Unmanned aerial vehicle ISAR instantaneous imaging method based on differential evolution optimization SPWVD

Abstract

The invention discloses an unmanned aerial vehicle ISAR instantaneous imaging method based on differential evolution optimization SPWVD. The method comprises the following steps of performing distance compression, envelope alignment and phase compensation on a received echo signal to obtain processed data; the method comprises the following steps of: processing data, replacing fast Fourier transform (FFT) in an orientation direction in a distance Doppler algorithm with a time-frequency analysis method of smooth pseudo Wiener distribution (SPWVD) by using the processed data to obtain a preliminary distance-instantaneous Doppler image, optimizing two smooth windows of the SPWVD by using a differential evolution (DE) algorithm, and imaging again. According to the method, the inverse synthetic aperture radar algorithm of SPWVD is improved by introducing the differential evolution algorithm optimization process, the imaging quality of the maneuvering target inverse synthetic aperture radar can be effectively improved, the target imaging result is clearer, and the entropy value of the imaging image is smaller.

Description

Technical Field

[0001] The present invention relates to the field of radar imaging technology, and in particular to an unmanned aerial vehicle instantaneous inverse synthetic aperture radar imaging method based on differential evolution optimization SPWVD. Background Art

[0002] In order to meet the needs of inverse synthetic aperture radar (ISAR) imaging of complex maneuvering targets, researchers have proposed a variety of range-instantaneous Doppler (RID) imaging algorithms in recent years to address the defocus and smearing problems of traditional range-Doppler (RD) imaging algorithms when dealing with non-uniformly rotating targets. These methods aim to improve the imaging quality of maneuvering targets, especially in ISAR imaging applications of high-speed, variable-acceleration aircraft such as drones.

[0003] Traditional RD imaging algorithms rely on Fourier transform, which assumes that the target is performing uniform or approximately uniform rotational motion. However, in actual operating environments, the motion patterns of UAVs are often complex and unpredictable, which causes the Doppler frequency in the echo signal to vary over time, thus affecting the quality of imaging. To address this problem, two main types of RID imaging algorithms have been proposed: parametric methods and non-parametric time-frequency analysis methods. The former obtains the instantaneous Doppler frequency by estimating the scatterer parameters, but its computational cost is high and it is easily affected by model mismatch; the latter uses techniques such as short-time Fourier transform (STFT), Wigner-Ville distribution (WVD) and its improved versions such as smoothed pseudo-Wiener distribution (SPWVD) to provide more flexible time-frequency representation, but these methods either have insufficient resolution or have the problem of cross-term interference, which affects the clarity of the final ISAR image. Summary of the invention

[0004] In view of the existing technical problems, the present invention provides a UAV instantaneous inverse synthetic aperture radar imaging method based on differential evolution optimization of SPWVD. This algorithm can obtain better imaging effect when performing ISAR imaging on targets with non-stationary motion.

[0005] In order to achieve the above-mentioned object of the invention, the technical solution adopted by the present invention is:

[0006] A UAV ISAR instantaneous imaging method based on differential evolution optimization of SPWVD is provided, and the method comprises the following steps:

[0007] S1, performing distance compression, envelope alignment and phase compensation on the received echo signal to obtain processed data;

[0008] S2, using an improved range Doppler algorithm to image the data obtained in step 1, wherein the improved range Doppler algorithm uses a SPWVD time-frequency analysis method instead of a fast Fourier transform in azimuth;

[0009] S3, optimizing the two smoothing windows in SPWVD by differential optimization method to obtain the optimization result;

[0010] S4. Use the optimized smoothing window to image the signal again.

[0011] Furthermore, the time-frequency analysis method of SPWVD in step 2 is:

[0012]

[0013] Among them, SPWVD s (t,f) represents smooth pseudo-Wiener distribution, t represents time, f represents frequency, and The function is the instantaneous correlation function of the signal s(t), h(τ) and g(u) are smoothing windows, and the differential optimization algorithm is used to optimize the window in SPWVD.

[0014] Furthermore, the sub-steps of step S3 are as follows:

[0015] S3-1, initialize the parameters, and randomly generate a window population, set the number of spatial populations to NP, and perform the Gth iteration of the smoothing window h(τ) and g(u) on the solution space L G for:

[0016]

[0017] Among them, X i,G represents the individuals of a population, represents the i-th individual in the G-th iteration corresponding to h(τ), represents the i-th individual in the G-th iteration corresponding to g(u), where G represents the number of iterations;

[0018] S3-2, randomly select 3 individuals X from the current window population r1,G , X r2,G , X r3,G And perform difference processing, assuming the variation factor is F v , then we get the variant individual V i,G ;

[0019] S3-3. Each individual selected from the population is crossed with the mutant individual according to probability CR to obtain the individuals required for the experiment. in:

[0020]

[0021] Among them, rand j (0,1) represents a random number between 0 and 1, j rand is a random component, ensuring that at least one dimension of the experimental individuals after crossover is provided by the mutant individuals;

[0022] S3-4, bring the current individual and the experimental individual into the cost function calculation respectively, and select the one with smaller entropy value as the next generation individual, where the entropy function E(I) is:

[0023]

[0024] Where ln represents the natural logarithm, I(i,j) represents the two-dimensional image matrix, and |·| 2 It means to find the sum of squares;

[0025] The cost function f(ρ) is:

[0026] f(ρ)=E(I)

[0027] Among them, ρ represents the independent variable in the cost function, and I represents the two-dimensional image matrix;

[0028] The process of selecting individuals for the next generation is as follows:

[0029]

[0030] S3-5, reaching the number of iterations G max , or when the population optimal solution reaches the predetermined error accuracy, the process ends and the optimal result is output.

[0031] Furthermore, the calculation method of the variant individuals in step 3-2 is:

[0032] V i,G =X r1,G +F v (X r2,G -X r3,G )

[0033] Among them, F v () represents the coefficient of variation in the variation treatment.

[0034] The beneficial effects of the present invention are as follows: the UAV instantaneous inverse synthetic aperture radar imaging method based on differential evolution optimization of SPWVD can optimize the smoothing window, reduce the influence of adding the smoothing window on the time-frequency analysis effect, and thus improve the ISAR imaging quality. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] Figure 1 The specific flow chart of the UAV instantaneous inverse synthetic aperture radar imaging method based on differential evolution optimization of SPWVD;

[0036] Figure 2 This is the model diagram used in the simulation experiment;

[0037] Figure 3 This is the imaging result of the STFT-RID algorithm;

[0038] Figure 4 This is the imaging result of the SPWVD-RID algorithm;

[0039] Figure 5 This is an imaging result diagram of the algorithm provided by the present invention. DETAILED DESCRIPTION

[0040] The specific implementation modes of the present invention are described below so that those skilled in the art can understand the present invention. However, it should be clear that the present invention is not limited to the scope of the specific implementation modes. For those of ordinary skill in the art, as long as various changes are within the spirit and scope of the present invention as defined and determined by the attached claims, these changes are obvious, and all inventions and creations utilizing the concept of the present invention are protected.

[0041] This paper proposes a UAV ISAR instantaneous imaging method based on SPWVD optimized by differential evolution. This method not only inherits the ability of SPWVD in suppressing cross terms, but also further optimizes the design of the smoothing window by introducing the differential evolution algorithm, thereby minimizing the impact of cross terms while maintaining high resolution. The goal of this study is to explore a method that can effectively reduce cross-term interference and maintain or even improve the resolution of ISAR images, providing a new solution for achieving high-quality UAV ISAR instantaneous imaging.

[0042] like Figure 1 As shown, a UAV instantaneous inverse synthetic aperture radar imaging method based on differential evolution optimization of SPWVD includes the following steps:

[0043] S1, performing distance compression, envelope alignment and phase compensation on the received echo signal to obtain processed data;

[0044] S2, the processed data is subjected to the SPWVD time-frequency analysis method instead of the azimuth FFT in the range Doppler algorithm to obtain the SPWVD-RID preliminary imaging;

[0045] S3, optimizing the two smoothing windows in SPWVD by using a differential optimization algorithm to obtain an optimization result;

[0046] S4. Use the optimized smoothing window to image the signal again.

[0047] Furthermore, the sub-steps of step S3 are as follows:

[0048] S3-1. Analyze the processed data by using the SPWVD time-frequency analysis method instead of the azimuth fast Fourier transform (FFT):

[0049]

[0050] in and The function is the instantaneous correlation function of the signal s(t), h(τ) and g(u) are smoothing windows. The differential optimization algorithm is used to optimize the window in SPWVD.

[0051] S3-2, initialize the parameters, and randomly generate a window population. Set the space population size to NP. The solution space after the Gth iteration of the smoothing window h(τ) and g(u) is expressed as follows:

[0052]

[0053] S3-3, randomly select 3 individuals X from the current window population r1,G , X r2,G , X r3,G And perform differential processing and set the variation factor as F v , then we can get the variant individuals:

[0054] V i,G =X r1,G +F v (X r2,G -X r3,G )

[0055] S3-4. In order to increase the diversity of the above population, each individual selected from the population is crossed with the mutant individual according to the probability CR to obtain the individuals required for the experiment. in:

[0056]

[0057] j rand is a random component, ensuring that at least one dimension of the experimental individuals after crossover is provided by the mutant individuals.

[0058] S3-5, bring the current individual and the experimental individual into the cost function calculation respectively, and select the one with smaller entropy value and better effect as the next generation individual, where the entropy value function is:

[0059]

[0060] So the cost function is converted to:

[0061] f(ρ)=E(I)

[0062] The process of selecting individuals for the next generation is as follows:

[0063]

[0064] S3-6, reaching the number of iterations G max , or when the population optimal solution reaches the predetermined error accuracy, the algorithm ends and outputs the optimal result.

[0065] In one embodiment of the present invention, the target model is simulated by Matlab under the same simulation conditions using the STFT-RID algorithm, the SPWVD-RID algorithm, and the UAV instantaneous inverse synthetic aperture radar imaging method based on differential evolution optimization of SPWVD.

[0066] The following simulation conditions are set: the center frequency of the FM signal is 35 GHz and the bandwidth is 10 GHz.

[0067] The simulation target is a quad-rotor drone based on target scattering point simulation. The rotors perform autorotation motion and the wings perform variable speed motion with the drone body. The STFT-RID algorithm (corresponding to Figure 3 ), SPWVD-RID algorithm (corresponding to Figure 4 ) and differential evolution algorithm (DE) improved SPWVD-range-instantaneous Doppler (RID) inverse synthetic aperture radar imaging algorithm (corresponding to Figure 5 ) is used for simulation. The imaging results of the STFT-RID algorithm are shown in Figure 3 As shown in Figure 1, the time-frequency analysis method of short-time Fourier transform in STFT-RID is easily affected by noise and has a fuzzy imaging. The imaging results of the SPWVD-RID algorithm are shown in Figure 1. Figure 4 As shown in Figure 1, the imaging is blurred due to the influence of the window function on the time-frequency analysis effect. The imaging results of the SPWVD-range-instantaneous Doppler (RID) algorithm improved by the differential evolution algorithm (DE) are shown in Figure 1. Figure 5 As shown, due to the addition of the differential evolution algorithm to optimize the smoothing window, the contour is clearer and the imaging is clearer than that of the imaging results of the first two methods.

[0068] Table 1 is a comparison of the results of different algorithms under the same simulation conditions. Image entropy is used to measure the quality of imaging results.

[0069] Table 1

[0070]

[0071] From the data in Table 1, it can be seen that the entropy value of the imaging result of the SPWVD-range-instantaneous Doppler (RID) algorithm improved by the differential evolution algorithm (DE) is lower than that of the other two methods. When the image entropy is smaller, the image quality is better. Therefore, we can infer that the imaging effect of the SPWVD-range-instantaneous Doppler (RID) imaging algorithm improved by the differential evolution algorithm (DE) is better than that of the other two methods under the same test environment, and it can well reduce the influence of noise, thereby obtaining better imaging effects.

[0072] In summary, the improved SPWVD-range-instantaneous Doppler (RID) imaging algorithm based on the differential evolution algorithm (DE) of the present invention reduces the influence of reducing the smoothing window on the imaging effect of the time-frequency analysis method and improves the imaging quality.

Claims

1. A UAV ISAR instantaneous imaging method based on differential evolution optimization of SPWVD, the method comprising the following steps: S1, performing distance compression, envelope alignment and phase compensation on the received echo signal to obtain processed data; S2, using an improved range Doppler algorithm to image the data obtained in step 1, wherein the improved range Doppler algorithm uses a SPWVD time-frequency analysis method instead of a fast Fourier transform in azimuth; S3, optimizing the two smoothing windows in SPWVD by differential optimization method to obtain the optimization result; S4. Use the optimized smoothing window to image the signal again.

2. The method for UAV ISAR instantaneous imaging based on differential evolution optimization of SPWVD as claimed in claim 1, characterized in that: The time-frequency analysis method of SPWVD in step 2 is: Among them, SPWVD s (t,f) represents smooth pseudo-Wiener distribution, t represents time, f represents frequency, and The function is the instantaneous correlation function of the signal s(t), h(τ) and g(u) are smoothing windows, and the differential optimization algorithm is used to optimize the window in SPWVD.

3. The method for UAV ISAR instantaneous imaging based on differential evolution optimization of SPWVD as claimed in claim 1, characterized in that: The sub-steps of step S3 are as follows: S3-1, initialize the parameters, and randomly generate a window population, set the number of spatial populations to NP, and perform the Gth iteration of the smoothing window h(τ) and g(u) on the solution space L G for: Among them, X i,G represents individuals in a population, represents the i-th individual in the G-th iteration corresponding to h(τ), represents the i-th individual in the G-th iteration corresponding to g(u), where G represents the number of iterations; S3-2, randomly select 3 individuals X from the current window population r1,G , X r2,G , X r3,G And perform difference processing, let the variation factor be F v , then we get the variant individual V i,G ; S3-3. Each individual selected from the population is crossed with the mutant individual according to probability CR to obtain the individuals required for the experiment. in: Among them, rand j (0,1) represents a random number between 0 and 1, j rand is a random component, ensuring that at least one dimension of the experimental individuals after crossover is provided by the mutant individuals; S3-4, bring the current individual and the experimental individual into the cost function calculation respectively, and select the one with smaller entropy value as the next generation individual, where the entropy function E(I) is: Where ln represents the natural logarithm, I(i,j) represents the two-dimensional image matrix, and |·| 2 It means to find the square sum; The cost function f(ρ) is: f(ρ)=E(I) Among them, ρ represents the independent variable in the cost function, and I represents the two-dimensional image matrix; The process of selecting individuals for the next generation is as follows: S3-5, reaching the number of iterations G max , or when the population optimal solution reaches the predetermined error accuracy, the process ends and the optimal result is output.

4. The method for UAV ISAR instantaneous imaging based on differential evolution optimization of SPWVD as claimed in claim 1, characterized in that: The calculation method of the variant individuals in step 3-2 is: V i,G =X r1,G +F v (X r2,G -X r3,G ) Among them, F v (·) represents the coefficient of variation in the variation treatment.

Images