Radar data technology system for monitoring tiny target

Abstract

The invention relates to the field of radar detection and signal processing, and particularly discloses a radar data technology system for monitoring a tiny target, and the system comprises a signal receiving and transmitting module which is used for generating and transmitting a linear frequency modulation signal based on a linear frequency modulation technology, and receiving an original echo signal reflected by the target; the pulse compression processing module is used for performing complex convolution operation on the original echo signal based on a matched filter to generate a compressed pulse signal; the multi-pulse coherent accumulation module is used for performing frequency domain energy accumulation on a plurality of compressed pulse signals of the same distance unit based on a discrete Fourier transform technology; and the Kalman filtering tracking module is used for processing the signals after multi-pulse coherent accumulation through a Kalman filtering algorithm, predicting and updating a target state in real time, and generating final trace point data. According to the invention, through key technologies of large-time wide-bandwidth pulse compression, multi-pulse coherent accumulation and the like, the detection capability of the radar on a tiny target is remarkably improved.

Description

Technical Field

[0001] The present application relates to the field of radar detection and signal processing, and more specifically, to a radar data technology system for monitoring tiny targets. Background Art

[0002] In many fields, such as aerospace, military defense, meteorological observation, traffic monitoring, and industrial inspection, the detection and monitoring of tiny targets are of extremely important significance. However, due to their small size, low radar cross-section (RCS), weak signals, and complex motion characteristics, traditional radar systems are difficult to effectively detect and track tiny targets.

[0003] Traditional radar systems mainly adopt continuous-wave radar or pulse radar technology. Although continuous-wave radar can measure the speed of targets, its detection ability for tiny targets is limited because it is difficult to distinguish the target echo from background noise, especially when the target is far away and the signal is weak, it is easily interfered and unable to accurately detect the target. And although traditional pulse radar has a certain detection range and target resolution ability, for tiny targets, the received echo signal is very weak after the transmitted pulse signal is reflected by the target and is easily submerged by noise, resulting in a decrease in detection probability. In addition, the pulse width of traditional pulse radar is relatively wide, which makes its range resolution not high enough to accurately distinguish adjacent tiny targets, and it is easy to have target overlap or misjudgment. Moreover, when traditional radar faces multi-target detection in a complex environment, it is easily interfered by clutter, further affecting the detection performance of tiny targets.

[0004] Therefore, a radar data technology system for monitoring tiny targets is provided. Summary of the Invention

[0005] To solve the above technical problems, the present application is proposed.

[0006] Specifically, according to one aspect of the present application, a radar data technology system for monitoring tiny targets is provided, which includes:

[0007] A signal transceiver module for generating and transmitting a linearly frequency-modulated signal based on linear frequency modulation technology and receiving the original echo signal reflected by the target;

[0008] A pulse compression processing module for performing complex convolution operation on the original echo signal based on a matched filter to generate a compressed pulse signal;

[0009] A multi-pulse coherent integration module for performing frequency-domain energy integration on multiple compressed pulse signals in the same range cell based on discrete Fourier transform technology;

[0010] Kalman filter tracking module, which is used to process the signal after multi-pulse coherent accumulation through the Kalman filter algorithm, predict and update the target state in real time, and generate the final track data.

[0011] Preferably, in the signal transceiver module, the linear frequency modulation signal is represented by the following formula:

[0012]

[0013] In the formula, A represents the amplitude; rect(t / τ) represents the rectangular function; τ represents the pulse width; f0 represents the transmit center frequency; μ = B / τ represents the frequency modulation slope, and B represents the bandwidth. The complex form of the linear frequency modulation signal x(t) is expressed as: Its spectrum is:

[0014]

[0015] Preferably, the pulse compression processing module defines the compression ratio D = Bτ. When D >> 1, the spectrum piecewise function of the original echo signal is:

[0016]

[0017] Among them, in the main lobe region where |f - f0| ≤ B / 2, it is a Gaussian function, and the main lobe width Δf ≈ 1 / B; in the sidelobe region where |f - f0| > B / 2, the spectrum value is zero;

[0018] Based on the conjugate symmetry of the spectrum X(f), a matched filter is designed: calculate the frequency response H(f) of the matched filter:

[0019]

[0020] Among them, is the normalization constant; φ(f) is the phase compensation term;

[0021]

[0022] In the formula, the group delay characteristic t d (f) of the compression filter is:

[0023]

[0024] Among them, is the additional delay;

[0025] The original echo signal is subjected to complex convolution operation through the matched filter to generate a compressed pulse signal, and the specific formula is expressed as:

[0026]

[0027] Take its real part:

[0028]

[0029] Preferably, the multi-pulse coherent accumulation module includes:

[0030] After orthogonal demodulation of the compressed pulse signal using an orthogonal phase detector, it is expressed in complex form: where a(t) represents the signal amplitude, f d represents the Doppler frequency, represents the initial phase; let Sampling x(t) at equal time intervals T s to discretize the continuous-time signal, obtaining a digital signal containing N range cells within a single detection period. The specific formula is expressed as: In the formula, ω d = 2πf d T s , representing the Doppler frequency of the digital signal, in radians; within M adjacent detection periods, sampling the echo signals of the same range cell to obtain a coherent pulse train signal x n (m). The specific formula is expressed as: x n (m) (0 ≤ n ≤ N - 1; 0 ≤ m ≤ M - 1), where m represents the pulse number and n represents the range cell index; performing an L ≥ M-point discrete Fourier transform on the pulse train signal x n (m) (0 ≤ n ≤ N - 1; 0 ≤ m ≤ M - 1) for each range cell respectively to obtain the signal after multi-pulse coherent accumulation. The specific formula is expressed as:

[0031]

[0032] Preferably, the Kalman filter tracking module includes: setting the initial state estimate X(1|1) ∈ R n×1 , covariance matrix P(1|1) ∈ R n×n , and establishing a system model; where the state equation X(k + 1) and measurement equation Z(k) of the system are expressed by the following formulas:

[0033] X(k + 1) = F(k)X(k) + G(k)W(k)

[0034] Z(k) = H(k)X(k) + V(k)

[0035] where X(k) ∈ R n×1 is the target state vector; Z(k) ∈ R m×1is the system measurement vector (the signal after multi-pulse coherent accumulation); W(k) ∈ R p×1 is the system noise; F(k) ∈ R n×n is the state transition matrix; G(k) ∈ R n×p is the system noise transfer matrix; H(k) ∈ R m×n is the system observation matrix; V(k) ∈ R m×1 is the system observation noise; Set k = 2 as the initial time, use the filter, and predict the state and covariance at the current time (k + 1) according to the state estimate and covariance at the previous time (k). The specific formula is expressed as:

[0036] X(k + 1|k) = FX(k|k)

[0037] P(k + 1|k) = FP(k|k)F T + GQ(k)G T ;

[0038] According to the measurement value Z(k + 1) ∈ R at the current time m×1 Calculate the Kalman gain, optimize the prediction result, and obtain the updated state estimate and covariance matrix. The specific formula is expressed as:

[0039] Innovation: y(k + 1) = Z(k + 1) - HX(k + 1|k)

[0040] Innovation covariance: S(k + 1) = HP(k + 1|k)H T + R

[0041] Kalman gain: K(k + 1) = P(k + 1|k)H T S -1 (k + 1)

[0042] Update state estimate: X(k + 1|k + 1) = X(k + 1|k) + K(k + 1)y(k + 1)

[0043] Update covariance matrix: P(k + 1|k + 1) = [I - K(k + 1)H]P(k + 1|k);

[0044] Pack the final state estimate and covariance matrix into structured data to obtain the trace data.

[0045] Compared with the prior art, a radar data technology system for monitoring small targets provided by the present application has the following remarkable effects:

[0046] 1), Through the linear frequency modulation signal generated by the linear frequency modulation technology, high-precision measurement and high-sensitivity detection of the target are realized, the detection ability and adaptability of the radar to small targets are improved, and it can better meet the detection requirements in different environments.

[0047] 2) Through the synergistic effect of large time-bandwidth pulse compression and multi-pulse coherent integration, effective detection and tracking of multiple targets in a complex environment are achieved, clutter interference is effectively suppressed, the anti-jamming ability of the radar in a complex environment is enhanced, and the detection performance for small targets is further improved.

[0048] 3) Through the Kalman filter tracking module, real-time and accurate estimation of the dynamic state of the target is achieved. It can track the position, velocity, and acceleration of the target in real time, generate stable track information, and improve the tracking accuracy and reliability of the radar for small targets. Description of the Drawings

[0049] By describing the embodiments of the present application in more detail in conjunction with the drawings, the above and other objects, features, and advantages of the present application will become more obvious. The drawings are used to provide a further understanding of the embodiments of the present application, and constitute a part of the specification. Together with the embodiments of the present application, they are used to explain the present application and do not constitute a limitation to the present application. In the drawings, the same reference numerals generally represent the same components or steps.

[0050] Figure 1 The system block diagram according to an embodiment of the present application is illustrated.

[0051] Figure 2 The linear frequency modulation signal and its amplitude and frequency modulation relationship diagram according to an embodiment of the present application are illustrated.

[0052] Figure 3 The pulse compression algorithm flowchart according to an embodiment of the present application is illustrated.

[0053] Figure 4 The output diagram of the matched filter for the linear frequency modulation signal according to an embodiment of the present application is illustrated.

[0054] Figure 5 The pulse integration algorithm flowchart according to an embodiment of the present application is illustrated.

[0055] Figure 6 The received signal M×N data matrix for M adjacent detection periods according to an embodiment of the present application is illustrated.

[0056] Figure 7 The target tracking algorithm flowchart according to an embodiment of the present application is illustrated.

[0057] Figure 8 The tracking processing software diagram according to an embodiment of the present application is illustrated.

[0058] Figure 9 The small target tracking display diagram according to an embodiment of the present application is illustrated. Detailed Embodiments

[0059] Next, embodiments according to the present application will be described in detail with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all embodiments of the present application. It should be understood that the present application is not limited by the exemplary embodiments described herein.

[0060] Embodiment:

[0061] Figure 1 A system block diagram according to an embodiment of the present application is illustrated. As Figure 1 shown, the radar data technology system 100 for monitoring small targets according to an embodiment of the present application includes: a signal transceiver module 110, configured to generate and transmit a chirp signal based on chirp technology and receive the original echo signal reflected by the target; a pulse compression processing module 120, configured to perform a complex convolution operation on the original echo signal based on a matched filter to generate a compressed pulse signal; a multi-pulse coherent integration module 130, configured to perform frequency-domain energy integration on multiple compressed pulse signals in the same range cell based on discrete Fourier transform technology; and a Kalman filter tracking module 140, configured to process the signal after multi-pulse coherent integration through a Kalman filter algorithm, predict and update the target state in real time, and generate the final track data.

[0062] In an embodiment of the present application, the signal transceiver module 110 is configured to generate and transmit a chirp signal based on chirp technology and receive the original echo signal reflected by the target. It should be understood that chirp (LFM) technology can significantly improve the performance of the radar system. The LFM signal obtained through LFM technology has the characteristics of high bandwidth and pulse compression, which can greatly improve the range resolution of the radar, enabling it to more accurately distinguish adjacent small targets and avoid target overlap or misjudgment. Therefore, in an embodiment of the present application, a chirp signal is generated and transmitted based on LFM technology, and then the original echo signal reflected by the target is received.

[0063] Specifically, the transmitting end generates an LFM signal through a voltage-controlled oscillator (VCO), and after power amplification, it is radiated through an antenna. The receiving end captures the original echo signal (i.e., the LFM signal reflected by the target) through a low-noise amplifier (LNA), and after conjugate matching with the transmitted signal, it is sent to the subsequent processing module.

[0064] Among them, Figure 2 A diagram of a chirp signal and its amplitude and frequency modulation relationship according to an embodiment of the present application is illustrated. Among them, (a) is the chirp signal, (b) is the signal envelope, and (c) is the modulation characteristic of the carrier frequency. As Figure 2 shown, based on the chirp (LFM) principle, a chirp signal with a center frequency of f0, a bandwidth of B, and an amplitude of A is transmitted, which is represented by the following formula:

[0065]

[0066] In the formula, represents the rectangular function; τ represents the pulse width; μ = B / τ represents the frequency modulation slope.

[0067] From Figure 2 it can be seen that within the pulse width τ, the angular frequency of the LFM signal changes from to

[0068] In particular, in a coherent radar, the complex form representation and spectrum analysis of the LFM signal are the core design bases. The complex form ensures the phase consistency of the signal during transmission, propagation, and reception by retaining the phase information (such as the quadratic phase term of linear frequency modulation), provides a mathematical basis for matched filtering and pulse compression, and thus improves the signal-to-noise ratio. Spectrum analysis can reveal the frequency characteristics of the signal (main lobe width, sidelobe attenuation), guide the selection of bandwidth and the optimization of the compression ratio, and at the same time support clutter suppression, Doppler frequency shift compensation, etc. Both together provide a theoretical support for high-resolution detection and stable tracking of small targets. Therefore, the complex form representation of the LFM signal x(t) is:

[0069]

[0070] Its spectrum is:

[0071]

[0072] In the embodiment of the present application, the pulse compression processing module 120 is configured to perform a complex convolution operation on the original echo signal based on a matched filter to generate a compressed pulse signal. It should be understood that pulse compression technology can compress a long pulse signal into a short pulse signal, thereby significantly reducing the pulse width. The smaller the pulse width, the higher the range resolution of the radar. For the detection of small targets, high range resolution is the key, which can more accurately distinguish adjacent small targets and avoid target overlap or misjudgment. Therefore, in order to better improve the performance of the radar in detecting small targets in a complex environment, in the embodiment of the present application, a complex convolution is performed on the original echo signal based on a matched filter to generate a compressed pulse signal.

[0073] Specifically, the pulse compression algorithm in the pulse compression processing module 120 is completed in the DSP, and its flow is as Figure 3 shown, specifically including:

[0074] 1>. Define the compression ratio D = Bτ (the compression ratio D represents the degree of pulse compression), and analyze the main lobe and sidelobe characteristics;

[0075] When D >> 1, the spectral piecewise function of the original echo signal is:

[0076]

[0077] Among them, in the main lobe region where |f - f0| ≤ B / 2, it is a Gaussian function, and the main lobe width Δf ≈ 1 / B; in the sidelobe region where |f - f0| > B / 2, the spectral value is zero, and there is no sidelobe interference ideally.

[0078] 2>. Based on the conjugate symmetry of the spectrum X(f), design a matched filter:

[0079] Calculate the frequency response H(f) of the matched filter:

[0080]

[0081] Among them, is a normalization constant used to adjust the filter gain; φ(f) is the phase compensation term;

[0082] It can be obtained that

[0083]

[0084] In the formula, the group delay characteristic (frequency - delay characteristic) t d (f) is

[0085]

[0086] Among them, is an additional delay related to the physical implementation of the filter.

[0087] 3>. Perform a complex convolution operation on the original echo signal through the matched filter to generate a compressed pulse signal, and the specific formula is expressed as:

[0088]

[0089] Since the actual signal is real, its real part is taken:

[0090]

[0091] Among them, the output signal waveform of the matched filter is as follows Figure 4 , and the signal envelope can be expressed as:

[0092]

[0093] From Figure 4 it can be seen that the output pulse envelope has the form of a sinc function. At -4dB of the output pulse amplitude, it is the output pulse width τ0, which is approximately equal to the reciprocal of the transmit signal spectral width B, that is The effective width τ0 of the output pulse is reduced by D times compared to the input pulse width τ, and the output pulse amplitude A0 is times, i.e.,

[0094] It is worth mentioning that from Figure 4 It can be seen that the ratio of the main lobe energy to the sidelobe energy of the output pulse is about 13.2 dB (the first sidelobe level). Since excessive sidelobes can cause false detection of adjacent targets or interference to real targets, measures are usually taken to reduce the sidelobes. The most common method is window function weighting (such as Hanning window, Hamming window, etc.), which can greatly reduce the sidelobe level. However, although window function weighting can effectively reduce the sidelobe level (usually suppressed to below -40 dB), it will be accompanied by main lobe broadening (the broadening amplitude is about 10%-20%) and signal energy loss. Therefore, it is necessary to balance and optimize between resolution and anti-interference performance.

[0095] In the embodiment of the present application, the multi-pulse coherent accumulation module 130 is used to perform frequency-domain energy accumulation on multiple compressed pulse signals in the same range cell based on the discrete Fourier transform technology. It should be understood that in a radar system, due to the limited echo energy of a single pulse, directly making a decision based on a single received pulse will result in a low signal-to-noise ratio and it is difficult to effectively detect weak and small target signals. In order to improve the signal-to-noise ratio and enhance the detection ability of small targets, a pulse accumulation processing technology is introduced to improve the signal-to-noise ratio. The accumulation is divided into coherent accumulation and non-coherent accumulation. In the embodiment of the present application, the coherent accumulation method is adopted to perform frequency-domain energy accumulation on multiple compressed pulse signals in the same range cell based on the discrete Fourier transform technology. In this way, the signal energies of multiple pulses can be superimposed when the phases are consistent, thereby maximizing the signal-to-noise ratio, enhancing the detectability of target signals, and effectively suppressing the interference of noise and clutter at the same time. This method not only improves the detection probability of small targets, but also enhances the anti-interference ability and target recognition accuracy of the radar system in a complex environment.

[0096] Specifically, as Figure 5 shown, the specific implementation process of the multi-pulse coherent accumulation module 130 is as follows:

[0097] 1), Use an orthogonal phase detector to demodulate the compressed pulse signal orthogonally and represent it in complex form: where a(t) represents the signal amplitude, f d represents the Doppler frequency, represents the initial phase;

[0098] 2), Let Sample x(t) at equal time intervals T s to discretize the continuous-time signal, and obtain a digital signal containing N range cells in a single detection period. The specific formula is expressed as:

[0099]

[0100] In the formula, ω d = 2πf d T s , representing the Doppler frequency of the digital signal, with the unit of radian (rad).

[0101] 3), constructing the digital signals with consistent Doppler frequencies in adjacent M detection periods into a coherent pulse train signal x n (m). That is, within M adjacent detection periods, the echo signals of the same range cell n are sampled to obtain the sequence: x n (m) (0 ≤ n ≤ N - 1; 0 ≤ m ≤ M - 1), where m represents the pulse sequence number (the m-th detection period), and n represents the range cell index (covering the entire detection area). If the detection periods T r of adjacent M detections are the same (or pulse group stagger, batch processing), and the carrier frequency f0 of the transmitted signal remains unchanged (or pulse group frequency modulation, batch processing), the M target echo signals of the same range cell are coherent pulse train signals with a length of M.

[0102] 4), arranging the coherent pulse train signals of all range cells into an M×N data matrix. Specifically, as Figure 6 shown, assuming that each detection period of the radar divides the whole processing process into N range cells starting from the transmitted synchronization pulse, the received signals of adjacent M detection periods form an M×N data matrix.

[0103] 5), performing an L≥M-point discrete Fourier transform (DFT) on the coherent pulse train signal x n (m) (0 ≤ m ≤ M - 1) of the n-th range cell and adjacent M detection periods to obtain the signal matrix after pulse accumulation. The specific formula is expressed as:

[0104]

[0105] In this way, the coherent accumulation of the pulse train signal x n (m) (0 ≤ m ≤ M - 1) of the n-th range cell and adjacent M detection periods is achieved. By performing an L≥M-point discrete Fourier transform on the pulse train signals x n (m) (0 ≤ n ≤ N - 1; 0 ≤ m ≤ M - 1) of each range cell in the whole processing process, the coherent accumulation of the pulse train signals of each range cell in the whole process is achieved.

[0106] In the embodiments of the present application, the Kalman filter tracking module 140 is used to process the signal after multi-pulse coherent accumulation through the Kalman filter algorithm, predict and update the target state in real time, and generate the final track data. It should be understood that the Kalman filter algorithm can estimate dynamic information such as the position, velocity, and acceleration of the target in real time and accurately through its prediction and update recursive mechanism, significantly improving the accuracy of target state estimation. It effectively suppresses the influence of measurement noise and system noise, and optimizes the estimation result of the target state by dynamically adjusting the Kalman gain. In addition, using the Kalman filter algorithm for target tracking can integrate discrete track data into continuous tracks, realizing stable tracking of the target. Even when the target signal is temporarily lost or interfered, the target tracking can be maintained through prediction. This not only improves the radar system's ability to track small targets but also enhances the system's adaptability and robustness in complex environments, providing strong support for the efficient operation of the radar system. Therefore, in the embodiments of the present application, the Kalman filter algorithm is further used to process the signal after multi-pulse coherent accumulation, track and predict the target state in real time to obtain the final track data.

[0107] Specifically, the Kalman filter based on a non-linear motion model is called the extended Kalman filter. In the embodiments of the present application, it is assumed that the system is a linear system.

[0108] Specifically, as Figure 7 shown, the specific implementation process of the Kalman filter tracking module 150 is as follows:

[0109] 1) Set the initial state estimate X(1|1) ∈ R n×1 and the covariance matrix P(1|1) ∈ R n×n , and establish the system model; where, the state equation X(k + 1) and the measurement equation Z(k) of the system are expressed by the following formulas:

[0110] X(k + 1) = F(k)X(k) + G(k)W(k)

[0111] Z(k) = H(k)X(k) + V(k)

[0112] where, X(k) ∈ R n×1 is the target state vector; Z(k) ∈ R m×1 is the system measurement vector (from the signal after multi-pulse coherent accumulation); W(k) ∈ R p×1 is the system noise; F(k) ∈ R n×n is the state transition matrix; G(k) ∈ R n×p is the system noise transition matrix; H(k) ∈ R m×n is the system observation matrix; V(k) ∈ Rm×1 is the system observation noise.

[0113] 2), Set k = 2 as the initial time. Use the filter to predict the state and covariance at the current time (k + 1) based on the state estimate and covariance at the previous time (k). The specific formulas are as follows:

[0114] X(k + 1|k) = FX(k|k)

[0115] P(k + 1|k) = FP(k|k)F T + GQ(k)G T .

[0116] 3), According to the measurement value Z(k + 1) ∈ R at the current time m×1 Calculate the Kalman gain, optimize the prediction result, and obtain the updated state estimate and covariance matrix. The specific calculation formulas are as follows:

[0117] Innovation: y(k + 1) = Z(k + 1) - HX(k + 1|k)

[0118] Innovation covariance: S(k + 1) = HP(k + 1|k)H T + R

[0119] Kalman gain: K(k + 1) = P(k + 1|k)H T S -1 (k + 1)

[0120] Updated state estimate: X(k + 1|k + 1) = X(k + 1|k) + K(k + 1)y(k + 1)

[0121] Updated covariance matrix: P(k + 1|k + 1) = [I - K(k + 1)H]P(k + 1|k).

[0122] 4), Package the final state estimate (position, velocity, acceleration) and covariance matrix into structured data to obtain the track data.

[0123] The system adopts an open architecture design, which is convenient for subsequent function expansion and upgrade. The core module integrates signal transceiver, pulse compression, multi-pulse coherent accumulation, and Kalman filter tracking technologies. Combined with the independently developed tracking processing software (as Figure 8 shown), it realizes high-precision and stable tracking of small targets, significantly improves the anti-noise ability and target state estimation accuracy in complex environments, and provides reliable technical support for military, ocean monitoring and other fields. In practical applications, the system can stably track small targets such as buoys, as Figure 9 shown. The buoy is within the red frame, and the system can stably track the buoy or other small targets.

[0124] In summary, the radar data technology system for monitoring tiny targets according to the embodiments of the present application is elucidated. It includes: a signal transceiver module for generating and transmitting a chirp signal based on chirp technology and receiving the original echo signal reflected by the target; a pulse compression processing module for performing complex convolution operations on the original echo signal based on a matched filter to generate a compressed pulse signal; a multi-pulse coherent integration module for performing frequency-domain energy integration on multiple compressed pulse signals in the same range cell based on discrete Fourier transform technology; a Kalman filter tracking module for processing the signal after multi-pulse coherent integration through the Kalman filter algorithm, predicting and updating the target state in real time, and generating the final track data. Through key technologies such as large time-bandwidth pulse compression and multi-pulse coherent integration, the present application significantly improves the detection ability of the radar for tiny targets.

[0125] For those skilled in the art, it is obvious that the present invention is not limited to the details of the above exemplary embodiments, and without departing from the spirit or basic characteristics of the present invention, the present invention can be implemented in other specific forms.

[0126] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit them. Although the present invention has been described in detail with reference to the preferred embodiments, those of ordinary skill in the art should understand that the technical solutions of the present invention can be modified or equivalently replaced without departing from the spirit of the technical solutions of the present invention.

Claims

1. A radar data technology system for monitoring tiny targets, characterized in that, Including: A signal transceiver module, configured to generate and transmit a chirp signal based on chirp technology and receive the original echo signal reflected by a target; A pulse compression processing module, configured to perform a complex convolution operation on the original echo signal based on a matched filter to generate a compressed pulse signal; A multi-pulse coherent integration module, configured to perform frequency-domain energy integration on multiple compressed pulse signals in the same range cell based on discrete Fourier transform technology; A Kalman filter tracking module, configured to process the signal after multi-pulse coherent integration through a Kalman filter algorithm, predict and update the target state in real time, and generate final track data.

2. The radar data technology system for monitoring tiny targets according to claim 1, characterized in that, In the signal transceiver module, the chirp signal is represented by the following formula: where, A represents amplitude; rect(t / τ) represents a rectangular function; τ represents the pulse width; f0 represents the transmit center frequency; μ = B / τ represents the chirp rate, and B represents the bandwidth.

3. The radar data technology system for monitoring minute targets according to claim 2, wherein In the signal transceiver module, the complex form of the chirp signal x(t) is represented as: Its spectrum is:

4. The radar data technology system for monitoring tiny targets according to claim 3, characterized in that, The pulse compression processing module includes: Define the compression ratio D = Bτ. When D >> 1, the spectral piecewise function of the original echo signal is: where, in the main lobe region |f - f0| ≤ B / 2, it is a Gaussian function, and the main lobe width Δf ≈ 1 / B; in the sidelobe region |f - f0| > B / 2, the spectral value is zero; Design a matched filter based on the conjugate symmetry of the spectrum X(f); Perform a complex convolution operation on the original echo signal through the matched filter to generate a compressed pulse signal.

5. The radar data technology system for monitoring tiny targets according to claim 4, wherein Design a matched filter based on the conjugate symmetry of the spectrum X(f), including: Calculate the frequency response H(f) of the matched filter: wherein, is a normalization constant; φ(f) is a phase compensation term; It can be obtained that: In the formula, the group delay characteristic t d (f) is as follows: where t d0 is the additional delay time.

6. The radar data technology system for monitoring tiny targets according to claim 5, characterized in that, Perform a complex convolution operation on the original echo signal through the matched filter to generate a compressed pulse signal. The specific formula is represented as: Take its real part:

7. The radar data technology system for monitoring tiny targets according to claim 6, wherein The multi-pulse coherent integration module includes: After orthogonally demodulating the compressed pulse signal using an orthogonal phase detector, it is expressed in complex form: where a(t) represents the signal amplitude, f d represents the Doppler frequency, represents the initial phase; Let Sample \(x(t)\) at equal time intervals \(T\) s to discretize the continuous-time signal and obtain a digital signal containing \(N\) range cells within a single detection period. The specific formula is expressed as: where \(\omega\) d \(= 2\pi f\) d \(T\) s represents the Doppler frequency of the digital signal, in radians; Within M adjacent detection periods, the echo signals of the same range cell are sampled to obtain the coherent pulse train signal x n (m), which is specifically expressed by the formula: x n (m) (0 ≤ n ≤ N - 1; 0 ≤ m ≤ M - 1), where m represents the pulse number and n represents the range cell index; Perform discrete Fourier transform of length L≥M on the burst signal x n (m) (0 ≤ n ≤ N - 1; 0 ≤ m ≤ M - 1) for each range cell to obtain the signal after multi-pulse coherent integration. The specific formula is as follows:​ 8. The radar data technology system for monitoring tiny targets according to claim 7, characterized in that, The Kalman filter tracking module includes: Set the initial state estimate of the system \(X(1|1)\in\mathbb{R}\) n×1 and the covariance matrix \(P(1|1)\in\mathbb{R}\) n×n , and establish the system model; where the state equation \(X(k + 1)\) and the measurement equation \(Z(k)\) of the system are expressed by the following formulas: X(k + 1) = F(k)X(k) + G(k)W(k) Z(k) = H(k)X(k) + V(k) where X(k) ∈ R n×1 is the target state vector; Z(k) ∈ R m×1 is the system measurement vector (from the signal after multi-pulse coherent integration); W(k) ∈ R p×1 is the system noise; F(k) ∈ R n×n is the state transition matrix; G(k) ∈ R n×p is the system noise transition matrix; H(k) ∈ R m×n is the system observation matrix; V(k) ∈ R m×1 is the system observation noise; Set k = 2 as the initial time. Use the filter to predict the state and covariance at the current time (k + 1) according to the state estimate and covariance at the previous time (k). The specific formula is represented as: X(k + 1|k) = FX(k|k) P(k + 1|k) = FP(k|k)F T + GQ(k)G T ; Based on the measurement value \(Z(k + 1)\in\mathbb{R}\) at the current moment m×1 Calculate the Kalman gain, optimize the prediction result, and obtain the updated state estimate and covariance matrix; Pack the final state estimate and covariance matrix into structured data to obtain track data.

9. The radar data technology system for monitoring tiny targets according to claim 8, wherein, Based on the measurement value \(Z(k + 1)\in\mathbb{R}\) at the current time m×1 Calculate the Kalman gain, optimize the prediction result, and obtain the updated state estimate and covariance matrix, including: Innovation: y(k + 1) = Z(k + 1) - HX(k + 1|k) Innovation covariance: S(k + 1) = HP(k + 1|k)H T + R Kalman gain: K(k + 1) = P(k + 1|k)H T S -1 (k + 1) Update the state estimate: X(k + 1|k + 1) = X(k + 1|k) + K(k + 1)y(k + 1) Update the covariance matrix: P(k + 1|k + 1) = [I - K(k + 1)H]P(k + 1|k).

Images