is to address the problem of poor sorting accuracy caused by pulse loss and interference pulses in linear sliding pulse sorting. This invention combines random consistency sampling and Radon transform to design a new linear sliding pulse sorting algorithm, which increases the accuracy and robustness of linear sliding pulse sorting. At the same time, a herd foraging algorithm is designed, embedding a hunger factor to add a hunger attribute to wild horses, optimizing the herd position, and solving the problems of no upper limit on the number of iterations and threshold setting based on expert experience in the detection algorithm. This allows for high linear sliding pulse sorting accuracy even in the presence of pulse loss and interference pulses. Attached Figure Description

[0043] To more clearly illustrate the technical solution of the present invention, the drawings used in the embodiments are briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0044] Figure 1 This is a flowchart of a linear sliding pulse sorting method based on a horse herd foraging algorithm according to an embodiment of the present invention;

[0045] Figure 2 This is a schematic diagram of the horse herd foraging algorithm according to an embodiment of the present invention. Detailed Implementation

[0046] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0047] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0048] Example 1:

[0049] like Figure 1 As shown, this embodiment of the invention provides a linear sliding pulse sorting method based on a horse herd foraging algorithm, comprising the following steps:

[0050] Step S1: Obtain a linear slip pulse sequence containing interference pulses;

[0051] Step S2: Perform a difference operation on the sequence to estimate the first pulse repetition period and the basic slip interval of the linear slip pulse;

[0052] Step S3: Based on the first pulse repetition period and the basic slip interval, map the linear slip pulse sequence onto the plane. ;

[0053] Step S4: Randomly select a plane For two non-coincident points, establish a straight line model and calculate the set of interior points;

[0054] Step S5: Determine the plane The final interior set ND final ;

[0055] Step S6: Based on the final interior point set ND final Accurate sorting of linear slip pulses is achieved through a herd foraging algorithm.

[0056] As one embodiment of the present invention, step S3 includes:

[0057] The cumulative transform of the first-order difference sequence is calculated based on the first pulse repetition period and the basic slip interval.

[0058] Based on the cumulative transform of the first-order difference sequence, the linear sliding pulse sequence is mapped to the plane through a second-order modulus transform. .

[0059] As one embodiment of the present invention, step S5 includes:

[0060] plane The largest interior set ND max The corresponding point is transformed to the Radon plane Ω;

[0061] Transform the local brightness maximum of the Radon plane Ω back to the plane. , to obtain a plane The upper straight line model is used to determine the final interior point set ND. final .

[0062] As one embodiment of the present invention, step S6 includes:

[0063] The parameters of linear slip pulse sorting are optimized using a herd foraging algorithm.

[0064] According to the final interior point set ND final By optimizing the parameters, the optimal interior point set ND is obtained. best ;

[0065] ND is removed from linear slip pulse sequences containing interference pulses. best The corresponding pulses in the data are used to achieve accurate sorting of linear sliding pulses.

[0066] The linear sliding pulse sorting method of this invention solves the problems of poor sorting accuracy and weak anti-interference ability caused by pulse loss and interference pulses in existing linear sliding pulse sorting methods. This invention designs a new linear sliding pulse sorting method based on line detection, and further designs a horse herd foraging algorithm to intelligently set the key parameters of the sorting method. A hunger attribute is added to the wild horses, embedding a hunger factor and improving the herd update method. The optimized algorithm can accurately detect pulses that conform to a linear pattern, thus achieving linear sliding pulse sorting.

[0067] Example 2:

[0068] like Figure 1 As shown, this embodiment of the invention provides a linear sliding pulse sorting method based on a horse herd foraging algorithm, comprising the following steps:

[0069] Step 1: Generate a linear slip pulse sequence Pulse containing interference pulses, with a pulse loss rate of c. In this embodiment, c = 0.1.

[0070] Pulse = {PDW1, PDW2, ..., PDW} i ,...,PDW num}

[0071] Among them, PDW i This is the pulse descriptor (PDW) for the i-th pulse in the Pulse, where i = 1, 2, ..., num, and num is the number of pulses in the Pulse. i The formula is:

[0072] PDW i ={TOA i BW i ,CF i PA i PW i}

[0073] Among them, TOA i For PDW i The pulse arrival time, BW i For PDW i bandwidth, CF i For PDW icarrier frequency, PA i For PDW i Pulse amplitude, PW i For PDW i The pulse width is used to merge overlapping pulses.

[0074] Step 2: Perform a difference operation on the sequence generated in Step 1 to estimate the first pulse repetition period and the basic slip interval of the linear slip pulse, which includes:

[0075] Step 2.1: Export the Time of Arrival (TOA) from the Pulse result in Step 1:

[0076] TOA = {TOA1, TOA2, ..., TOA} i ,...,TOA num}

[0077] Step 2.2: Calculate the multi-level difference sequence diff of TOA in Step 2:

[0078] diff = {diff1, diff2, ..., diff} j ,...,diff num-1}

[0079] Among them, diff j Let j be the j-th level difference sequence of diff, j = 1, 2, ..., num-1, and let diff be the difference sequence of diff. j The calculation formula is:

[0080] diff j ={TOA j+1 -TOA1,TOA j+2 -TOA2,...,TOA j+k -TOA k TOA num -TOA mum-j}

[0081] ={P1,P2,...,P k ,....,P num-j}

[0082] Among them, TOA k Let TOA be the arrival time of the kth pulse of Pulse. j+k Let P be the arrival time of the (j+k)th pulse of Pulse. k TOA j+k with TOA k The difference, k = 1, 2, ..., num-j.

[0083] Step 2.3: Based on the diff in Step 2.2, calculate the repetitive periodic sequence D, and take the mode of D as the minimum pulse repetition period PRI of the linear slip pulse sequence. min :

[0084] PRI min =mode[D]

[0085] Where mode[g] represents the mode operator, and the formula for calculating the repetitive periodic sequence D is:

[0086]

[0087] Where j = 1, 2, ..., num-1, it represents the j-level difference operation.

[0088] Step 2.4: Calculate the multi-level difference sequence diff_P of diff1 in the repeating periodic sequence D from step 2.3:

[0089] diff_P={diff_P1,diff_P2,...,diff_P l ,...,diff_P num-2}

[0090] Among them, diff_P l Let diff_P represent the l-th level difference sequence of diff1, where l = 1, 2, ..., num-2. l The calculation formula is:

[0091] diff_P l ={P l+1 -P1,P l+2 -P2,...,P l+m -P m ,..,P num-1 -P mum-l-1}

[0092] Among them, P m TOA m+1 with TOA m The difference, P l+m TOA l+m+1 with TOA l+m The difference, m = 1, 2, ..., num-l-1.

[0093] Step 2.5: Based on the multi-level difference sequence diff_P from Step 2.4, calculate the slip interval sequence E, and calculate the mode of the sequence, which serves as the slip interval C of the linear slip sequence.

[0094] C = mode[E]

[0095] The formula for calculating E is:

[0096]

[0097] Step 3: Calculate the cumulative transform of the first-order difference sequence diff1, and map the linear sliding pulse sequence to the plane through a second-order modulus transform. include:

[0098] Step 3.1: Calculate the cumulative transform Trans of the first-order difference sequence diff1 in Step 2.3:

[0099] Trans = {T1,T2,...,T} n ,...,T num-1}

[0100] Among them, T n This represents the cumulative transformation result of the nth element of the first-order difference sequence diff1, where n = 1, 2, ..., num-1, and num is the total number of pulses in the linear sliding pulse sequence. n The calculation formula is:

[0101]

[0102] Where diff1(o) is the o-th element in the first-order difference sequence, o = 1, 2, ..., n.

[0103] Step 3.2: Calculate the quadratic magnitude of Trans from Step 3.1 and map it onto the plane. superior:

[0104]

[0105] Where x and y are linearly slipped pulse sequences in the plane The coordinates on the coordinates, mod(g) is the modulo operator, and ceil(g) is the floor function.

[0106] Step 4: Randomly select a plane For two non-overlapping points, establish a straight line model and calculate the set of interior points. Set a maximum number of iterations K, and continuously update the straight line model and the set of interior points, including:

[0107] Step 4.1: Select a plane Establish a straight line model L based on two non-coincident points a and b. α :

[0108] L α :(b2-a2)x α +(a1-b1)y α +(b1·a2-a1·b2)=0

[0109] Where a1 and a2 are The x and y coordinates of point a on the plane, b1 and b2 are... The x and y coordinates of point b on the plane, x α and y α For line L α The set of x and y coordinates of the points above.

[0110] Step 4.2: Calculation All points on the plane to line L α Interior point deviation vector A:

[0111]

[0112] in, for The distance from the e-th point on the plane to the line L α The distance, e = 1, 2, ..., num, where num is the total number of pulses in the linear slip pulse sequence Pulse. The calculation formula is:

[0113]

[0114] Among them, e1 and e2 are The x and y coordinates of point e on the plane.

[0115] Step 4.3: Based on A in Step 4.2, calculate the interior point decision vector B:

[0116]

[0117] in, for The formula for calculating the decision result of the e-th point on the plane is:

[0118]

[0119] Where Q represents the interior point decision threshold.

[0120] Step 4.4: Add the points corresponding to the non-zero elements of B in Step 4.3 to the interior point set ND:

[0121] ND = [nd1,nd2,...,nd] g ,...,nd z ]

[0122] Among them, nd g Let g represent the point corresponding to the g-th non-zero element in B, where g = 1, 2, ..., z, and z represents... On the plane, belonging to line L α The number of interior points.

[0123] Step 4.5: Select a plane Establish a straight line model for the two points other than points a and b, and find the set of interior points.

[0124] Step 4.6: Return to step 4.2 and repeat steps 4.2 to 4.5 until the number of iterations equals the maximum number of iterations K, and output the maximum interior point set ND. max and linear model L β .

[0125] Step 5: Retain ND, the largest interior set of points on the plane max The corresponding point is transformed to the Radon plane Ω, and the local brightness maximum value of the Ω plane is transformed back. Plane, to obtain A straight line model on a plane determines the final set of interior points ND. final ,include:

[0126] Step 5.1: Retain The largest interior set ND on the plane max For the corresponding points, perform a Radon transform and calculate the interior point brightness vector M:

[0127] M=[R1(ρ,θ),R2(ρ,θ),...,R r (ρ,θ),...,R w (ρ,θ)]

[0128] Among them, R r ND max The Radon transform result of the r-th point in the matrix, r = 1, 2, ..., w, where w is the value of ND. max The number of midpoints, R r The calculation formula is:

[0129]

[0130] in, Representing a plane Double integral operation on, f(u,v) is ND max The gray value corresponding to the midpoint r coordinates (u,v), where δ is the Dirac function:

[0131]

[0132] Where ρ is a plane From the origin to the line L β The distance is θ, where θ is the angle between the perpendicular line from the origin to the line and the x-axis.

[0133] Step 5.2: Calculate the luminance decision vector N based on M from Step 5.1:

[0134]

[0135] in, The brightness decision result for the r-th point on the Ω-plane is calculated using the following formula:

[0136]

[0137] Where S represents the brightness decision threshold.

[0138] Step 5.3: Add the brightness points corresponding to the non-zero elements of N in Step 4.3 to the local brightness vector JL:

[0139] JL=[R1(ρ,θ),R2(ρ,θ),...,R s (ρ,θ),...,R t (ρ,θ)]

[0140] Among them, R s (ρ,θ) represents the brightness point corresponding to the s-th non-zero element in N, where s = 1, 2, ..., t, and t represents the number of points with local maximum brightness on the Ω plane.

[0141] Step 5.4: Calculate the plane based on the corresponding parameters of JL from Step 5.3. The corresponding line reconstruction set

[0142]

[0143] in, R represents the s-th local luminance maximum value on the Ω plane. s The reconstructed linear model (ρ,θ) is given by the following formula:

[0144]

[0145] Where, ρ s and θ s For R s (ρ,θ) are the coordinates of the plane Ω, where u and v are straight lines. The coordinates of the points on the surface. Step 5.5: Calculate the maximum interior set ND. max arrive Standard deviation (BP):

[0146] BP = [bp1, bp2, ..., bp] s ,...,bp t ]

[0147] Among them, bp s For ND max to the straight line The standard deviation of is calculated using the following formula:

[0148]

[0149] Where ω represents the straight line The number of interior points and ND max The ratio of the number in the middle.

[0150] Step 5.6: Output the linear model corresponding to the minimum standard deviation and the final interior point set ND. final .

[0151] Step 6: This invention further designs a herd foraging algorithm to optimize Q in step 4.3, K in step 4.6, and S in step 5.2. The improved herd foraging algorithm is based on the wild horse algorithm, assigning a hunger attribute to each wild horse and changing the wild horse position update mechanism to obtain the optimal solution. ND is removed from the linear slip pulse sequence containing interference pulses generated in step 1. final Corresponding pulses, thus achieving accurate sorting of linear sliding pulses. For example... Figure 2 As shown, it includes the following steps:

[0152] Step 6.1: Set the fitness function of the horse herd foraging algorithm. In this embodiment of the invention, sorting accuracy is specifically used as the evaluation index. The fitness function is defined as follows:

[0153]

[0154] Where ZQ is the number of pulses sorted out, and ZT is the actual number of pulses in the linear sliding pulse sequence.

[0155] Step 6.2: Initialize the horse herd. In Embodiment 1 of this invention, the position of each horse is composed of the interior point decision threshold Q, the maximum number of iterations K, and the brightness decision threshold S. Initialize the position of the μ-th horse in the herd:

[0156] MQ μ =min_MQ μ +U(0,1)×(max_MQ μ -min_MQ μ )

[0157] Among them, MQ μ Let U be the initial position of the μ-th horse in the herd, where μ = 1, 2, ..., G, G is the herd size, U(0,1) is a uniformly distributed random number within (0,1), and min_MQ is the initial position of the μ-th horse in the herd. μ max_MQ represents the lower bound of the position of the μ-th horse in the herd. μ This represents the upper limit of the position of the μth horse. In Embodiment 1 of this invention, the population size G is set to 30, and the maximum number of iterations maxiter is set to 500.

[0158] Step 6.3: Calculate the fitness value of each horse, select η leaders of the herd based on the maximum fitness value, and distribute the remaining horses equally among the leaders. In Embodiment 1 of this invention, η is set to 5.

[0159] Step 6.4: Herd members forage around the leader. The local foraging formula is:

[0160]

[0161] in, Let Z be the new position of the γ-th member in the ε-th group when grazing, and let Z be a uniformly random number in the range [-2, 2], π = 3.14, ST. ε This indicates the location of the leader of group ε. Let H be the current position of the γ-th member in the ε-th group. H is the adaptive foraging factor, calculated using the following formula:

[0162]

[0163] Where U(0,1) is a uniformly random number in the range [0,1]. A random vector between [0,1] The tensor product operator is used, and IDX is the foraging factor index. The calculation formula is as follows:

[0164]

[0165] The find[g] operation returns the index of the index that meets the condition. Let it be a random vector between [0,1], where iter is the current iteration number and maxiter is the maximum iteration number.

[0166] Step 6.5: The leader leads the group members to a location with higher fitness to forage. If the current group occupies a leadership position, the location is updated; otherwise, the group leaves the location.

[0167]

[0168] Among them, ST_new ε The new position of the leader of group ε, SW is the position with the highest current fitness value, J is the leadership factor, and it is determined whether the current group occupies a leadership position in this position based on J.

[0169] Step 6.6: Set the hunger attribute, introduce hungry individuals into the herd, and define their position update formula:

[0170]

[0171] Where new_JE is the new position of the hungry individual, JE represents the current position of the hungry individual, randn(1) is a random number that satisfies the standard normal distribution, p = 1 - iter / maxiter, rand is a random number between [0,1], σ and ξ are the food attractiveness factor and satiety factor, respectively, and are random numbers between [0,1], F is the inverse hyperbolic cotangent function, and W1 and W2 are the food sufficiency factor and hunger factor, respectively. The calculation formula is shown in the following formula:

[0172]

[0173] Wherein, JEF is the fitness value of the starving individual JE.

[0174]

[0175] W2=2×χ×(1-e -|HG-SHG| )

[0176] Where HG represents the hunger level of the horse, G represents the population size, SHG represents the sum of the hunger levels of all horses, υ and χ are food stimulants and horse ghrelin, respectively, and are random numbers between [0,1], and ζ is the food gravitation factor, which is a random number between [0,2]. The calculation formulas for HG and SHG are shown in the following formulas.

[0177]

[0178] SHG = sum(HG)

[0179] Wherein, HH is the hunger factor, and the calculation formula is:

[0180]

[0181] Where UB and LB are the upper and lower bounds of the search space, and WF is the worst fitness value in the current population.

[0182] Step 6.7: Return to step 6.3 and repeat steps 6.3 to 6.6 until the number of iterations iter equals the maximum number of iterations maxiter, and output the global optimal solution.

[0183] In practice, Q, K, and S, optimized by the horse herd foraging algorithm, are set as the interior point decision threshold in step 4.4, the maximum number of iterations in step 4.6, and the brightness decision threshold in step 5.3, to obtain the optimal interior point set ND. best .

[0184] Step 6.8: Remove ND from the linear slip pulse sequence Pulse containing the interference pulse from Step 1. best Corresponding pulses are used to achieve accurate sorting of linear sliding pulses.

[0185] Example 3:

[0186] This invention provides a linear sliding pulse sorting device based on a horse herd foraging algorithm, comprising:

[0187] The acquisition module is used to acquire a linear sliding pulse sequence containing interference pulses;

[0188] The first calculation module is used to perform differential operations on the sequence to estimate the first pulse repetition period and the basic slip interval of the linear slip pulse;

[0189] The processing module maps the linear slip pulse sequence onto a plane based on the first pulse repetition period and the basic slip interval. ;

[0190] The second calculation module is used to randomly select a plane. For two non-coincident points, establish a straight line model and calculate the set of interior points;

[0191] Determine the module, used to determine the plane. The final interior set ND final ;

[0192] The sorting module is used to sort the final interior point set ND. final Accurate sorting of linear slip pulses is achieved through a herd foraging algorithm.

[0193] As one embodiment of the present invention, the processing module includes:

[0194] The first calculation unit is used to calculate the cumulative transformation of the first-order difference sequence based on the first pulse repetition period and the basic slip interval.

[0195] The second computational unit is used to map the linear sliding pulse sequence to a plane through a quadratic modulus transformation based on the cumulative transformation of the first-order difference sequence. .

[0196] As one embodiment of the present invention, the determining module includes:

[0197] Transformation unit, used to transform a plane The largest interior set ND max The corresponding point is transformed to the Radon plane Ω;

[0198] The defining element is used to transform the local maximum brightness of the Radon plane Ω back to the plane. , to obtain a plane The upper straight line model is used to determine the final interior point set ND. final .

[0199] As one embodiment of the present invention, the sorting device includes:

[0200] The third calculation unit is used to optimize the parameters of linear slip pulse sorting using the herd foraging algorithm;

[0201] Optimization unit, used to optimize based on the final interior set ND final By optimizing the parameters, the optimal interior point set ND is obtained. best ;

[0202] The rejection unit is used to reject ND from a linear slip pulse sequence containing interference pulses. best The corresponding pulses in the data are used to achieve accurate sorting of linear sliding pulses.

[0203] The embodiments described above are merely preferred embodiments of the present invention and are not intended to limit the scope of the present invention. Various modifications and improvements made to the technical solutions of the present invention by those skilled in the art without departing from the spirit of the present invention should fall within the protection scope defined by the claims of the present invention.

Claims

1. A linear ramp pulse sorting method based on a herding foraging algorithm, characterized in that, The method comprises the following steps: Step S1, acquiring a linear-ramping pulse sequence containing interference pulses; Step S2, performing a difference operation on the sequence to estimate a first pulse repetition period and a basic ramping interval of the linear-ramping pulses; Step S3, mapping the linear ramp pulse sequence to the plane according to the first pulse repetition period and the basic ramp interval Step S4, randomly selecting a plane The upper two points do not coincide, and the straight line model is established and the inner point set is calculated; Step S5, determining the plane upper final inner point set ND final ; Step S6, according to the final inner point set ND final The linearly sliding pulse is accurately sorted by the herd foraging algorithm.

2. The linear-ramping pulse sorting method based on the herding foraging algorithm of claim 1, wherein, Step S3 comprises: According to the first pulse repetition period and the basic ramping interval, calculating a cumulative transform of a first-order difference sequence; According to a cumulative transformation of the first-order difference sequence, the linear ramp pulse sequence is mapped to the plane by a quadratic modulo transformation 3. The linear-ramping pulse sorting method based on the herding foraging algorithm of claim 2, wherein, Step S5 comprises: The plane The upper maximal inner point set ND max The corresponding points are transformed to the Radon plane Ω; transforming local brightness maxima of a Radon plane Ω back to a plane obtaining a plane the upper straight line model to determine the final set of inliers ND final .

4. The linear-ramping pulse sorting method based on the herding foraging algorithm of claim 3, wherein, Step S6 comprises: Optimizing parameters of linear-ramping pulse sorting by a horse herd foraging algorithm; According to the final inner point set ND final and the optimization parameter, the optimal inner point set ND best is obtained. Rejecting ND in a linear-ramping pulse sequence containing interfering pulses by rejecting the corresponding pulses in best best , enabling accurate sorting of linear-ramping pulses.

5. A linear ramped pulse sorting device based on a herding foraging algorithm, characterized in that, The method comprises: An acquisition module is configured to acquire a linear-ramping pulse sequence containing interference pulses; A first calculation module is configured to perform a difference operation on the sequence to estimate a first pulse repetition period and a basic ramping interval of the linear-ramping pulses; a processing module for mapping the linearly ramped pulse sequence to the plane according to the first pulse repetition period and the base ramp interval The second computing module is configured to randomly select a plane The upper two points that do not coincide are used to establish a straight line model and calculate an inner point set. determining module configured to determine a plane the uppermost final inlier set ND final ; a sorting module for sorting the final set of inliers ND according to the distance between the inliers and the center of the circle C final The accurate sorting of linearly sliding pulses is realized by the herding algorithm.

6. The linear ramped pulse sorting device based on horse herd foraging algorithm as claimed in claim 5, wherein, The processing module comprises: A first calculation unit is configured to calculate a cumulative transform of a first-order difference sequence according to the first pulse repetition period and the basic ramping interval; a second computing unit for mapping the linearly ramped pulse sequence to the plane by a quadratic modulo transformation according to a cumulant transformation of the first-order difference sequence 7. The linear-ramping pulse sorting device based on the herdsman foraging algorithm of claim 6, wherein, The determination module comprises: a transformation unit for transforming the planar the upper maximal inner point set ND max the corresponding points are transformed to the Radon plane Ω; determining unit for transforming the local luminance maximum of the Radon plane Ω back to the plane obtaining a plane the upper straight line model to determine the final set of inliers ND final .

8. The linear-ramping pulse sorting device based on the herding foraging algorithm of claim 7, wherein, The sorting device comprises: A third calculation unit is configured to optimize parameters of linear-ramping pulse sorting by a horse herd foraging algorithm; an optimization unit for obtaining an optimal set of interior points ND from the final set of interior points ND final and the optimization parameters best ; The rejection unit is used to reject ND from a linear slip pulse sequence containing interference pulses. best The corresponding pulses in the data are used to achieve accurate sorting of linear sliding pulses.

Images