Radar target detection method and system based on cumulative permutation entropy weighting
Through the radar target detection method based on the accumulated arrangement entropy weighting, the spectrum leakage and noise distinction problems under low signal-to-noise ratio conditions are solved, and the significant peak enhancement of the target signal and detection performance are achieved.
Patent Information
- Application Number
- CN202510764362.6
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2025-06-10
- Publication Date
- 2025-08-22
- Estimated Expiration
- 2045-06-10
AI Technical Summary
The existing radar target detection methods have severe spectrum leakage under low signal-to-noise ratio conditions, resulting in less obvious peaks in the target signal, and the arrangement entropy characteristics cannot effectively distinguish pure background noise from target signal + background noise.
Using a radar target detection method based on accumulated arrangement entropy weighting, by dividing the target detection distance into distance units, calculating the product of normalized accumulated arrangement entropy and FFT results, and determining the target detection threshold using detection statistics.
Suppresses spectrum leakage at low pulse counts, enhances the target signal peak, effectively distinguishes pure noise from target signal, and improves detection performance.
Smart Images
Figure CN120275931B_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the field of radar target detection technology, and in particular to a radar target detection method and system based on cumulative permutation entropy weighting. Background Art
[0002] Target detection under low signal-to-noise ratio (SNR) conditions is a critical technical challenge in modern radar research. Traditional radar target detection methods based on fast Fourier transforms (FFTs) improve the SNR through coherent accumulation. However, when the number of pulses is low, severe spectral leakage occurs, resulting in unclear spectral peaks in the target signal and reduced detection performance.
[0003] Furthermore, the permutation entropy feature uses the relative size of adjacent points in a time series, rather than their absolute size, to evaluate the complexity of the time series. This is a simple and robust method. However, in a noisy background, the presence or absence of a target signal does not change the random state of the echo sequence's amplitude and phase. Therefore, this feature cannot be directly used to distinguish pure background noise from a combination of target signal and background noise.
[0004] Therefore, there is an urgent need for a radar target detection method and system based on cumulative permutation entropy weighting to solve the problem that the existing radar target detection method has serious spectrum leakage when the number of pulses is small, resulting in unclear spectrum peak of the target signal. In addition, the scheme characteristics for evaluating the complexity of the time series cannot be directly used to distinguish pure background noise from "target signal + background noise". Summary of the Invention
[0005] The present application provides a radar target detection method and system based on cumulative permutation entropy weighting to address the problem that existing radar target detection methods suffer from severe spectrum leakage when the number of pulses is small, resulting in unclear spectrum peaks of target signals. In addition, the scheme characteristics for evaluating the complexity of time series cannot be directly used to distinguish pure background noise from "target signal + background noise".
[0006] In a first aspect, the present application provides a radar target detection method based on cumulative permutation entropy weighting, the method comprising:
[0007] Divide the target detection distance into several distance units, use the distance units as the distance units to be detected in turn, and determine the reference distance units corresponding to the distance units to be detected;
[0008] According to the echo pulse sequence of the range unit, the normalized cumulative permutation entropy of the echo pulse sequence is calculated;
[0009] Perform fast Fourier transform on the echo pulse sequence to obtain the FFT result, and then take the modulus square of the FFT result of the Doppler channel where the detection range unit is located to obtain the modulus square calculation result;
[0010] The product result is obtained by using the squared modular calculation result and the normalized cumulative permutation entropy;
[0011] Traverse all the distance units to be detected, and obtain the detection statistic corresponding to the current distance unit to be detected based on the average of the product result of the current distance unit to be detected and the product result of the corresponding reference distance unit;
[0012] The detection statistics are used to determine the target detection threshold and then perform radar target detection.
[0013] In one implementation of the present application, determining a reference distance unit corresponding to a distance unit to be detected specifically includes:
[0014] When the distance unit to be detected is the first distance unit, the adjacent distance unit of the distance unit to be detected is determined to be the protection unit, and the R distance units after the protection unit are the reference distance units; wherein R is the total number of preset reference distance units and R is an even number;
[0015] When the distance unit to be detected is not the first distance unit and is located in the first R / 2+1 positions, the two distance units adjacent to the distance unit to be detected are determined as protection units, and the distance units in the first R+3 positions that are not the distance units to be detected and the non-protection units are determined as reference distance units;
[0016] When the distance unit to be detected is not in the front R / 2+1 position and is not in the rear R / 2+1 position, the two distance units adjacent to the distance unit to be detected are determined as protection units, the distance unit in the front R / 2 position of the front protection unit is the reference distance unit, and the distance unit in the rear R / 2 position of the rear protection unit is the reference distance unit;
[0017] When the distance unit to be detected is at the last R / 2+1 position and is not the last distance unit, the two distance units adjacent to the distance unit to be detected are determined as protection units, and the distance units at the last R+3 positions that are not the distance units to be detected and the non-protection units are determined as reference distance units;
[0018] When the distance unit to be detected is the last distance unit, the distance units adjacent to the distance unit to be detected are protection units, and the first R distance units of the protection unit are reference distance units.
[0019] In one implementation of the present application, the normalized cumulative permutation entropy of the echo pulse sequence is calculated according to the echo pulse sequence of the range unit, specifically including:
[0020] For length Echo pulse sequence Perform Doppler search and get the sequence ,in, ;
[0021] will sequence and Upper triangular matrix Multiply to achieve the accumulation of sequence elements and obtain the accumulated sequence , ;
[0022] Calculation sequence The permutation entropy of .
[0023] In one implementation of the present application, the calculation sequence The permutation entropy of , specifically including:
[0024] right Reconstruct the phase space to obtain ,
[0025] ,
[0026] in, is the embedding dimension, is the number of reconstructed vectors;
[0027] Use sequence Each element in the tag sequence The serial number of the element at the corresponding position in;
[0028] will sequence The elements in are sorted in ascending order, and then the sequence is sorted in the order of sorting. Sort and get the index sequence ; Among them, the element ;
[0029] like If there are equal elements in , they are arranged in the original order;
[0030] For each , get the corresponding index sequence , index sequence set In , different index sequences constitute a new index sequence set, denoted as ;in, express The number of different index sequences in ;
[0031] By formula:
[0032] , calculate the probability of each index sequence appearing ;
[0033] in, ;
[0034] By formula:
[0035] , calculate the cumulative permutation entropy ;
[0036] when ,and , When the cumulative permutation entropy reaches its maximum ;
[0037] Using the maximum value of cumulative permutation entropy right Normalize and get the normalized cumulative permutation entropy .
[0038] In one implementation of the present application, a fast Fourier transform is performed on the echo pulse sequence to obtain an FFT result, and then the modulus square of the FFT result of the Doppler channel where the detection range unit is located is taken to obtain a modulus square calculation result, specifically including:
[0039] Perform fast Fourier transform on the echo pulse sequence to obtain the FFT result:
[0040] ; where x represents the echo pulse sequence ;
[0041] The Doppler channel where the detection range unit is located The result is squared modulo:
[0042] , and obtain the result of the modulus square calculation.
[0043] In one implementation of the present application, the product result is obtained by using the modular square calculation result and the normalized cumulative permutation entropy, specifically including:
[0044] By formula:
[0045] , obtain the product result t;
[0046] in, represents the normalized cumulative permutation entropy, Indicates the result of modular square calculation.
[0047] In one implementation of the present application, the detection statistic corresponding to the current distance unit to be detected is obtained according to the mean of the product result of the current distance unit to be detected and the product result of the corresponding reference distance unit, specifically including:
[0048] By formula:
[0049] , calculate the detection statistic T;
[0050] in, Indicates the product result of the current distance unit to be detected, represents the product result of the rth reference distance unit, and R represents the total number of reference distance units.
[0051] In one implementation of the present application, the target detection threshold is determined using detection statistics, specifically including:
[0052] Through MATLAB software, N Monte Carlo simulation experiments are performed to generate the target-free distance unit and reference distance unit, and N detection statistics are obtained, among which, , is the false alarm probability set by the system, Indicates rounding down; sort the N detection statistics from large to small, and take the 100th detection statistic as the target detection threshold.
[0053] In a second aspect, the present application provides a radar target detection system based on cumulative permutation entropy weighting, the system comprising:
[0054] The determination module is used to divide the target detection distance into several distance units, and use the distance units as the distance units to be detected in turn, and at the same time determine the reference distance units corresponding to the distance units to be detected; the calculation module is used to calculate the normalized cumulative permutation entropy of the echo pulse sequence based on the echo pulse sequence of the distance unit; perform fast Fourier transform on the echo pulse sequence to obtain an FFT result, and then take the modulus square of the FFT result of the Doppler channel where the distance unit to be detected is located to obtain a modulus square calculation result; use the modulus square calculation result and the normalized cumulative permutation entropy to obtain a product result; the statistics module is used to traverse all the distance units to be detected, and obtain the detection statistic corresponding to the current distance unit to be detected based on the average of the product result of the current distance unit to be detected and the product result of the corresponding reference distance unit; the detection module is used to use the detection statistic to determine the target detection threshold, and then perform radar target detection.
[0055] In one implementation of the present application, the detection module includes a detection unit,
[0056] It is used to enter N Monte Carlo simulation experiments through MATLAB software, generate the distance unit to be detected and the reference distance unit under the condition of no target, and obtain N detection statistics, among which, , is the false alarm probability set by the system, Indicates rounding down; sort the N detection statistics from large to small, and take the 100th detection statistic as the target detection threshold.
[0057] It can be seen from the above technical solutions that this application has the following advantages:
[0058] 1. Suppresses spectrum leakage at low pulse counts and enhances target signal peaks:
[0059] The complexity of a time series is quantified by calculating the cumulative permutation entropy of the echo pulse sequence. The presence of a target signal alters the periodic structure of the sequence, resulting in a decrease in the permutation entropy. Pure noise, on the other hand, is more random and has a higher entropy. Through normalization, the entropy value can be converted into a weight coefficient in the range of 0-1.
[0060] Traditional FFTs, due to insufficient sampling at low pulse counts, can cause spectrum leakage, drowning out target peaks in noise. By multiplying the FFT modulus squared result (reflecting frequency domain energy) with the normalized cumulative permutation entropy (reflecting time domain structural complexity), the target's frequencies, which have both high energy and low entropy (low complexity), are significantly amplified, while the product of noise frequencies, due to their high entropy, is suppressed. This weighting strategy effectively highlights the target's spectral peaks and reduces the impact of spectrum leakage on detection.
[0061] By using entropy weighting to suppress noise frequencies, the energy of target frequencies is selectively enhanced, making peaks easier for detection algorithms to identify. In low-pulse conditions, energy dispersion caused by spectrum leakage is compensated, reducing the risk of missed detections.
[0062] 2. Distinguishing between pure noise and "target + noise" scenarios:
[0063] Traditional permutation entropy relies solely on the relative sizes of adjacent points and cannot directly distinguish target signals from noise. This application enhances sensitivity to sequence periodicity by accumulating permutation entropy (possibly by introducing time windows or segmented accumulation). The periodic echoes of a target can result in highly random sequences with entropy values significantly lower than pure noise.
[0064] The mean of the product of the reference cell (usually a neighboring cell, assumed to contain no target) is used as an estimate of the background noise. By comparing the statistics of the target cell with the reference cell, the baseline interference of the environmental noise can be eliminated and the target signal can be further distinguished from the noise.
[0065] The accumulation operation captures the local structural changes caused by the target, making the entropy feature the key indicator for distinguishing signal from noise. The detection threshold is dynamically set based on the statistics of the reference unit, improving the algorithm's generalization ability in complex noisy environments.
[0066] 3. Integrating time-frequency domain information to improve detection performance:
[0067] FFT provides frequency-domain energy distribution, while cumulative permutation entropy reflects time-domain structural characteristics. The product of the two integrates information from both time and frequency domains, breaking through the limitations of single-domain analysis.
[0068] The detection statistic is constructed by locally comparing the product results (the mean of the target distance unit and the reference unit). This design suppresses the influence of non-stationary noise while preserving the joint characteristics of the target signal.
[0069] The combination of frequency domain energy and time domain complexity reduces reliance on a single feature and improves detection reliability. Under low signal-to-noise ratio conditions, noise may exhibit localized high energy or low complexity in the time or frequency domain, but the probability of both occurring simultaneously is low. This combined criterion effectively reduces false alarms. BRIEF DESCRIPTION OF THE DRAWINGS
[0070] In order to more clearly illustrate the technical solution of the present invention, the following is a brief introduction to the drawings required for the description. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.
[0071] Figure 1 This is a flow chart of a radar target detection method based on cumulative permutation entropy weighting provided in an embodiment of the present application.
[0072] Figure 2 This is a schematic diagram of a principle for constructing detection statistics provided in an embodiment of the present application.
[0073] Figure 3 This is a detection threshold diagram under different noise power levels provided in an embodiment of the present application.
[0074] Figure 4 This is a spectrum diagram of an 8-pulse range unit pulse sequence to be detected provided in an embodiment of the present application.
[0075] Figure 5 This is a spectrum diagram of a 32-pulse range unit pulse sequence to be detected provided in an embodiment of the present application.
[0076] Figure 6 This is a spectrum diagram of a 128-pulse to-be-detected distance unit pulse sequence provided in an embodiment of the present application.
[0077] Figure 7 This is a comparison chart of the detection performance curves of the two methods when there are 128 pulses provided in an embodiment of the present application.
[0078] Figure 8 This is a comparison chart of the detection performance curves of the two methods when there are 32 pulses provided in an embodiment of the present application.
[0079] Figure 9 This is a comparison chart of the detection performance curves of the two methods when there are 8 pulses provided in an embodiment of the present application.
[0080] Figure 10 This is a schematic diagram of the internal structure of a radar target detection system based on cumulative permutation entropy weighting provided in an embodiment of the present application. DETAILED DESCRIPTION
[0081] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.
[0082] Those skilled in the art should understand that the embodiments described below are merely preferred embodiments of the present disclosure and do not imply that the present disclosure can only be implemented through these preferred embodiments. These preferred embodiments are merely intended to explain the technical principles of the present disclosure and are not intended to limit the scope of protection of the present disclosure. Based on the preferred embodiments provided by the present disclosure, all other embodiments obtained by those of ordinary skill in the art without creative effort should still fall within the scope of protection of the present disclosure.
[0083] It should also be noted that the terms "comprises," "includes," or any other variations thereof are intended to encompass non-exclusive inclusion, such that a process, method, commodity, or apparatus that includes a series of elements includes not only those elements but also other elements not explicitly listed, or includes elements inherent to such process, method, commodity, or apparatus. In the absence of further limitations, an element defined by the phrase "comprises a ..." does not exclude the presence of other identical elements in the process, method, commodity, or apparatus that includes the element.
[0084] The technical solutions proposed in the embodiments of the present application are described in detail below with reference to the accompanying drawings.
[0085] The embodiment provides a radar target detection method based on cumulative permutation entropy weighting, such as Figure 1 As shown, the method provided in the embodiment of the present application mainly includes the following steps:
[0086] Step 110 : Divide the target detection distance into a number of distance units, use the distance units as distance units to be detected in sequence, and simultaneously determine reference distance units corresponding to the distance units to be detected.
[0087] In some embodiments, determining the reference distance unit corresponding to the distance unit to be detected may specifically be:
[0088] When the distance unit to be detected is the first distance unit, the adjacent distance unit of the distance unit to be detected is determined to be the protection unit, and the R distance units after the protection unit are the reference distance units; wherein R is the total number of preset reference distance units and R is an even number;
[0089] When the distance unit to be detected is not the first distance unit and is located in the first R / 2+1 positions, the two distance units adjacent to the distance unit to be detected are determined as protection units, and the distance units in the first R+3 positions that are not the distance units to be detected and the non-protection units are determined as reference distance units;
[0090] When the distance unit to be detected is not in the front R / 2+1 position and is not in the rear R / 2+1 position, the two distance units adjacent to the distance unit to be detected are determined as protection units, the distance unit in the front R / 2 position of the front protection unit is the reference distance unit, and the distance unit in the rear R / 2 position of the rear protection unit is the reference distance unit;
[0091] When the distance unit to be detected is at the last R / 2+1 position and is not the last distance unit, the two distance units adjacent to the distance unit to be detected are determined as protection units, and the distance units at the last R+3 positions that are not the distance units to be detected and the non-protection units are determined as reference distance units;
[0092] When the distance unit to be detected is the last distance unit, the distance units adjacent to the distance unit to be detected are protection units, and the first R distance units of the protection unit are reference distance units.
[0093] Those skilled in the art will appreciate that, no matter where the distance unit to be detected is located, protection units are provided at adjacent positions thereof (such as adjacent units behind the first unit, adjacent units on both sides of the middle unit, etc.).
[0094] The selection of the reference cell avoids the protection cell and the distance cell to be detected itself, ensuring that the reference sample contains only background noise or irrelevant interference. This step prevents the target signal energy from spreading to adjacent cells and contaminating the reference cell. For example, if the target occupies an adjacent cell, the traditional method of directly selecting the adjacent cell as a reference will result in an overestimate of the background noise estimate, thereby raising the detection threshold and causing missed detection. This step ensures the statistical independence of the reference cell by isolating the protection cell. In strong clutter or multi-target scenarios, the impact of the non-uniform environment on the reference cell is reduced, improving the accuracy of the background noise power estimate.
[0095] When the range unit to be detected is the first or last range unit, valid range units exist only on one side. By adjusting the reference unit selection direction (e.g., selecting the first unit backward and the last unit forward) and expanding the number of reference units (R+3), insufficient samples in the boundary area can be compensated.
[0096] For non-boundary intermediate units, a front-to-back symmetrical reference unit selection method (front R / 2 and back R / 2) is adopted to ensure that the number of reference samples is sufficient and the distribution is balanced.
[0097] Traditional methods may lead to noise estimation bias in boundary cells due to an insufficient number of reference cells (e.g., only one-sided selection is possible). This step dynamically adjusts the number and direction of reference cells to ensure that the number of reference samples for the first and last cells is consistent with that for the middle cells, thus avoiding fluctuations in detection performance due to positional differences.
[0098] By combining symmetric sampling (middle unit) with extended sampling (boundary unit), the algorithm's adaptability to non-stationary noise in the radar range dimension is enhanced.
[0099] Preset the total number of reference units R to an even number: Limit the size of the reference units by fixing the R value (such as R=8, 10, 16, etc.) to avoid the computational burden caused by unrestricted search.
[0100] Step 120: Calculate the normalized cumulative permutation entropy of the echo pulse sequence based on the echo pulse sequence of the range unit; perform a fast Fourier transform on the echo pulse sequence to obtain an FFT result, and then take the modulus square of the FFT result of the Doppler channel where the range unit to be detected is located to obtain a modulus square calculation result; and obtain a product result using the modulus square calculation result and the normalized cumulative permutation entropy.
[0101] According to the echo pulse sequence of the range unit, the normalized cumulative permutation entropy of the echo pulse sequence is calculated, which can be specifically:
[0102] For length Echo pulse sequence Perform Doppler search and get the sequence ,in, ;
[0103] will sequence and Upper triangular matrix Multiply to achieve the accumulation of sequence elements and obtain the accumulated sequence , ;
[0104] Calculation sequence The permutation entropy of .
[0105] Among them, the calculation sequence The permutation entropy of , specifically including:
[0106] right Reconstruct the phase space to obtain ,
[0107] ,
[0108] in, is the embedding dimension, is the number of reconstructed vectors;
[0109] Use sequence Each element in the tag sequence The serial number of the element at the corresponding position in;
[0110] will sequence The elements in are sorted in ascending order, and then the sequence is sorted in the order of sorting. Sort and get the index sequence ; Among them, the element ;
[0111] like If there are equal elements in , they are arranged in the original order;
[0112] For each , get the corresponding index sequence , index sequence set In , different index sequences constitute a new index sequence set, denoted as ;in, express The number of different index sequences in ;
[0113] By formula:
[0114] , calculate the probability of each index sequence appearing ;
[0115] in, ;
[0116] By formula:
[0117] , calculate the cumulative permutation entropy ;
[0118] when ,and , When the cumulative permutation entropy reaches its maximum ;
[0119] Using the maximum value of cumulative permutation entropy right Normalize and get the normalized cumulative permutation entropy .
[0120] The echo pulse sequence is subjected to a fast Fourier transform to obtain an FFT result, and then the modulus square of the FFT result of the Doppler channel where the detection range unit is located is taken to obtain a modulus square calculation result, which can be specifically:
[0121] Perform fast Fourier transform on the echo pulse sequence to obtain the FFT result:
[0122] ; where x represents the echo pulse sequence ;
[0123] The Doppler channel where the detection range unit is located The result is squared modulo:
[0124] , and obtain the result of the modulus square calculation.
[0125] The product result is obtained by using the squared modulus calculation result and the normalized cumulative permutation entropy, which can be specifically:
[0126] By formula:
[0127] , obtain the product result t;
[0128] in, represents the normalized cumulative permutation entropy, Indicates the result of modular square calculation.
[0129] Step 130: Traverse all the distance units to be detected, and obtain the detection statistic corresponding to the current distance unit to be detected based on the average of the product result of the current distance unit to be detected and the product result of the corresponding reference distance unit; use the detection statistic to determine the target detection threshold, and then perform radar target detection.
[0130] In some embodiments, the detection statistic corresponding to the current distance unit to be detected is obtained based on the average of the product result of the current distance unit to be detected and the product result of the corresponding reference distance unit, which can be specifically:
[0131] By formula:
[0132] , calculate the detection statistic T;
[0133] in, Indicates the product result of the current distance unit to be detected, represents the product result of the rth reference distance unit, and R represents the total number of reference distance units.
[0134] like Figure 2As shown, taking the distance unit to be detected that is not in the first R / 2+1 positions and not in the last R / 2+1 positions as an example, the two distance units adjacent to the distance unit to be detected are protection units, the distance unit in the first R / 2 positions of the front protection unit is the reference distance unit, and the distance unit in the last R / 2 positions of the last protection unit is the reference distance unit. For the reference distance units in the first R / 2 positions and the reference distance units in the last R / 2 positions, the product results of the reference distance units are calculated. The sum of the reference distance units in the first R / 2 positions is X, and the sum of the reference distance units in the last R / 2 positions is Y. The sum is divided by R to obtain the mean Z of the product results of the reference distance units. The detection statistic corresponding to the current distance unit to be detected is obtained by the ratio of the product result of the current distance unit to be detected and the mean of the product results of the corresponding reference distance units.
[0135] The target detection threshold is determined by using the detection statistic, which can be specifically:
[0136] Through MATLAB software, N Monte Carlo simulation experiments are performed to generate the target-free distance unit and reference distance unit, and N detection statistics are obtained, among which, , is the false alarm probability set by the system, Indicates rounding down; sort the N detection statistics from large to small, and take the 100th detection statistic as the target detection threshold.
[0137] In summary, as a specific embodiment:
[0138] The parameters determined are 、 、 The noise background is complex Gaussian noise with a standard deviation of 1 and a normalized Doppler of 0.2. When the number of pulses is 8, the signal-to-noise ratio is set to -10~10dB; when the number of pulses is 32, the signal-to-noise ratio is set to -10~5dB; when the number of pulses is 128, the signal-to-noise ratio is set to -15~0dB;
[0139] 10 by MATLAB software 6 Monte Carlo simulation experiments were conducted to generate the target-free distance unit data and reference distance unit data, and 10 6 Detection statistics are sorted from large to small, and the 100th value is taken as the detection threshold. Repeat the above simulation experiment under different background noise power levels to obtain the change of detection threshold, as shown in Figure 3 The figure shows that as the noise power level changes, the detection threshold of the CPE-FFT method fluctuates very little and remains approximately unchanged, indicating that the CPE-FFT method has a constant false alarm characteristic.
[0140] Through Monte Carlo simulation experiments, the detection range unit data containing the target plus noise and the reference unit data of pure noise are generated respectively, and the number of pulses is set to 8, 32 and 128 respectively. When the signal-to-noise ratio (SNR) is 0dB, the spectrum of the pulse sequence in the generated detection range unit is as follows Figure 4 (Spectrum of the pulse sequence of the distance unit to be detected with 8 pulses), Figure 5 (Spectrum of the pulse sequence of the distance unit to be detected with 32 pulses), Figure 6 (Spectrum of the pulse sequence of the distance unit to be detected with a pulse number of 128) is shown. Figure 4-Figure 6 It shows that when the number of pulses is small, the spectrum leakage is serious and the signal-to-noise ratio of the Doppler unit where the target is located decreases; when the number of pulses is large, spectrum leakage also exists, but the coherent pulse accumulation is large, and the Doppler unit where the target is located can still maintain a large signal-to-noise ratio.
[0141] Monte Carlo simulation experiments were used to compare the detection performance of the FFT method and the CPE-FFT method. When the number of pulses is 128, the simulation results are as follows: Figure 7 As shown in Figure 2, the detection probability of the CPE-FFT method is slightly higher than that of the FFT method, but the advantage is not obvious. When the detection probability is 0.8, the CPE-FFT method reduces the signal-to-noise ratio requirement by about 0.5dB compared to the FFT method. When the number of pulses is 32, the simulation results are as follows: Figure 8 As shown in Figure 2, the detection performance of the CPE-FFT method is better than that of the FFT method. When the detection probability is 0.8, the signal-to-noise ratio requirement is reduced by about 1.5dB. When the number of pulses is 8, the simulation results are as follows: Figure 9 As shown in the figure, the CPE-FFT method has obvious advantages over the FFT method. When the detection probability is 0.8, the requirement for the signal-to-noise ratio is reduced by about 3dB.
[0142] In addition, this application Figure 10 The embodiment of the present application provides a radar target detection system based on cumulative permutation entropy weighting. Figure 10 As shown, the system provided in the embodiment of the present application mainly includes:
[0143] The determination module 210 is configured to divide the target detection distance into a number of distance units, sequentially use the distance units as distance units to be detected, and simultaneously determine reference distance units corresponding to the distance units to be detected.
[0144] The calculation module 220 is used to calculate the normalized cumulative permutation entropy of the echo pulse sequence based on the echo pulse sequence of the range unit; perform a fast Fourier transform on the echo pulse sequence to obtain an FFT result, and then take the modulus square of the FFT result of the Doppler channel where the range unit to be detected is located to obtain a modulus square calculation result; and use the modulus square calculation result and the normalized cumulative permutation entropy to obtain a product result.
[0145] The statistics module 230 is used to traverse all the distance units to be detected and obtain the detection statistics corresponding to the current distance unit to be detected according to the average of the product result of the current distance unit to be detected and the product result of the corresponding reference distance unit.
[0146] The detection module 240 is used to determine the target detection threshold using the detection statistics, and then perform radar target detection.
[0147] The detection module 240 includes a detection unit,
[0148] It is used to enter N Monte Carlo simulation experiments through MATLAB software, generate the distance unit to be detected and the reference distance unit under the condition of no target, and obtain N detection statistics, among which, , is the false alarm probability set by the system, Indicates rounding down; sort the N detection statistics from large to small, and take the 100th detection statistic as the target detection threshold.
[0149] The above description of the disclosed embodiments is intended to enable one skilled in the art to implement or use the present invention. Various modifications to these embodiments will be readily apparent to one skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention is not limited to the embodiments shown herein but is intended to conform to the widest scope consistent with the principles and novel features disclosed herein.
Claims
1. A radar target detection method based on cumulative permutation entropy weighting, characterized in that: The method comprises: Divide the target detection distance into several distance units, use the distance units as the distance units to be detected in turn, and determine the reference distance units corresponding to the distance units to be detected; Determining the reference distance unit corresponding to the distance unit to be detected specifically includes: When the distance unit to be detected is the first distance unit, the adjacent distance unit of the distance unit to be detected is determined to be the protection unit, and the R distance units after the protection unit are the reference distance units; wherein R is the total number of preset reference distance units and R is an even number; When the distance unit to be detected is not the first distance unit and is located in the first R / 2+1 positions, the two distance units adjacent to the distance unit to be detected are determined as protection units, and the distance units in the first R+3 positions that are not the distance units to be detected and the non-protection units are determined as reference distance units; When the distance unit to be detected is not in the front R / 2+1 position and is not in the rear R / 2+1 position, the two distance units adjacent to the distance unit to be detected are determined as protection units, the distance unit in the front R / 2 position of the front protection unit is the reference distance unit, and the distance unit in the rear R / 2 position of the rear protection unit is the reference distance unit; When the distance unit to be detected is at the last R / 2+1 position and is not the last distance unit, the two distance units adjacent to the distance unit to be detected are determined as protection units, and the distance units at the last R+3 positions that are not the distance units to be detected and the non-protection units are determined as reference distance units; When the distance unit to be detected is the last distance unit, the adjacent distance unit of the distance unit to be detected is the protection unit, and the first R distance units of the protection unit are the reference distance units; According to the echo pulse sequence of the range unit, the normalized cumulative permutation entropy of the echo pulse sequence is calculated; specifically, the following steps are performed: For length Echo pulse sequence Perform Doppler search and get the sequence ,in, ; will sequence and Upper triangular matrix Multiply to achieve the accumulation of sequence elements and obtain the accumulated sequence , ; Calculation sequence The permutation entropy of Perform fast Fourier transform on the echo pulse sequence to obtain the FFT result, then take the modulus square of the FFT result of the Doppler channel where the detection range unit is located to obtain the modulus square calculation result; use the modulus square calculation result and the normalized cumulative permutation entropy to obtain the product result; Traverse all the distance units to be detected, and obtain the detection statistics corresponding to the current distance unit to be detected based on the average of the product results of the current distance unit to be detected and the corresponding reference distance unit; use the detection statistics to determine the target detection threshold, and then perform radar target detection.
2. The radar target detection method based on cumulative permutation entropy weighting according to claim 1, characterized in that: Calculation sequence The permutation entropy of , specifically including: right Reconstruct the phase space to obtain , ， in, is the embedding dimension, is the number of reconstructed vectors; Use sequence Each element in the tag sequence The serial number of the element at the corresponding position in; will sequence The elements in are sorted in ascending order, and then the sequence is sorted in the order of sorting. Sort and get the index sequence ; Among them, the element ; like If there are equal elements in , they are arranged in the original order; For each , get the corresponding index sequence , index sequence set In , different index sequences constitute a new index sequence set, denoted as ;in, express The number of different index sequences in ; By formula: , calculate the probability of each index sequence appearing ; in, ; By formula: , calculate the cumulative permutation entropy ; when ,and , When the cumulative permutation entropy reaches its maximum ; Using the maximum value of cumulative permutation entropy right Normalize and get the normalized cumulative permutation entropy .
3. The radar target detection method based on cumulative permutation entropy weighting according to claim 1, characterized in that: Perform fast Fourier transform on the echo pulse sequence to obtain the FFT result, and then take the modulus square of the FFT result of the Doppler channel where the detection range unit is located to obtain the modulus square calculation result, which specifically includes: Perform fast Fourier transform on the echo pulse sequence to obtain the FFT result: ; where x represents the echo pulse sequence ; The Doppler channel where the detection range unit is located The result is squared modulo: , and obtain the result of the modulus square calculation.
4. The radar target detection method based on cumulative permutation entropy weighting according to claim 1, characterized in that: The product result is obtained by using the squared modular calculation result and the normalized cumulative permutation entropy, including: By formula: , obtain the product result t; in, represents the normalized cumulative permutation entropy, Indicates the result of modular square calculation.
5. The radar target detection method based on cumulative permutation entropy weighting according to claim 1, characterized in that: According to the mean of the product result of the current distance unit to be detected and the product result of the corresponding reference distance unit, the detection statistic corresponding to the current distance unit to be detected is obtained, specifically including: By formula: , calculate the detection statistic T; in, Indicates the product result of the current distance unit to be detected, represents the product result of the rth reference distance unit, and R represents the total number of reference distance units.
6. The radar target detection method based on cumulative permutation entropy weighting according to claim 1, characterized in that: Use detection statistics to determine the target detection threshold, including: Through MATLAB software, N Monte Carlo simulation experiments are performed to generate the target-free distance unit and reference distance unit, and N detection statistics are obtained, among which, , is the false alarm probability set by the system, Indicates rounding down; sort the N detection statistics from large to small, and take the 100th detection statistic as the target detection threshold.
7. A radar target detection system based on cumulative permutation entropy weighting, characterized in that: The system comprises: The determination module is used to divide the target detection distance into a number of distance units, sequentially use the distance units as the distance units to be detected, and simultaneously determine the reference distance units corresponding to the distance units to be detected; wherein, determining the reference distance units corresponding to the distance units to be detected specifically includes: When the distance unit to be detected is the first distance unit, the adjacent distance unit of the distance unit to be detected is determined to be the protection unit, and the R distance units after the protection unit are the reference distance units; wherein R is the total number of preset reference distance units and R is an even number; when the distance unit to be detected is not the first distance unit and is in the first R / 2+1 positions, the two adjacent distance units of the distance unit to be detected are determined to be the protection units, and the distance units of the non-to-be-detected distance units and non-protection units in the first R+3 positions are the reference distance units; when the distance unit to be detected is not in the first R / 2+1 positions and is not in the last R / 2+1 positions, the adjacent distance units of the distance unit to be detected are determined to be the protection units. Two distance units are protection units, the distance units at the first R / 2 positions of the front protection unit are reference distance units, and the distance units at the last R / 2 positions of the rear protection unit are reference distance units; when the distance unit to be detected is at the last R / 2+1 position and is not the last distance unit, the two distance units adjacent to the distance unit to be detected are determined to be protection units, and the distance units at the last R+3 positions that are not distance units to be detected and non-protection units are determined to be reference distance units; when the distance unit to be detected is the last distance unit, the distance unit adjacent to the distance unit to be detected is the protection unit, and the first R distance units of the protection unit are reference distance units; A calculation module is used to calculate the normalized cumulative permutation entropy of the echo pulse sequence based on the echo pulse sequence of the range unit; perform fast Fourier transform on the echo pulse sequence to obtain an FFT result, and then take the modulus square of the FFT result of the Doppler channel where the range unit to be detected is located to obtain a modulus square calculation result; and use the modulus square calculation result and the normalized cumulative permutation entropy to obtain a product result; The normalized cumulative permutation entropy of the echo pulse sequence is calculated according to the echo pulse sequence of the range unit. Specifically, the method includes: For length Echo pulse sequence Perform Doppler search and get the sequence ,in, ;Change the sequence and Upper triangular matrix Multiply to achieve the accumulation of sequence elements and obtain the accumulated sequence , ; Calculate sequence The permutation entropy of A statistics module is used to traverse all the distance units to be detected and obtain the detection statistics corresponding to the current distance unit to be detected based on the average of the product result of the current distance unit to be detected and the product result of the corresponding reference distance unit; The detection module is used to use detection statistics to determine the target detection threshold and then perform radar target detection.
8. The radar target detection system based on cumulative permutation entropy weighting according to claim 7, characterized in that: The detection module includes a detection unit, It is used to enter N Monte Carlo simulation experiments through MATLAB software, generate the distance unit to be detected and the reference distance unit under the condition of no target, and obtain N detection statistics, where, , is the false alarm probability set by the system, Indicates rounding down; sort the N detection statistics from large to small, and take the 100th detection statistic as the target detection threshold.