A long-time accumulation-based air-space high-speed target detection method

Abstract

The application belongs to the technical field of radar signal processing, and specifically discloses a method for detecting airspeed targets based on long-time accumulation, which comprises the following steps: firstly, segmenting the echo signals after pulse compression according to the number of pulses, and compensating the distance migration and Doppler spread of the envelope of the echo signals in each segment; secondly, performing in-segment coherent accumulation on the compensated results in each segment; and finally, performing envelope migration and non-coherent accumulation between segments on all the echo signals, so that the energy of the echo signals can be effectively accumulated. The long-time accumulation method of time segmentation, in-segment coherent accumulation and inter-segment non-coherent accumulation can solve the problems of long-time coherent accumulation calculation and echo signal coherent deterioration, improve the detection performance, and is easy to implement in engineering.

Description

Technical Field

[0001] This invention relates to the field of radar signal processing technology, and in particular to a method for detecting high-speed targets in the air based on long-term accumulation, which can be used to improve accumulation gain, reduce computational load, and is easy to implement in engineering. Background Technology

[0002] With the development of aerospace technology, hypersonic vehicles have attracted the attention of major military powers due to their extremely attractive military application prospects. Integrating new technologies from numerous disciplines in the aerospace field, they represent the future research and development direction of aerospace and are considered a key technological area after stealth technology in the military field. Hypersonic vehicles generally fly in near-space between 20km and 100km, possessing characteristics such as long operational range, high speed, and strong maneuverability. They can penetrate radar beams in a very short time, breaking through existing airspace defense systems and posing a significant threat to national airspace security. Therefore, research on high-speed, highly maneuverable target detection technology is of great significance in the face of various penetration methods employed by near-space hypersonic vehicles.

[0003] Typically, increasing observation time and employing long-term accumulation techniques can improve the target's output signal-to-noise ratio, trading time for energy to enhance radar detection performance against such targets. However, during long-term accumulation, the high-speed, highly maneuverable nature of near-space high-speed targets leads to range migration and Doppler spread, severely degrading the effectiveness of traditional coherent accumulation. Furthermore, considering that high-speed target motion in the atmosphere causes air ionization and the generation of a plasma sheath, further degrading the coherence of the echo signal, continuing long-term coherent accumulation under these conditions would actually worsen the results. Moreover, from a hardware implementation perspective, long-term coherent accumulation requires a large amount of equipment, resulting in prohibitively high costs. Therefore, researching effective and easily engineering-implementable long-term accumulation algorithms for detecting near-space high-speed targets is of urgent practical significance. Summary of the Invention

[0004] To address the problems existing in the prior art, the present invention aims to provide a long-term accumulation-based method for detecting high-speed targets in the airspace. This method employs a long-term accumulation approach that combines time segmentation, intra-segment coherent accumulation, and inter-segment non-coherent accumulation. This approach solves both the computational burden of long-term coherent accumulation and the problem of deteriorating echo signal coherence. It improves detection performance and is easy to implement in engineering.

[0005] The technical principle of this invention is as follows: First, the pulse-compressed echo signal is segmented according to the number of pulses, and the distance migration and Doppler spread of the echo signal envelope within each segment are compensated; then, the compensated results within each segment are coherently accumulated; finally, the envelope shift and non-coherent accumulation between segments are performed on all echo signals, so that the echo signal energy can be effectively accumulated.

[0006] To achieve the above objectives, the present invention employs the following technical solutions.

[0007] A method for detecting high-speed targets in the airspace based on long-term accumulation includes the following steps:

[0008] Step 1: Establish a high-speed maneuvering target motion model, construct a radar echo data model under long-term observation conditions, and sequentially perform down-conversion and low-pass filtering on the raw echo data to obtain the baseband echo data s received by the radar. r (t k , t m Then, a matched filter is used to process the baseband echo data s. r (t k , t m Pulse compression was performed to obtain the frequency-slow time domain echo data S after pulse compression. P (f,t) m );

[0009] Where m = 0, 1, ..., N-1, and N represents the total number of accumulated pulses; t k To save time, t m Slow time;

[0010] Step 2, convert the pulse compression frequency-slow time domain echo data S according to the slow time dimension. P (f,t) m Divided into M echo data segments

[0011] Here, the repetition interval of each Q pulse is considered as a coherent processing time period, where Q is an integer power of 2, and m i =(i-1)×Q, (i-1)×Q+1,..., i×Q-1, N=M×Q;

[0012] Step 3: Construct the traversal interval and step size for the search speed, based on each search speed v. s Construct the corresponding phase compensation factor to form the compensation matrix H v (v s ), and together with the first echo data Multiply the data to obtain the first segment of compensated echo data and estimate the target velocity. Based on the target velocity estimate, form the optimal compensation matrix and compensate for each echo data segment separately to obtain M phase-compensated echo data segments. Transform the M phase-compensated echo data segments to the range-time domain using IFFT to obtain the phase-compensated time-domain signal.

[0013] Step 4: Use the de-line frequency modulation method to estimate the Doppler frequency modulation of the time-domain signal, that is, estimate the target acceleration. Construct a Doppler compensation term based on the target acceleration estimate and compensate all phase-compensated time-domain signals to achieve Doppler extension compensation.

[0014] Step 5: Perform Q-point Fast Fourier Transform on each range cell of the M echo data segments after compensation in Step 4, and concentrate the energy in each echo data segment into a range cell and a Doppler cell respectively to achieve coherent accumulation within the data segment, and obtain the M echo data segments after coherent accumulation.

[0015] Step 6: Determine the range cell difference and Doppler channel difference between different echo data, construct a range compensation factor using the inter-segment range cell difference, and perform range compensation on the M echo data segments after coherent accumulation to obtain the M echo data segments after inter-segment envelope alignment.

[0016] Step 7: Normalize the M echo data segments after inter-segment envelope alignment to obtain M normalized echo data segments; using the inter-segment Doppler channel difference determined in Step 6, extract the data corresponding to the Doppler channel of each data segment in the M normalized echo data segments, take the modulus value, and then add them together to achieve inter-segment non-coherent accumulation, obtaining echo data after inter-segment non-coherent accumulation; target detection can then be performed on the echo data after inter-segment non-coherent accumulation.

[0017] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0018] (1) Long-term coherent accumulation techniques are limited for highly maneuverable targets such as aircraft. Long-term coherent accumulation techniques assume that the target's motion is stable and its radial velocity remains constant. However, during long-term detection, the target's trajectory and velocity usually change, making coherent accumulation impossible. This invention solves the problem of computational complexity in long-term coherent accumulation and the problem of deteriorating echo signal coherence by employing a long-term accumulation method that uses time segmentation, intra-segment coherent accumulation, and inter-segment non-coherent accumulation. At the same time, it can reduce the amount of traversal search for target velocity and acceleration when performing intra-segment range travel correction and Doppler spread compensation, thereby reducing the computational complexity of the algorithm and making it easier to implement in engineering.

[0019] (2) If MTD detection is performed on all echoes over a long period, even if the target does not move, its velocity changes will cause its energy to be dispersed across multiple velocity units. This is because the velocity (Doppler) units are divided too finely. This invention uses time segmentation, intra-segment coherent accumulation, and inter-segment non-coherent accumulation to prevent velocity splitting of the target. After segmenting the signal, the number of coherent accumulation pulses within each segment is small, and the velocity units are coarse. When the target's velocity changes little, it still falls within a single velocity unit. By superimposing the magnitudes of the coherent accumulation results of multiple batches of data over multiple time periods to achieve non-coherent accumulation, velocity splitting will also not occur. Attached Figure Description

[0020] The present invention will now be described in further detail with reference to the accompanying drawings and specific embodiments.

[0021] Figure 1 This is a flowchart illustrating the implementation of the present invention;

[0022] Figure 2(a) is a contour map of the first echo signal after pulse compression in this invention;

[0023] Figure 2(b) is a contour map of the first echo signal after distance travel correction in this invention;

[0024] Figure 3(a) is a diagram showing the result of coherent accumulation of the first echo signal after distance travel correction in this invention;

[0025] Figure 3(b) is a diagram showing the result of coherent accumulation of the first echo signal after Doppler spread compensation in this invention;

[0026] Figure 4(a) shows the peak positions of the first, fourth, and eighth segments of the echo signal after intra-segment coherent accumulation in this invention.

[0027] Figure 4(b) shows the results of the peak positions of the first, fourth, and eighth segments of the echo signal after inter-segment envelope shift in this invention.

[0028] Figure 5 This is a diagram showing the result of the echo signal after inter-segment non-coherent accumulation in this invention.

[0029] Figure 6 This is a comparison chart showing the trend of algorithm computational complexity of the proposed method and the traditional long-term coherent accumulation method as the number of accumulation pulses changes. Detailed Implementation

[0030] The embodiments of the present invention will be described in detail below with reference to examples. However, those skilled in the art will understand that the following examples are only for illustrating the present invention and should not be regarded as limiting the scope of the present invention.

[0031] Reference Figure 1The present invention provides a method for detecting high-speed targets in the airspace based on long-term accumulation, comprising the following steps:

[0032] Step 1: Establish a high-speed maneuvering target motion model, construct a radar echo data model under long-term observation conditions, and sequentially perform down-conversion and low-pass filtering on the raw echo data to obtain the baseband echo data s received by the radar. r (t k , t m Then, a matched filter is used to process the baseband echo data s. r (t k , t m Pulse compression was performed to obtain the frequency-slow time domain echo data S after pulse compression. P (f,t) m ); where m = 0, 1, ..., N-1, and N represents the total number of accumulated pulses; t k To save time, t m Where f is the slow time interval and f is the frequency.

[0033] Specifically, it includes:

[0034] 1.1 Establishing a high-speed maneuvering target motion model: For a moving target, the distance between the target and the radar changes with slow time t. m Time-varying, can be represented as t m A polynomial function, when expanded by a Taylor series, can be expressed as:

[0035]

[0036] Where P is the order of the target motion, α p Let P be the motion parameter of order p. In this invention, the first four terms of the Taylor series are retained, i.e., P = 3, then the above equation can be simplified to:

[0037]

[0038] In the formula, R(t) m R0 represents the distance between the target and the radar during the m-th pulse repetition period, v represents the target's initial slant range, a represents the target's radial velocity, and a represents the target's radial acceleration.

[0039] 1.2 Constructing a radar echo data model under long-term observation conditions:

[0040] Under narrowband point target conditions, the model of the baseband echo signal received by the radar is constructed as follows:

[0041]

[0042] Where σ0 represents the reflection coefficient of the target, For a rectangular window function, Tp B is the rectangular pulse width, and B is the modulation bandwidth. To adjust the frequency, f c Let t be the carrier frequency, j be the imaginary unit, and c be the speed of electromagnetic wave propagation in air; let t m =mT r (m = 0, 1, ..., N-1) represents slow time, T r denoted as the pulse repetition interval, and N as the total number of accumulated pulses.

[0043] The time-domain expression of the matched filter is: The frequency-slow time domain expression of the echo signal after pulse compression is:

[0044]

[0045] Where A0 is the amplitude of the pulse after compression.

[0046] Step 2, convert the pulse compression frequency-slow time domain echo data S according to the slow time dimension. P (f,t) m Divided into M echo data segments Here, the repetition interval of each Q pulse is considered as a coherent processing time period, where Q is an integer power of 2, and m i =(i-1)×Q, (i-1)×Q+1,..., i×Q-1, N=M×Q;

[0047] For echo data S P (f,t) m When segmenting, the real-time processing capability of the actual hardware device and the search accuracy of velocity and acceleration in steps 3 and 4 should be considered; at the same time, the number of pulses Q in each segment should satisfy an integer power of 2 to facilitate Fast Fourier Transform (FFT).

[0048] Step 3: Construct the traversal interval and step size for the search speed, based on each search speed v. s Construct the corresponding phase compensation factor to form the compensation matrix H v (v s ), and together with the first echo data Multiply the data to obtain the first segment of compensated echo data and estimate the target velocity. Based on the target velocity estimate, form the optimal compensation matrix and compensate for each echo data segment separately to obtain M phase-compensated echo data segments. Transform the M phase-compensated echo data segments to the range-time domain using IFFT to obtain the phase-compensated time-domain signal.

[0049] First, based on radar system parameters and prior information about the target, a method for estimating the target's velocity is designed:

[0050] 3.1 Determine the maximum speed v for the search traversal based on the prior information of the target (the maximum possible flight speed of the detected target). max ;

[0051] 3.2, Determine the maximum search traversal interval Δv for velocity based on radar system parameters. During velocity search, when the search velocity v... s When matched with the target's true velocity v, the following relationship holds:

[0052] |vv s |≤Δv s / 2

[0053] Where, Δv s The search interval is the speed. When the search speed v... s When matching the target's true velocity v, after range travel correction using the frequency domain phase compensation method, the distance between the peak value of the first echo pulse and the peak value of the last echo pulse cannot exceed one range cell, i.e., satisfying the following relationship:

[0054]

[0055] Where Q represents the number of pulses in each segment after segmentation, c is the speed of light, B is the signal bandwidth, and T is the signal strength. r This is the pulse repetition period.

[0056] Because the speed found differs from the target's actual speed by Δv s / 2, After compensation, the distance traveled due to the remaining velocity of the k-th echo signal is:

[0057]

[0058] Among them, R res = c / 2B, which is the range resolution unit.

[0059] To further ensure distance positioning between segments, the search interval needs to be reduced by a factor of M-1. Therefore, the maximum search interval for velocity is set to...

[0060] Δv=c / (QBT r ) / (M-1)

[0061] 3.3, Based on steps 3.1 and 3.2, the search traversal range for speed is determined to be: [-v max ,-Δv]∪[Δv,v max ]

[0062] 3.4, based on search speed v s Constructing the phase compensation factor:

[0063] Hv (v s )=exp(j2πf·2v s m1T r / c)

[0064] Where m1 = 0, 1, ..., Q-1;

[0065] Phase compensation factor H v (v s For the first segment of echo data Perform frequency domain compensation, then perform inverse fast Fourier transform (IFFT) along the fast time frequency to obtain the compensated (t) k -t m The first segment of echo data in the domain:

[0066]

[0067] Where sin c(x)=sin(πx) / (πx) represents the Sa function, and λ=c / f c This is the radar wavelength.

[0068] Ignoring the distance traveled due to acceleration, then when v s When the search velocity equals the target's true velocity, the peak values ​​of the first echo signal envelope are located within the same distance cell, from which the target velocity estimate can be obtained. Based on the target velocity estimate Construct the optimal compensation matrix Frequency domain compensation is performed on all echo data segments, and IFFT is performed along the fast time frequency to obtain the phase-compensated time domain signal:

[0069]

[0070] Where i = 1, 2, ..., M, This represents the time-domain signal after compensation for the i-th segment.

[0071] Step 4: Use the de-line frequency modulation method to estimate the Doppler frequency modulation of the time-domain signal, that is, estimate the target acceleration. Construct a Doppler compensation term based on the target acceleration estimate and compensate all phase-compensated time-domain signals to achieve Doppler extension compensation.

[0072] This process is similar to step 3, specifically as follows:

[0073] 4.1 Determine the maximum acceleration 'a' for the search traversal based on the target's prior information (the maximum possible flight acceleration of the target being detected). max ;

[0074] 4.2, Determine the maximum search traversal interval Δa for acceleration based on radar system parameters. When the searched acceleration a... s When matched with the target's true acceleration 'a', the following relationship holds:

[0075] |aa s |≤Δa s / 2

[0076] Where, Δa s This is the acceleration search interval. Therefore, after compensating the echo signal with the searched acceleration, the velocity change caused by the remaining acceleration should not exceed one Doppler unit, i.e.

[0077]

[0078] Right now

[0079] Δa s ≤λ / (QT r ) 2

[0080] Where Q represents the number of pulses in each segment after segmentation, λ is the wavelength, and f r Let be the pulse repetition frequency. Using the above search method for acceleration, the maximum difference between the searched acceleration and the target's true acceleration is Δa. s / 2. After compensation, the velocity change caused by the target's residual acceleration is shown below:

[0081]

[0082] Where k represents the k-th signal after segmentation, v res This represents the Doppler resolution unit.

[0083] Since the Doppler cells of the target segment obtained after acceleration compensation using the above search method are different, to prevent this from happening, the search interval is reduced by a factor of M-1, where M is the final number of segments. Therefore, the maximum search interval for acceleration is:

[0084] Δa s =λ / (QT) r ) 2 / (M-1)

[0085] 4.3, Based on steps 4.1 and 4.2, the search traversal range for acceleration is determined to be: [-a max ,-Δa]∪[Δa,a max ];

[0086] 4.4, based on the search acceleration a s Obtain the search frequency γ s=-2a s / λ, constructing the compensation factor And the first echo data Frequency domain compensation is performed, and then the maximum energy peak is searched using the following formula to calculate the optimal tuning frequency:

[0087]

[0088] in, This represents the slow-time dimension signal corresponding to the distance cell where the target is located.

[0089] when equal The true modulation frequency γ a When =-2a / λ, an energy peak will appear, and the frequency corresponding to this peak is the true frequency modulation.

[0090] Then, the compensation factor constructed using the actual frequency modulation was used. Compensation is performed on all echo data segments after distance travel correction to obtain:

[0091]

[0092] Where i = 1, 2, ..., M, This represents the echo after compensation for the i-th segment. It can be seen from the formula that the higher-order time terms corresponding to acceleration in the signal have been eliminated, and the Doppler spread has been compensated.

[0093] Step 5: Perform Q-point Fast Fourier Transform on each range cell of the M echo data segments after compensation in Step 4, and concentrate the energy in each echo data segment into a range cell and a Doppler cell respectively to achieve coherent accumulation within the data segment, and obtain the M echo data segments after coherent accumulation.

[0094] Step 6: Determine the range cell difference and Doppler channel difference between different echo data, construct a range compensation factor using the inter-segment range cell difference, and perform range compensation on the M echo data segments after coherent accumulation to obtain the M echo data segments after inter-segment envelope alignment.

[0095] 6.1, from step 4 The expression shows that the distance cell difference between the peak position of the i-th echo data segment and the first echo data segment is (i-1)vQT. r After intra-segment coherent accumulation, the Doppler channel difference at the peak position between the i-th echo data segment and the first echo data segment is (i-1)aQT. r These two differences are used to achieve envelope shift and non-coherent accumulation between segments in subsequent steps.

[0096] 6.2 Utilize the velocity estimate obtained in step 3 Construct the phase compensation factor exp(j2πft′) i ),in This represents the time delay difference between the i-th signal segment and the first signal segment; the echo signal after coherent accumulation in step 5 is subjected to a fast time-dimension FFT, compensated in the distance-frequency domain, and then subjected to a fast time-dimension IFFT to obtain the shifted signal:

[0097]

[0098] At this point, the peak position of the signal within each segment has moved to... Next, inter-segment noncoherent accumulation can be performed to accumulate the energy of all echo signals, which will facilitate subsequent target detection.

[0099] Step 7: Normalize the M echo data segments after inter-segment envelope alignment to obtain M normalized echo data segments; using the inter-segment Doppler channel difference determined in Step 6, extract the data corresponding to the Doppler channel of each data segment in the M normalized echo data segments, take the modulus value, and then add them together to achieve inter-segment non-coherent accumulation, obtaining echo data after inter-segment non-coherent accumulation; target detection can then be performed on the echo data after inter-segment non-coherent accumulation.

[0100] 7.1, regarding the shifted signal obtained in step 6 Perform a normalization operation to obtain Initialize the Doppler channel number j = 1;

[0101] 7.2 Extract the data corresponding to the Doppler channel within each segment. Then, by taking the modulus and adding them together, we obtain the data of the j-th Doppler channel from the non-coherent accumulation, that is:

[0102]

[0103] Where d i，j =j+(i-1)aQT r This represents the Doppler channel corresponding to the i-th data segment;

[0104] 7.3 If j < Q, let j = j + 1 and return to step 7.2; otherwise, obtain the inter-segment non-coherent accumulation result y. out (t k , t l ), where t l =1, 2, ..., Q;

[0105] 7.4, regarding the accumulated result y out (t k , t lPerform constant false alarm rate (CFAR) detection to obtain target information.

[0106] Simulation Experiment

[0107] The effects of the present invention will be further illustrated below through simulation experiments.

[0108] 1. Simulation conditions

[0109] Assume the radar transmits a linear frequency modulated (LFM) signal with a carrier frequency of 1 GHz, a bandwidth of 1 MHz, a sampling rate of 2 MHz, a pulse width of 2 ms, a pulse repetition period of 8.4 ms, an accumulated pulse count of 256, and the target's initial velocity is 6800 m / s with an acceleration of 60 m / s². 2 The initial distance between the radar and the target is 150 km, and the signal-to-noise ratio is -20 dB.

[0110] 2. Simulation Content

[0111] Simulation 1: 256 pulses were divided into 8 groups of 32 pulses each. These 32 pulses were coherently accumulated, and then the coherent accumulation results of the 8 groups were non-coherently accumulated. During the coherent accumulation time, the first group of echo signals was simulated after pulse compression, and its contour plot is shown in Figure 2(a). A velocity search was performed using the proposed velocity search method and search interval. The optimal velocity value was used to correct the distance travel of the first group of echo signals, and the corrected contour plot is shown in Figure 2(b). Coherent accumulation was performed on the corrected first segment of echo signals, and the result is shown in Figure 3(a). An acceleration search was performed using the proposed acceleration search method and search interval. The optimal acceleration value was used to compensate for Doppler spread of the first segment of echo signals. The compensated results are coherently accumulated, and the accumulation result is shown in Figure 3(b). The optimal velocity and acceleration are used to perform distance travel correction and Doppler spread compensation on all segmented echo signals, and intra-segment coherent accumulation is performed. The peak positions of the first, fourth, and eighth segments are shown in Figure 4(a). The frequency domain envelope shift method is used to compensate for the inter-segment distance, and the result is shown in Figure 4(b). Finally, the signal is processed non-coherently between segments. To ensure the uniformity of the data scale after shifting, the data needs to be normalized before non-coherent accumulation, and the result is shown in Figure 4(b). Figure 5 ;

[0112] Simulation 2 compares the computational complexity of the proposed method and the traditional long-term coherent accumulation algorithm with the increase of the number of accumulation pulses. The results are as follows: Figure 6 .

[0113] 3. Simulation Result Analysis

[0114] As can be seen from Figures 2(a) and 2(b), although the number of accumulated pulses in each group is only 32, the distance movement in the echo signal appears due to the relatively high speed of the target. After speed compensation, the echo signals of different pulses are corrected to the same distance unit, and the distance movement between pulses is corrected.

[0115] As can be seen from Figures 3(a) and 3(b), due to the acceleration of the target, when coherently accumulating the echo directly, the Doppler spread of the range cell where the target is located is severe, and the energy of the echo signal cannot be effectively accumulated, resulting in energy loss, which is not conducive to subsequent target detection. Therefore, certain methods are needed to compensate for this to ensure the signal-to-noise ratio during target detection. After acceleration compensation, coherent accumulation is performed, and the energy of the echo signal is effectively accumulated within a Doppler cell, at which point inter-segment processing can continue.

[0116] As can be seen from Figures 4(a) and 4(b), after a series of processing of the signal within the segment, the energy of the target can be effectively accumulated, but the distance difference between segments still needs to be addressed; after frequency domain envelope shifting, the peak position of each segment is in the same distance unit, effectively compensating for the distance difference between segments.

[0117] from Figure 5 It can be seen that after inter-segment noncoherent accumulation, the target energy is effectively accumulated, which is beneficial to subsequent target detection and verifies the effectiveness of the proposed method.

[0118] from Figure 6 It can be seen that as the number of accumulated pulses increases, the number of searches for the traditional long-term coherent accumulation method increases rapidly, which means that the computational load of the algorithm increases dramatically. However, the method proposed in this invention has a slow growth trend, and the computational load changes very little, which is more conducive to engineering implementation and real-time processing.

[0119] Although the present invention has been described in detail in this specification with general description and specific embodiments, some modifications or improvements can be made to it, which will be obvious to those skilled in the art. Therefore, all such modifications or improvements made without departing from the spirit of the present invention are within the scope of protection claimed by the present invention.

Claims

1. A method for detecting high-speed targets in airspace based on long-term accumulation, characterized in that, Includes the following steps: Step 1: Establish a motion model for a high-speed maneuvering target, construct a radar echo data model under long-term observation conditions, and sequentially perform down-conversion and low-pass filtering on the raw echo data to obtain the baseband echo data received by the radar. ; Then, a matched filter is used to process the baseband echo data. Pulse compression was performed to obtain echo data in the frequency-slow time domain after pulse compression. ; in, , This represents the total number of accumulated pulses; To save time, Slow time; Step 2: Convert the pulse compression frequency-slow time domain echo data into a slow time dimension. Divided into Each echo data segment ; Among them, each Each pulse repetition interval is used as a coherent processing time period. an integer power of 2 , ; Step 3: Construct the traversal interval and step size for the search speed, based on each search speed. Construct the corresponding phase compensation factor to form the compensation matrix. and the first echo data Multiply the data to obtain the first segment of compensated echo data and estimate the target's velocity. Based on the target velocity estimate, form the optimal compensation matrix and compensate for each echo data segment separately to obtain the phase-compensated data. Each echo data segment will be phase-compensated. Each echo data segment is transformed into the distance-time domain using IFFT to obtain the phase-compensated time-domain signal. ; Step 4: Use the de-line frequency modulation method to estimate the Doppler frequency modulation of the time-domain signal, that is, estimate the target acceleration. Construct a Doppler compensation term based on the target acceleration estimate and compensate all phase-compensated time-domain signals to achieve Doppler extension compensation. Step 5, after compensation in step 4 Each range cell of each echo data segment is performed separately. The Fast Fourier Transform of the points focuses the energy within each echo data segment into a range cell and a Doppler cell, respectively, achieving coherent accumulation within the data segment and obtaining the coherently accumulated energy. One echo data segment; Step 6: Determine the range cell difference and Doppler channel difference between different echo data, construct a range compensation factor using the inter-segment range cell difference, and apply it to the coherently accumulated data. Range compensation is performed on each echo data segment to obtain the inter-segment envelope aligned data. One echo data segment; Step 7, after aligning the segment envelopes Each echo data segment is normalized to obtain the normalized data. Each echo data segment; using the inter-segment Doppler channel difference determined in step 6, the normalized data... The data corresponding to the Doppler channel of each echo data segment is extracted, the modulus value is taken, and then they are added together to achieve non-coherent accumulation between segments, thus obtaining the echo data after non-coherent accumulation between segments; target detection can then be performed on the echo data after non-coherent accumulation between segments.

2. The method for detecting high-speed targets in airspace based on long-term accumulation as described in claim 1, characterized in that, In step 1, the motion model of the high-speed maneuvering target is as follows: For a moving target, the distance between the target and the radar varies with slow time. Time-varying, can be represented as A polynomial function, which can be represented by a Taylor series expansion, is: in, For the target motion order, For the first The first-order motion parameters; if the first four terms of the Taylor series are retained, that is, taking... Then the above formula can be simplified to: In the formula, Indicates the first The distance between the target and the radar within each pulse repetition cycle Indicates the initial slant range of the target. For the target radial velocity, The target radial acceleration; The baseband echo data is: in, This represents the reflectance coefficient of the target. For rectangular window functions, The width is a rectangular pulse. For modulation bandwidth, To adjust the frequency, For carrier frequency, The imaginary unit, Let be the speed at which electromagnetic waves propagate in the air; let Indicates slow time, where m =0,1…, N -1, The pulse repetition interval, This represents the total number of accumulated pulses.

3. The method for detecting high-speed targets in airspace based on long-term accumulation as described in claim 1, characterized in that, In step 3, the estimation process of the target's velocity is as follows: 3.1, Set the maximum possible flight speed of the target to the maximum speed of the search traversal. ; 3.2 Determine the maximum search traversal interval for velocity based on radar system parameters. When performing a speed search, when the search speed relative to the target's true speed When they match, the following relationship holds: in, The search interval is the speed interval; When search speed relative to the target's true speed When matching, after using frequency domain phase compensation for distance travel correction, the distance between the peak value of the first echo pulse and the peak value of the last echo pulse cannot exceed one distance unit, i.e., satisfying the following relationship: in, This indicates the number of pulses in each segment after segmentation. At the speed of light, For signal bandwidth, The pulse repetition period; Because the speed found differs from the target's actual speed by a certain amount. After compensation, the first The distance traveled due to the residual velocity of the segment echo signal is: in, , which is a distance-resolved unit; To further ensure distance-based positioning between segments, the search interval needs to be reduced. If the speed is times higher, then the maximum search interval is set to... ； 3.3, Based on steps 3.1 and 3.2, the search traversal range for speed is determined as follows: ; 3.4, Based on search speed Constructing the phase compensation factor: in, ; Phase compensation factor For the first segment of echo data Perform frequency domain compensation, then perform an inverse fast Fourier transform along the fast time frequency to obtain the compensated result. The first segment of echo data in the domain: in, Represents the Sa function. The radar wavelength; Ignoring the distance traveled due to acceleration, then when When the search speed equals the target's true speed, the peak values ​​of the first echo signal envelope are located within the same distance cell, from which the target speed estimate is obtained. .

4. The method for detecting high-speed targets in airspace based on long-term accumulation according to claim 3, characterized in that, The time-domain signal after phase compensation is: in, , Indicates the first The time-domain signal after segment compensation, express The signal amplitude.

5. The method for detecting high-speed targets in airspace based on long-term accumulation according to claim 1, characterized in that, The process of estimating the Doppler modulation frequency of the time-domain signal using the de-line frequency modulation method is as follows: 4.1 The maximum possible flight acceleration of the target is determined as the maximum acceleration for the search traversal. ; 4.2 Determine the maximum search traversal interval for acceleration based on radar system parameters. When the acceleration found The actual acceleration of the target When they match, the following relationship holds: in, The acceleration search interval; After compensating the echo signal with the searched acceleration, the velocity change caused by the residual acceleration should not exceed one Doppler unit, i.e. Right now in, This indicates the number of pulses in each segment after segmentation. For wavelength, Let be the pulse repetition frequency; using the above search method to perform an acceleration search, the maximum difference between the searched acceleration and the target's true acceleration is . After compensation, the velocity change caused by the target's remaining acceleration is as follows: in, Indicates the first segment after segmentation One signal, Indicates the Doppler resolution unit; To avoid discrepancies in the Doppler cells of the target segment obtained after acceleration compensation, the search interval is reduced. times, of which Given the final number of segments, the maximum search interval for acceleration is: 4.3, Based on steps 4.1 and 4.2, the search traversal range for acceleration is determined as follows: ; 4.4, Based on search acceleration Get the search frequency Constructing compensation factors and the first echo data Frequency domain compensation is performed, and then the maximum energy peak is searched using the following formula to calculate the optimal tuning frequency: in, Indicates the modulo value. This represents the slow-time dimension signal corresponding to the distance cell where the target is located.

6. The method for detecting high-speed targets in airspace based on long-term accumulation according to claim 5, characterized in that, The process of constructing a Doppler compensation term based on the target acceleration estimate and compensating all phase-compensated time-domain signals is as follows: in, , Indicates the first The time-domain signal after phase compensation. express The signal amplitude.

7. The method for detecting high-speed targets in airspace based on long-term accumulation as described in claim 6, characterized in that, The specific process for determining the distance cell difference and Doppler channel difference between different echo data is as follows: from The expression shows that the first The distance cell difference between the peak positions of the first echo segment and the first echo segment is: ; After the intra-segment coherent accumulation, the first The Doppler channel difference at the peak position between the first echo segment and the first echo segment is .

8. The method for detecting high-speed targets in airspace based on long-term accumulation according to claim 1, characterized in that, The method employs the inter-segment distance unit difference to construct a distance-oriented compensation factor, which is then applied to the coherently accumulated data. Range compensation is performed on each echo data segment, specifically as follows: Using the velocity estimate obtained in step 3 Constructing phase compensation factor , where represents the first The time delay difference between the first segment signal and the second segment signal; the time delay difference after coherent accumulation in step 5. Perform a fast time-dimensional FFT on each echo data segment, compensate in the distance-frequency domain, and then perform a fast time-dimensional IFFT to obtain the echo data segments with inter-segment envelope alignment: in, Indicates the first The time-domain signal after phase compensation. , The speed at which electromagnetic waves propagate in the air; This is the pulse repetition interval.

9. The method for detecting high-speed targets in airspace based on long-term accumulation according to claim 1, characterized in that, The inter-segment noncoherent accumulation specifically refers to: 7.1, regarding the shifted signal obtained in step 6 Perform a normalization operation to obtain Initialize Doppler channel number ; 7.2 Extract the data corresponding to the Doppler channel within each segment. Then, by taking the modulus and adding them together, we obtain the first non-coherent accumulation. Data from each Doppler channel, namely: in Indicates the first The Doppler channel corresponding to the segment data; 7.3, if ,make Return to step 7.2; otherwise, obtain the inter-segment non-coherent accumulation result. ,in .

Images