Heavy-trailing radar sea clutter parameter estimation method

Abstract

The invention discloses a sea clutter parameter estimation method for a heavy-trailing radar. According to the method, firstly, a planning optimization problem is obtained based on KK distribution of heavy trailing sea clutter data; and then solving the planning optimization problem to obtain a KK distribution parameter set of the heavy trailing sea clutter data. According to the method, the fitting effect of the KK distribution and the sea clutter is good under the condition that the sea peak is strongly expressed.

Description

Technical Field

[0001] The present invention relates to the field of signal processing, and more specifically, to a method for estimating sea clutter parameters of a heavy-tail radar. Background Art

[0002] Sea clutter statistical modeling is a method to examine the random characteristics of data from the perspective of sea clutter echo fluctuations. This method can provide a reference for the design of radar sea target detection algorithms. With the continuous upgrading of radar systems and the increasing complexity of the marine environment, sea clutter is increasingly showing non-Gaussianity, and heavy tailing caused by discrete sea spikes often occurs.

[0003] The traditional statistical modeling of sea clutter adopts a composite structure model. When the sea peak is strong, the fitting effect of the heavy-tailed sea clutter data is not ideal.

[0004] The prior art discloses a method for detecting weak sea clutter signals based on a two-dimensional Duffing oscillator, which combines a quantum genetic algorithm with a stochastic resonance algorithm to enhance sea clutter signals, thereby obtaining an optimal detection result. This method has a poor detection effect on heavy tail sea clutter data. Summary of the invention

[0005] Aiming at the defect that the fitting effect of heavy-tail sea clutter data is not ideal in the prior art, the present invention provides a method for estimating sea clutter parameters of heavy-tail radar.

[0006] The primary purpose of the present invention is to solve the above technical problems. The technical solution of the present invention is as follows:

[0007] A method for estimating sea clutter parameters of a heavy-tail radar, comprising:

[0008] S1: Based on the KK distribution of heavy tailing sea clutter data, the planning optimization problem is obtained;

[0009] S2: Solve the planning optimization problem to obtain a set of KK distribution parameters of heavy-tailed sea clutter data.

[0010] Further, step S1 includes:

[0011] S101: define a probability density function of a KK distribution and obtain a distribution function of a KK distribution;

[0012] S102: Obtain a fitting error function according to the probability density function of the KK distribution and the distribution function of the KK distribution;

[0013] S103: According to the fitting error function, a definable optimization objective function and a planning optimization problem are obtained.

[0014] Furthermore, the probability density function of the KK distribution is as follows:

[0015]

[0016] Where k represents the parameter that affects the tailing characteristics of the KK distribution. They represent v1-1th order second kind modified Bessel function and v2-1th order second kind modified Bessel function respectively, v1 and v2 represent shape parameters, a1 and a2 represent scale parameters, x represents fast-varying sea clutter data, and Γ represents the gamma function.

[0017] Furthermore, the fitting error function is as follows:

[0018]

[0019] in, represents the probability density function of the KK distribution, represents the distribution function of KK distribution, represents the parameter vector to be determined, h(n) represents the statistical histogram of the probability density function of the KK distribution, H(n) represents the statistical histogram of the distribution function of the KK distribution, N represents all sea clutter data, and n represents one sea clutter data.

[0020] Furthermore, the optimization objective function is as follows:

[0021]

[0022] The planning optimization problem is as follows:

[0023] max|g object |

[0024] Among them, λ1 and λ2 represent weight coefficients, and ε1 and ε2 represent fitting error functions.

[0025] Furthermore, the KK distribution is formed by weighting and compounding two single K distributions with different parameters through a weight parameter;

[0026] The probability density function of the K distribution is as follows:

[0027]

[0028] Where p(y) represents the probability density function of the slowly varying component, represents the probability density function of the fast-varying component, x represents the fast-varying sea clutter data, Γ represents the gamma function, y represents the slowly-varying sea clutter data, a represents the scale parameter, v represents the shape parameter, and K v-1 They respectively represent the modified Bessel function of the second kind of order v-1.

[0029] Further, step S2 includes:

[0030] S201: constructing a fitness function according to the planning problem;

[0031] S202: according to the fitness function, set population parameters and randomly create multiple populations, each population containing a plurality of KK distribution parameter combinations;

[0032] S203: Solve using a multi-population genetic algorithm to obtain a KK distribution parameter set.

[0033] Furthermore, the fitness function is as follows:

[0034]

[0035] Among them, λ1 and λ2 represent weight coefficients, f KK (n; θ) represents the probability density function of KK distribution, F KK (n; θ) represents the distribution function of KK distribution, N represents all sea clutter data, h(n) represents the statistical histogram of the probability density function of KK distribution, H(n) represents the statistical histogram of the distribution function of KK distribution, n represents a sea clutter data, and θ represents the parameter vector to be determined.

[0036] Further, step S203 includes:

[0037] S20301: Perform random traversal selection, single-point crossover, random mutation and reinsertion operations on multiple populations respectively;

[0038] S20302: Select a set of parameter combinations with the smallest fitness and a set of parameter combinations with the largest fitness in each population as the first optimal parameter combination and the worst parameter combination; use the first optimal parameter combination of each population as the optimal parameter combination set for the planning optimization problem;

[0039] S20303: Use the first optimal parameter combination in the mth population to replace the worst parameter combination in the m+1th population, and use the first optimal parameter combination in the m+1th population to replace the worst parameter combination in the mth population;

[0040] S20304: extracting a set of parameter combinations with the smallest fitness from the set of optimal parameter combinations as the second optimal parameter combination;

[0041] S20305: Determine whether the number of iterations in which the second optimal parameter combination remains unchanged is higher than the preset number of iterations; if not, execute step S20301; if so, use the optimal parameter combination set as the KK distribution parameter set.

[0042] A device for estimating sea clutter parameters of a heavy-tail radar, comprising:

[0043] Problem acquisition module: Based on the KK distribution of heavy tailing sea clutter data, the planning optimization problem is obtained;

[0044] Problem solving module: solves the planning optimization problem and obtains a set of KK distribution parameters of heavy tailing sea clutter data.

[0045] Compared with the prior art, the present invention has the following beneficial effects:

[0046] The present invention obtains a planning optimization problem based on KK distribution, and obtains a KK distribution parameter set of heavy tail sea clutter data by solving the planning optimization problem. The method has a good fitting effect between KK distribution and sea clutter when the sea peak is strong. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] Figure 1 A flow chart of a method for estimating sea clutter parameters of a heavy-tail radar is provided in an embodiment.

[0048] Figure 2 A flow chart of a planning optimization problem is provided for an embodiment.

[0049] Figure 3 A flow chart for solving a planning optimization problem provided in an embodiment.

[0050] Figure 4 A flowchart of solving a problem using a multi-population genetic algorithm is provided for an embodiment.

[0051] Figure 5 A line graph showing the distribution function of the intensity coefficient variation provided in the embodiment.

[0052] Figure 6 A line graph of the distribution function of the weight parameter changes provided in the embodiment. DETAILED DESCRIPTION

[0053] The drawings are for illustrative purposes only and should not be construed as limiting the present patent;

[0054] In order to better illustrate the present embodiment, some parts in the drawings may be omitted, enlarged or reduced, and do not represent the size of the actual product;

[0055] It is understandable to those skilled in the art that some well-known structures and their descriptions may be omitted in the drawings.

[0056] The technical solution of the present invention is further described below in conjunction with the accompanying drawings and embodiments.

[0057] Example

[0058] like Figure 1 As shown, a method for estimating sea clutter parameters of a heavy-tail radar includes:

[0059] S1: Based on the KK distribution of heavy tailing sea clutter data, the planning optimization problem is obtained;

[0060] S2: Solve the planning optimization problem to obtain a set of KK distribution parameters of heavy-tailed sea clutter data.

[0061] It should be noted that the KK distribution is a weighted composite of two single K distributions with different parameters. Different K distributions can be regarded as Bragg / whitecap components and discrete sea peak components respectively.

[0062] Furthermore, if Figure 2 As shown, step S1 includes:

[0063] S101: define a probability density function of a KK distribution and obtain a distribution function of a KK distribution;

[0064] S102: Obtain a fitting error function according to the probability density function of the KK distribution and the distribution function of the KK distribution;

[0065] S103: According to the fitting error function, a definable optimization objective function and a planning optimization problem are obtained.

[0066] It should be noted that the abbreviation of probability density function is PDF. There is no explicit expression for distribution function.

[0067] Furthermore, the probability density function of the KK distribution is as follows:

[0068]

[0069] Where k represents the parameter that affects the tailing characteristics of the KK distribution. They represent v1-1th order second kind modified Bessel function and v2-1th order second kind modified Bessel function respectively, v1 and v2 represent shape parameters, a1 and a2 represent scale parameters, x represents fast-varying sea clutter data, and Γ represents the gamma function.

[0070] Furthermore, the fitting error function is as follows:

[0071]

[0072] in, represents the probability density function of the KK distribution, represents the distribution function of KK distribution, represents the parameter vector to be determined, h(n) represents the statistical histogram of the probability density function of the KK distribution, H(n) represents the statistical histogram of the distribution function of the KK distribution, N represents all sea clutter data, and n represents one sea clutter data.

[0073] Furthermore, the optimization objective function is as follows:

[0074]

[0075] The planning optimization problem is as follows:

[0076] max|g object |

[0077] Among them, λ1 and λ2 represent weight coefficients, and ε1 and ε2 represent fitting error functions.

[0078] In a specific embodiment, λ1 and λ2 are both 20.

[0079] It should be noted that the KK distribution contains 5 parameters to be determined, and the explicit expression of parameter estimation cannot be obtained using conventional moment estimation, maximum likelihood estimation and other methods. This method is used to decompose the target object parameter estimation problem into a probability density function curve and a distribution function curve fitting error function, and the weight coefficient is used to summarize the two constraint objects into corresponding nonlinear optimization problems, so that the parameter estimation can be accurately obtained.

[0080] In a specific embodiment, all parameter vectors to be determined are All need to be satisfied Among them, θ min ,θ max Represents the spatial range of potential solutions for KK distribution parameters.

[0081] Furthermore, the KK distribution is formed by weighting and compounding two single K distributions with different parameters through a weight parameter;

[0082] The probability density function of the K distribution is as follows:

[0083]

[0084] Where p(y) represents the probability density function of the slowly varying component, represents the probability density function of the fast-varying component, x represents the fast-varying sea clutter data, Γ represents the gamma function, y represents the slowly-varying sea clutter data, a represents the scale parameter, v represents the shape parameter, and K v-1 They respectively represent the modified Bessel function of the second kind of order v-1.

[0085] It should be noted that v represents the shape parameter, which affects the steepness of the peak area of ​​the probability density function curve. a represents the scale parameter, which represents the average intensity of the amplitude distribution of sea clutter data. p(y) represents the probability density function of the slowly varying component, which obeys the gamma distribution; represents the probability density function of the fast-varying component, which obeys the Rayleigh distribution.

[0086] In a specific embodiment, the probability density function of the KK distribution can be defined as:

[0087] f KK (x)=(1-k)f1(x; θ1)+kf2(x; θ2)

[0088] Among them, f1(x; θ1) represents the Bragg / whitecap component, f2(x; θ2) represents the sea peak component; θ1=(v1, a1), θ2=(v2, a2), θ1 and θ2 represent the parameter vectors of the probability density functions of two K distributions respectively, and the parameter vector θ=(θ1, θ2)=(v1, a1, v2, a2, k).

[0089] It should be noted that the weight parameter k affects the tailing characteristics of the KK distribution. If k = 0, the KK distribution degenerates into the K distribution, that is, there is no sea peak component. The different parameter combinations of the two K distributions in the KK distribution can construct various types of probability density function curves. Compared with the KA distribution, the KK distribution has a closed expression, and specific parameter indicators can be given for specific sea clutter, which is helpful to provide a reference for sea clutter analysis and sponge target detection algorithms.

[0090] Furthermore, if Figure 3 As shown, step S2 includes:

[0091] S201: constructing a fitness function according to the planning problem;

[0092] S202: according to the fitness function, set population parameters and randomly create multiple populations, each population containing a plurality of KK distribution parameter combinations;

[0093] S203: Solve using a multi-population genetic algorithm to obtain a KK distribution parameter set.

[0094] It should be noted that the use of multi-population genetic algorithm for solving the problem effectively avoids the stopping phenomenon during the evolution of a single population by controlling different parameters, while expanding the diversity of population evolution methods.

[0095] Furthermore, the fitness function is as follows:

[0096]

[0097] Among them, λ1 and λ2 represent weight coefficients, f KK (n; θ) represents the probability density function of KK distribution, F KK(n; θ) represents the distribution function of KK distribution, N represents all sea clutter data, h(n) represents the statistical histogram of the probability density function of KK distribution, H(n) represents the statistical histogram of the distribution function of KK distribution, n represents a sea clutter data, and θ represents the parameter vector to be determined.

[0098] It should be noted that the abbreviation of distribution function is CDF.

[0099] Furthermore, if Figure 4 As shown, step S203 includes:

[0100] S20301: Perform random traversal selection, single-point crossover, random mutation and reinsertion operations on multiple populations respectively;

[0101] S20302: Select a set of parameter combinations with the smallest fitness and a set of parameter combinations with the largest fitness in each population as the first optimal parameter combination and the worst parameter combination; use the first optimal parameter combination of each population as the optimal parameter combination set for the planning optimization problem;

[0102] S20303: Use the first optimal parameter combination in the mth population to replace the worst parameter combination in the m+1th population, and use the first optimal parameter combination in the m+1th population to replace the worst parameter combination in the mth population;

[0103] S20304: extracting a set of parameter combinations with the smallest fitness from the set of optimal parameter combinations as the second optimal parameter combination;

[0104] S20305: Determine whether the number of iterations in which the second optimal parameter combination remains unchanged is higher than the preset number of iterations; if not, execute step S20301; if so, use the optimal parameter combination set as the KK distribution parameter set.

[0105] The probability of random mutation is 0.001~0.05, and the probability of single-point crossover is 0.7~0.9.

[0106] In a specific embodiment, it is determined whether the number of iterations during which the second optimal parameter combination remains unchanged is higher than a preset number of iterations, thereby effectively avoiding the problem of a local optimal solution occurring in the population due to insufficient number of iterations.

[0107] It should be noted that if Figure 5 , Figure 6 As shown, if the strength coefficient The larger the value is and the larger the weight parameter k is, the more obvious the heavy tailing property of the KK distribution is. The intensity coefficient ρ controls the degree of deviation of the tailing region of the KK distribution relative to the K distribution, and the weight parameter k controls the deviation position and degree of deviation of the tailing region of the KK distribution.

[0108] A device for estimating sea clutter parameters of a heavy-tail radar, comprising:

[0109] Problem acquisition module: Based on the KK distribution of heavy tailing sea clutter data, the planning optimization problem is obtained;

[0110] Problem solving module: solves the planning optimization problem and obtains a set of KK distribution parameters of heavy tailing sea clutter data.

[0111] The same or similar reference numerals correspond to the same or similar components;

[0112] The terms used in the drawings to describe positional relationships are only used for illustrative purposes and should not be construed as limiting this patent;

[0113] Obviously, the above embodiments of the present invention are merely examples for clearly illustrating the present invention, and are not intended to limit the embodiments of the present invention. For those skilled in the art, other different forms of changes or modifications can be made based on the above description. It is not necessary and impossible to list all the embodiments here. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the protection scope of the claims of the present invention.

Claims

1. A method for estimating sea clutter parameters for a heavy tail radar, characterized in that: include: S1: Based on the KK distribution of heavy tailing sea clutter data, the planning optimization problem is obtained; S2: Solve the planning optimization problem to obtain a set of KK distribution parameters of heavy-tail sea clutter data.

2. A method for estimating sea clutter parameters of a heavy tail radar according to claim 1, characterized in that: Step S1 includes: S101: define a probability density function of a KK distribution and obtain a distribution function of a KK distribution; S102: Obtain a fitting error function according to the probability density function of the KK distribution and the distribution function of the KK distribution; S103: According to the fitting error function, a definable optimization objective function and a planning optimization problem are obtained.

3. A method for estimating sea clutter parameters of a heavy tail radar according to claim 2, characterized in that: The probability density function of the KK distribution is as follows: Where k represents the parameter that affects the tailing characteristics of the KK distribution. They represent v1-1th order second kind modified Bessel function and v2-1th order second kind modified Bessel function respectively, v1 and v2 represent shape parameters, a1 and a2 represent scale parameters, x represents fast-varying sea clutter data, and Γ represents the gamma function.

4. A method for estimating sea clutter parameters of a heavy tail radar according to claim 2, characterized in that: The fitting error function is as follows: in, represents the probability density function of the KK distribution, represents the distribution function of KK distribution, represents the parameter vector to be determined, h(n) represents the statistical histogram of the probability density function of the KK distribution, H(n) represents the statistical histogram of the distribution function of the KK distribution, N represents all sea clutter data, and n represents one sea clutter data.

5. The method for estimating sea clutter parameters of a heavy tail radar according to claim 2, characterized in that: The optimization objective function is as follows: The planning optimization problem is as follows: max|g object | Among them, λ1 and λ2 represent weight coefficients, and ε1 and ε2 represent fitting error functions.

6. A method for estimating sea clutter parameters of a heavy tail radar according to claim 2, characterized in that: The KK distribution is a composite of two single K distributions with different parameters weighted by a weight parameter; The probability density function of the K distribution is as follows: Where p(y) represents the probability density function of the slowly varying component, represents the probability density function of the fast-varying component, x represents the fast-varying sea clutter data, Γ represents the gamma function, y represents the slowly-varying sea clutter data, a represents the scale parameter, v represents the shape parameter, and K v-1 They respectively represent the modified Bessel function of the second kind of order v-1.

7. A method for estimating sea clutter parameters of a heavy tail radar according to claim 1, characterized in that: Step S2 includes: S201: constructing a fitness function according to the planning problem; S202: according to the fitness function, set population parameters and randomly create multiple populations, each population containing a plurality of KK distribution parameter combinations; S203: Solve using a multi-population genetic algorithm to obtain a KK distribution parameter set.

8. A method for estimating sea clutter parameters of a heavy tail radar according to claim 7, characterized in that: The fitness function is as follows: Among them, λ1 and λ2 represent weight coefficients, f KK (n; θ) represents the probability density function of KK distribution, F KK (n; θ) represents the distribution function of KK distribution, N represents all sea clutter data, h(n) represents the statistical histogram of the probability density function of KK distribution, H(n) represents the statistical histogram of the distribution function of KK distribution, n represents a sea clutter data, and θ represents the parameter vector to be determined.

9. A method for estimating sea clutter parameters of a heavy tail radar according to claim 7 or 8, characterized in that: Step S203 includes: S20301: Perform random traversal selection, single-point crossover, random mutation and reinsertion operations on multiple populations respectively; S20302: Select a set of parameter combinations with the smallest fitness and a set of parameter combinations with the largest fitness in each population as the first optimal parameter combination and the worst parameter combination; use the first optimal parameter combination of each population as the optimal parameter combination set for the planning optimization problem; S20303: Use the first optimal parameter combination in the mth population to replace the worst parameter combination in the m+1th population, and use the first optimal parameter combination in the m+1th population to replace the worst parameter combination in the mth population; S20304: extracting a set of parameter combinations with the smallest fitness from the set of optimal parameter combinations as the second optimal parameter combination; S20305: Determine whether the number of iterations in which the second optimal parameter combination remains unchanged is higher than the preset number of iterations; if not, execute step S20301; if so, use the optimal parameter combination set as the KK distribution parameter set.

10. A device for estimating sea clutter parameters of a heavy tail radar, applied to the parameter estimation method according to any one of claims 1 to 9, characterized in that: include: Problem acquisition module: Based on the KK distribution of heavy tailing sea clutter data, the planning optimization problem is obtained; Problem solving module: solves the planning optimization problem and obtains a set of KK distribution parameters of heavy tailing sea clutter data.

Images