Rotating radar detection method and system based on intelligent beam scheduling

Abstract

The application provides a rotating radar detection method and system based on intelligent beam scheduling, and relates to the technical field of radar detection. First, the rotating motion state description parameter of the current detection period of a radar platform and a preliminary radar echo data set are acquired, and a radar detection space dynamic division scheme for a subsequent detection period is generated accordingly. Then, the radar beam parameter scheduling instruction set is generated in combination with the radar detection space dynamic division scheme and the rotating motion state description parameter. The radar platform transmits and receives links are controlled based on the radar beam parameter scheduling instruction set to perform target detection and collect a subsequent echo data set. Finally, the relevant parameters and scheme for the next detection period are updated according to the echo data set and the rotating motion state description parameter. The application improves the pertinence, efficiency and adaptability of radar detection.

Description

Technical Field

[0001] This invention relates to the field of radar detection technology, and more specifically, to a rotating radar detection method and system based on intelligent beam scheduling. Background Technology

[0002] In the field of modern radar detection, rotating radar is widely used because it can cover a large detection airspace. However, traditional rotating radar detection methods have many limitations when dealing with complex and ever-changing detection environments.

[0003] On the one hand, traditional rotating radars typically employ fixed detection modes and beam parameters, making it impossible to dynamically adjust based on the radar platform's own rotational motion. In practical applications, the radar platform's rotational speed, acceleration, and other motion states constantly change, and fixed detection modes struggle to adapt to these changes, resulting in low detection efficiency and an inability to fully utilize radar resources.

[0004] On the other hand, the detection airspace is often complex and variable, potentially containing target areas of varying priorities, interference zones, and so on. Traditional radar detection methods do not dynamically divide the detection airspace appropriately, instead employing a uniform detection approach. This prevents the radar from concentrating resources for precise detection of key areas, while wasting significant energy and time on non-key areas. Furthermore, it cannot adjust detection strategies promptly in the face of interference, impacting the accuracy and reliability of detection. Summary of the Invention

[0005] In view of the aforementioned problems, and in conjunction with the first aspect of the present invention, embodiments of the present invention provide a rotating radar detection method based on intelligent beam scheduling, the method comprising: Acquire the rotational motion state description parameters of the radar platform during the current detection period and the preliminary radar echo data set corresponding to the current detection period; Based on the rotational motion state description parameters and the preliminary radar echo data set, a dynamic radar detection airspace partitioning scheme is generated for subsequent detection cycles. Based on the radar detection airspace dynamic division scheme and the rotation motion state description parameters, a set of radar beam parameter scheduling instructions is generated to control the operation of the radar platform in subsequent detection cycles. Based on the radar beam parameter scheduling instruction set, the transmission and reception links of the radar platform are controlled to perform target detection operations in the airspace defined by the radar detection airspace dynamic division scheme, and radar echo data sets corresponding to subsequent detection cycles are collected. Based on the radar echo data set and the rotational motion state description parameters, update the expected rotational motion state parameters of the radar platform at the start of the next detection cycle and the dynamic partitioning scheme of the radar detection airspace.

[0006] Furthermore, embodiments of the present invention also provide a rotating radar detection system based on intelligent beam scheduling, characterized in that it includes: A processor; a machine-readable storage medium for storing machine-executable instructions of the processor; wherein the processor is configured to execute the above-described rotating radar detection method based on intelligent beam scheduling by executing the machine-executable instructions.

[0007] In another aspect, embodiments of the present invention also provide a computer program product, the computer program product including machine-executable instructions, the machine-executable instructions being stored in a computer-readable storage medium, the processor of the rotating radar detection system based on intelligent beam scheduling reading the machine-executable instructions from the computer-readable storage medium, the processor executing the machine-executable instructions, causing the rotating radar detection system based on intelligent beam scheduling to perform the aforementioned rotating radar detection method based on intelligent beam scheduling.

[0008] Based on the above, by acquiring the radar platform's rotational motion state description parameters and preliminary radar echo data set during the current detection cycle, a dynamic radar detection airspace partitioning scheme for subsequent detection cycles is generated based on this data. This allows for flexible adjustment of the detection airspace according to actual conditions, concentrating limited radar resources on key areas, improving the targeting and efficiency of detection, and avoiding resource waste. A radar beam parameter scheduling command set is generated based on the dynamic radar detection airspace partitioning scheme and the rotational motion state description parameters, realizing intelligent radar beam scheduling. This enables the radar platform to precisely control the beam parameters of the transmit and receive links according to different detection airspaces and its own motion state during subsequent detection cycles, thereby more effectively capturing target information and enhancing the radar's detection capabilities. Target detection operations are executed based on the radar beam parameter scheduling command set, and radar echo data sets for subsequent detection cycles are collected. Then, the expected rotational motion state parameters at the start of the next detection cycle are updated in conjunction with the rotational motion state description parameters, along with the dynamic radar detection airspace partitioning scheme. This allows for continuous adaptation to changes in the radar platform's motion state and the detection environment, continuously optimizing the detection strategy and improving the accuracy, reliability, and adaptability of radar detection. Attached Figure Description

[0009] Figure 1 This is a schematic diagram of the execution flow of the rotating radar detection method based on intelligent beam scheduling provided in an embodiment of the present invention.

[0010] Figure 2This is a schematic diagram of exemplary hardware and software components of a rotating radar detection system based on intelligent beam scheduling provided in an embodiment of the present invention. Detailed Implementation

[0011] Figure 1 This is a flowchart illustrating a rotating radar detection method based on intelligent beam scheduling according to an embodiment of the present invention, which will be described in detail below.

[0012] Step S110: Obtain the rotational motion state description parameters of the radar platform in the current detection period and the preliminary radar echo data set corresponding to the current detection period.

[0013] In this embodiment, the radar platform involved is a vehicle-mounted rotating active phased array radar operating in the 60GHz band, employing digital beamforming technology to achieve flexible beam control. The radar platform's rotating mechanism can rotate continuously 360 degrees, with each detection cycle set to 10 seconds, during which a preliminary scan and data acquisition of the surrounding airspace is completed. In urban road environments, this radar needs to simultaneously detect multiple targets, including vehicles ahead, pedestrians on the roadside, and obstacles. To achieve intelligent beam scheduling, it is first necessary to accurately acquire the radar's own rotational motion state and preliminary echo data. The rotational motion state description parameters include the time-varying sequence of rotational angular velocity, azimuth angle, etc., while the preliminary radar echo data set contains target reflection signal information at different distances and azimuths. This data ensures that the radar can dynamically adjust its detection strategy according to the actual situation, improving the detection accuracy and efficiency of key targets.

[0014] For example, step S111: The radar platform integrates an inertial measurement unit and a rotary encoder to collect the original rotational angular velocity signal and the original rotational azimuth signal of the radar platform in the current detection cycle in real time.

[0015] The inertial measurement unit integrated into the radar platform employs a three-axis MEMS gyroscope with a sampling frequency set to 1000Hz. The output signal is an analog voltage signal, covering a range from -5 radians / second to +5 radians / second, corresponding to an output voltage range of 0 to 5V. The rotary encoder uses a 16-bit absolute photoelectric encoder, with its sampling frequency synchronized with the gyroscope. It outputs a parallel digital signal directly representing the current azimuth angle, with a resolution of 360 degrees divided by 65536, meaning each encoded value corresponds to an azimuth angle increment of approximately 0.0055 degrees. Within the current 10-second detection cycle, a total of 10,000 raw rotational angular velocity sampling points and 10,000 raw rotational azimuth angle sampling points were collected, forming two one-dimensional time series of length 10,000. These raw signals contain various noises from the radar platform's rotation process, such as high-frequency noise caused by mechanical vibration and pulse interference, which require subsequent processing to eliminate.

[0016] Step S112: Perform signal denoising filtering on the original rotational angular velocity signal and the original rotational azimuth signal to obtain the denoised rotational angular velocity signal sequence and the denoised rotational azimuth signal sequence.

[0017] The raw rotational angular velocity signal was processed using a Butterworth low-pass filter with a fourth-order filter and a cutoff frequency of 50Hz set according to the mechanical rotation characteristics of the radar platform. Filtering was achieved through convolution. Specifically, the raw signal was linearly convolved with the filter's unit impulse response. For each sampling point in the raw angular velocity signal sequence, high-frequency mechanical vibration noise was eliminated by weighted summation with the filter coefficients. The raw rotational azimuth signal was processed using median filtering with a window size of 5 sampling points. Impulse noise interference was suppressed by using a sliding window to take the median value. For example, for the 5 sampling values ​​within the window, they were sorted, and the median value was taken as the output value of the current window. After filtering, two denoised signal sequences of length 10000 were obtained, denoted as the angular velocity sequence ω(t) and the azimuth sequence θ(t), where t represents the sampling time, t=1, 2, ..., 10000. After denoising, the signal-to-noise ratio was significantly improved, more accurately reflecting the actual rotational state of the radar platform.

[0018] Step S113: Perform timestamp alignment and resampling on the denoised rotational angular velocity signal sequence and the denoised rotational azimuth signal sequence to generate a time series of radar platform rotational motion state parameters that are time-synchronized and have a uniform sampling rate.

[0019] The timestamps of the denoised rotational angular velocity signal sequence ω(t) and rotational azimuth signal sequence θ(t) are automatically generated by the acquisition system. However, due to a slight difference in the start-up time of the two sensors, synchronization errors may exist. First, the timestamps of the two sequences are precisely aligned using a linear interpolation method, with the system master clock as the reference, ensuring that the time stamp error of each sampling point is less than 1 microsecond. Then, resampling is performed, uniformly adjusting the sampling frequency to the system standard sampling rate of 200Hz, implemented using a linear interpolation algorithm. For two adjacent sampling points (t_i, x_i) and (t_{i+1}, x_{i+1}) in the original sequence, the sampling value x_new corresponding to the new sampling time t_new is obtained by calculating x_new = x_i + (x_{i+1} - x_i) × (t_new - t_i) / (t_{i+1} - t_i). After resampling, a synchronized time sequence of length 2000 is obtained, where each sampling point contains the corresponding angular velocity value, azimuth value, and unified timestamp. Time synchronization and consistent sampling rates ensured the accurate correspondence between angular velocity and azimuth data in subsequent data processing.

[0020] Step S114: Extract statistical feature vectors from the time series of the radar platform rotational motion state parameters that can characterize the overall rotational motion characteristics of the radar platform in the current detection period, and use them as components of the rotational motion state description parameters.

[0021] The time series of rotational motion state parameters contains angular velocity and azimuth data from 2000 sampling points. Extracted statistical features include: the mean μω, standard deviation σω, maximum value ωmax, minimum value ωmin, and mean μdω of the first-order difference sequence (representing the trend of angular acceleration); the initial value θstart, ending value θend, total change Δθ = θend - θstart, and mean rate of change μdθ = Δθ / 10 seconds of the azimuth sequence. These features are combined into a 1×8 feature vector Vmotion = [μω, σω, ωmax, ωmin, μdω, θstart, θend, μdθ], which serves as the core component describing the rotational motion state parameters. For example, if the mean of the angular velocity sequence is 0.5 radians / second, the standard deviation is 0.1 radians / second, the maximum value is 0.8 radians / second, and the minimum value is 0.2 radians / second, the mean of the first-order difference sequence is 0.02 radians / second², the initial azimuth angle is 0 degrees, the final azimuth angle is 360 degrees, the total change is 360 degrees, and the average rate of change is 36 degrees / second, then the eigenvector Vmotion is [0.5, 0.1, 0.8, 0.2, 0.02, 0, 360, 36]. This eigenvector reflects the rotational motion characteristics of the radar platform during the current detection period.

[0022] Step S115: Simultaneously, control the radar platform to perform uniform scanning detection of the entire airspace within the current detection period in a preset fixed beam parameter working mode, and receive the original radar echo simulation signal generated in this process.

[0023] While acquiring rotational motion parameters, the radar platform initiates its initial detection mode. At this time, the beam control module is set to fixed parameter mode: transmission frequency of 60 GHz, pulse repetition frequency of 5000 Hz, pulse width of 100 nanoseconds, beamwidth of 3 degrees, and scanning method of 360-degree mechanical rotation uniform scanning, with pulse transmission and reception occurring every 3 degrees at each azimuth angle. The radar transmitter feeds the radio frequency signal into the antenna array through a power amplifier. After propagating through space, the signal is reflected upon encountering a target, and the reflected signal is received by the antenna and sent to the receiver front end. In urban road environments, the radar needs to cover an area within 100 meters ahead; therefore, the detection range and range resolution are adjusted by controlling the pulse repetition frequency and pulse width. Within the current detection cycle, the radar transmits pulse signals at fixed time intervals and continuously receives echo signals. These raw echo analog signals contain information such as the target's distance, velocity, and azimuth, but require subsequent processing to extract useful target features.

[0024] Step S116: Perform intermediate frequency amplification and bandpass filtering on the original radar echo analog signal to obtain the intermediate frequency radar echo signal.

[0025] The original analog radar echo signal is first amplified by a low-noise amplifier. The amplification factor is dynamically adjusted according to the received signal strength; it increases when the signal is weak and decreases when the signal is strong to avoid signal distortion. The amplified signal is then fed into a bandpass filter. This filter has a center frequency of 3 GHz (radar intermediate frequency) and a bandwidth of 500 MHz. It is used to filter out out-of-band noise and interference, retaining the energy of the target echo signal's frequency band. The bandpass filter is implemented using an LC passive filter circuit, which has low insertion loss and good frequency selectivity. The processed intermediate frequency radar echo signal is an analog signal, and its amplitude and frequency vary with the target distance and velocity. For example, when the target is far away, the echo signal has a larger delay, corresponding to a lower intermediate frequency; when the target has radial velocity, the echo signal will experience a Doppler frequency shift, changing its frequency. Intermediate frequency amplification and bandpass filtering improve the signal-to-noise ratio of the echo signal.

[0026] Step S117: Perform orthogonal demodulation processing on the intermediate frequency radar echo signal to separate the in-phase branch baseband signal and the orthogonal branch baseband signal.

[0027] The intermediate frequency (IF) radar echo signal is mixed with two quadrature local oscillator (LOO) signals generated by a local oscillator. The LEO generates a 3GHz sine wave and a cosine wave, with the sine wave serving as the in-phase LO and the cosine wave as the quadrature LO, with a 90-degree phase difference. After mixing the IF echo signal with the in-phase LOO, a low-pass filter is applied to obtain the in-phase branch baseband signal. After mixing with the quadrature LOO, another low-pass filter is applied to obtain the quadrature branch baseband signal. The low-pass filter has a cutoff frequency of 250MHz to filter out high-frequency components generated during mixing, preserving the baseband signal. The in-phase and quadrature branch baseband signals constitute a complex baseband signal, where the in-phase component is the real part and the quadrature component is the imaginary part. This quadrature demodulation method can completely preserve the amplitude and phase information of the echo signal, which is beneficial for subsequent digital signal processing, such as pulse compression and Doppler analysis.

[0028] Step S118: Synchronously sample and quantize the baseband signal of the in-phase branch and the baseband signal of the orthogonal branch to generate the original digital radar echo data set corresponding to the current detection period.

[0029] The in-phase and quadrature branch baseband signals are each fed into a 12-bit analog-to-digital converter (ADC). The sampling frequency is set to 1 GHz to satisfy the Nyquist sampling theorem, ensuring distortion-free sampling. The ADC's quantization range is -1V to +1V, converting the analog signal into a 12-bit digital signal. The quantization value of each sampling point is between 0 and 4095. The sampled digital signals are arranged in chronological order to form the original digital radar echo data set. This original digital radar echo data set is a two-dimensional array. One dimension represents the fast time (range) dimension, corresponding to the number of sampling points per pulse, and the other dimension represents the slow time (azimuth / pulse) dimension, corresponding to different transmitted pulses. For example, if the number of sampling points per pulse is 2000 and the number of pulses in the detection period is 50000, then the size of the original digital radar echo data set is 50000 × 2000. Synchronous sampling ensures the time consistency of the in-phase and quadrature branch signals, while quantization converts the analog signal into a digital signal, facilitating subsequent digital signal processing algorithms.

[0030] Step S119: Perform range-direction pulse matched filtering on the original digital radar echo data set to improve the range resolution of the radar echo data and obtain the pre-processed radar echo data set.

[0031] Each pulse echo signal in the original digital radar echo dataset undergoes pulse matched filtering in the range direction. The impulse response of the matched filter is the conjugate inversion of the transmitted pulse. Compression of the echo signal is achieved by convolving the echo signal with the impulse response of the matched filter. For each slow-time sampling point (i.e., each pulse) in the original digital radar echo dataset, convolution is performed in the fast-time dimension. Specifically, for each sampling point in the fast-time dimension, it is weighted and summed with the coefficients of the matched filter to obtain the compressed signal. Pulse matched filtering can compress wide pulses into narrow pulses, thereby improving the radar's range resolution. For example, compressing a 100-nanosecond pulse can achieve a range resolution of approximately 15 meters. After pulse matched filtering, a preliminary processed radar echo dataset is obtained, where each element represents the target echo intensity at the corresponding range and azimuth.

[0032] Step S1110: Associate and encapsulate the pre-processed radar echo data set with the rotational motion state description parameters to form the preliminary radar echo data set and its matching rotational motion state description parameter data package corresponding to the current detection period.

[0033] The pre-processed radar echo data set and rotational motion state description parameters need to be associated and encapsulated for use in subsequent steps. Specifically, a corresponding rotational motion state parameter, including the angular velocity and azimuth angle at that sampling moment, is added to each slow-time sampling point in the pre-processed radar echo data set. The two are precisely associated using timestamps to ensure that each echo data sampling point corresponds to an accurate radar rotational state. The associated and encapsulated data packet uses a specific data format, consisting of a header and a body. The header contains basic information about the data packet, such as the detection period number, sampling time, and data size; the body contains the pre-processed radar echo data and the corresponding rotational motion state description parameters. This rotational motion state description parameter data packet is transmitted to the subsequent airspace dynamic partitioning module via an internal bus.

[0034] Step S120: Based on the rotational motion state description parameters and the preliminary radar echo data set, generate a dynamic radar detection airspace partitioning scheme for subsequent detection cycles.

[0035] After obtaining the rotational motion state description parameters and preliminary radar echo data set for the current detection cycle, a dynamic partitioning scheme for the radar detection airspace needs to be generated for subsequent detection cycles. This scheme divides the airspace into different regions based on the radar's rotational motion characteristics and the initially detected target information. By analyzing the rotational motion state description parameters, the radar's rotation trajectory and coverage area in subsequent detection cycles can be predicted; by processing the preliminary radar echo data set, the location and threat level of potential targets can be extracted. Integrating these two pieces of information allows for the identification of airspace areas requiring focused attention and the rational allocation of radar resources, thereby improving detection efficiency and target detection probability.

[0036] Step S121: Extract the real-time rotational angular velocity sequence and the real-time rotational azimuth sequence of the radar platform in the current detection period from the rotational motion state description parameters.

[0037] The rotational motion state description parameters encompass various motion characteristics of the radar platform within the current detection period. Extracting the real-time rotational angular velocity and azimuth sequences from these parameters is fundamental for subsequent processing. Specifically, the angular velocity and azimuth values ​​for each sampling point are extracted from the rotational motion state parameter time series, forming two one-dimensional sequences. For example, if the rotational motion state parameter time series contains 2000 sampling points, both the real-time rotational angular velocity and real-time rotational azimuth sequences are one-dimensional arrays of length 2000. These sequences reflect the changes in the radar platform's rotational angular velocity and azimuth over time within the current detection period. During the analysis process, it is crucial to ensure the accuracy and completeness of the data, eliminating potential outliers, such as angular velocities or azimuth values ​​exceeding the normal range.

[0038] Step S122: Based on the real-time rotational angular velocity sequence and the real-time rotational azimuth sequence, calculate the predicted rotational angular velocity value of the radar platform at the start of the subsequent detection cycle and the predicted rotational azimuth value of the radar platform at the start of the subsequent detection cycle.

[0039] Based on real-time rotational angular velocity and rotational azimuth sequences, a time-series prediction method is used to calculate the predicted rotational angular velocity and azimuth values ​​at the start of subsequent detection cycles. First, trend analysis is performed on the real-time rotational angular velocity sequence, and a trend line of angular velocity variation over time is fitted using the least squares method. The predicted rotational angular velocity value at the start of subsequent detection cycles is then extrapolated from this trend line. Similarly, the real-time rotational azimuth sequence is fitted, and the predicted rotational azimuth value is extrapolated. For example, if the trend line of the angular velocity sequence is ω(t) = a × t + b, where a and b are fitting coefficients, then the predicted angular velocity value corresponding to the start time t0 of the subsequent detection cycle is ω(t0) = a × t0 + b. During the fitting process, an appropriate fitting order needs to be selected to accurately reflect the changing trends of angular velocity and azimuth. The prediction results can provide information on the rotational state of the radar platform at the start of subsequent detection cycles for subsequent airspace division, ensuring that the airspace division matches the actual movement of the radar.

[0040] Step S123: Perform moving target feature extraction processing on the preliminary radar echo data set to obtain the azimuth distribution features and radial velocity distribution features of potential moving targets contained in the preliminary radar echo data set.

[0041] The initial radar echo dataset contains a large amount of target echo information, and it is necessary to extract the azimuth and radial velocity distribution characteristics of potential moving targets from it. First, constant false alarm rate (CFAR) detection is performed on the initial radar echo dataset to extract possible target traces. Then, cluster analysis is performed on these target traces to merge traces belonging to the same target. For each target trace cluster, its azimuth statistical distribution characteristics, such as mean and variance, are calculated to obtain the potential moving target azimuth distribution characteristics; simultaneously, its radial velocity statistical distribution characteristics are calculated to obtain the potential moving target radial velocity distribution characteristics. These potential moving target radial velocity distribution characteristics reflect the distribution and motion characteristics of potential moving targets in the airspace.

[0042] Step S1231: Perform range gate cell partitioning on the preliminary radar echo data set to obtain multiple radar echo range gate data segments arranged according to the range dimension.

[0043] The initial radar echo dataset has high range resolution, and to facilitate processing, it needs to be divided into multiple range gate units. The division of range gate units is based on the radar's range resolution, with each range gate unit corresponding to a specific range range. For example, if the range resolution is 15 meters, then each range gate unit is 15 meters wide. Dividing the initial radar echo dataset along the range direction into range gate units yields multiple range-gate data segments. Each range-gate data segment contains all echo signal sampling points within that range. Range gate unit division allows for the separation of target echo signals at different ranges, facilitating subsequent Doppler spectrum analysis and target detection.

[0044] Step S1232: Perform Doppler spectrum analysis on each radar echo range gate data segment to obtain the Doppler power spectral density distribution curve corresponding to each radar echo range gate data segment.

[0045] For each radar echo range gate data segment, a Fast Fourier Transform (FFT) is used for Doppler spectrum analysis. The range gate data segment is FFT-transformed in the slow-time dimension to obtain the Doppler power spectral density distribution curve corresponding to that range gate. The number of FFT points is determined based on the number of slow-time sampling points; for example, if each range gate data segment contains 512 slow-time sampling points, a 512-point FFT is used. The Doppler power spectral density distribution curve reflects the echo power distribution at different Doppler frequencies. By analyzing this curve, the radial velocity information of the target can be extracted. This is because there is a linear relationship between the target's radial velocity and the Doppler frequency, i.e., v = λ × f_d / 2, where λ is the radar wavelength and f_d is the Doppler frequency.

[0046] Step S1233: Detect significant spectral peaks exceeding the noise floor threshold from the Doppler power spectral density distribution curve corresponding to each radar echo range gate data segment, and record the Doppler frequency value and power amplitude value corresponding to each significant spectral peak.

[0047] After obtaining the Doppler power spectral density distribution curve, it is necessary to detect significant spectral peaks. First, a noise floor threshold is calculated, typically using the mean of the power spectral density distribution curve plus a certain multiple of the standard deviation as the threshold. Then, the power spectral density distribution curve is iterated through to identify all peaks exceeding this threshold. For each peak, its corresponding Doppler frequency and power amplitude are recorded. Significant spectral peaks correspond to potential moving targets, as moving targets generate strong Doppler frequency shift signals. By detecting these peaks, the presence of potential targets and their Doppler frequency information can be preliminarily determined.

[0048] Step S1234: Based on the Doppler frequency value corresponding to each significant spectral peak point and combined with the radar transmitted signal carrier wavelength parameter, calculate the estimated potential target radial velocity value corresponding to each significant spectral peak point.

[0049] Given that the carrier frequency of the radar transmitted signal is 60 GHz, the radar wavelength λ can be calculated using the relationship between wavelength and frequency λ = c / f, where c is the speed of light and f is the carrier frequency. For each significant spectral peak corresponding to the Doppler frequency value f_d, the estimated radial velocity v of the potential target is calculated using the formula v = λ × f_d / 2. For example, if the Doppler frequency is 1000 Hz and the radar wavelength is 0.005 meters, then the radial velocity v = 0.005 × 1000 / 2 = 2.5 m / s. This calculation converts the Doppler frequency into the target's radial velocity, facilitating subsequent analysis of the target's motion characteristics.

[0050] Step S1235: Based on the range gate number corresponding to the radar echo range gate data segment to which each significant spectral peak point belongs, and in conjunction with the radar range resolution parameters, calculate the potential target range estimate corresponding to each significant spectral peak point.

[0051] Each radar echo range gate data segment corresponds to a specific range gate number, and there is a correspondence between the range gate number and the range. Given that the radar's range resolution is ΔR and the starting range is R0, the range range corresponding to the nth range gate is from R0 + (n-1) × ΔR to R0 + n × ΔR. The center distance of the range gate is taken as the range estimate of the potential target, i.e., R = R0 + (n-0.5) × ΔR. For example, if the range gate number is 10, the range resolution is 15 meters, and the starting range is 0 meters, then the range estimate R = 0 + (10 - 0.5) × 15 = 142.5 meters. Through this calculation, the range estimate of the potential target corresponding to each significant spectral peak point is obtained.

[0052] Step S1236: Using the real-time rotating azimuth sequence, determine the instantaneous beam pointing azimuth of the radar platform corresponding to each significant spectral peak point through interpolation matching operation, and use the instantaneous beam pointing azimuth of the radar platform as the azimuth estimate of the corresponding potential target.

[0053] The real-time rotation azimuth angle sequence records the azimuth angle values of the radar platform at different times. Each significant spectral peak point corresponds to a specific slow-time sampling moment. Interpolation is performed in the real-time rotation azimuth angle sequence according to this moment to obtain the instantaneous beam pointing azimuth angle of the radar platform at this moment. For example, if the slow-time sampling moment corresponding to the significant spectral peak point is t, the azimuth angles at times t1 and t2 in the real-time rotation azimuth angle sequence are θ1 and θ2 respectively, and t1 < t < t2, then θ = θ1 + (θ2 - θ1) × (t - t1) / (t2 - t1) is calculated by linear interpolation to obtain the azimuth angle estimate value corresponding to this peak point. This azimuth angle estimate value is used as the azimuth angle of the corresponding potential target, thereby determining the azimuth position of the target in the airspace.

[0054] Step S1237: Aggregate the estimated radial velocities, estimated distances, and estimated azimuth angles of potential targets corresponding to all significant spectral peak points to form an original set of potential target traces.

[0055] The estimated radial velocity, distance, and azimuth angle of the potential target corresponding to each significant spectral peak point are combined into a trace. All these traces are aggregated to form an original set of potential target traces. The original set of potential target traces is a list containing multiple traces, and each trace contains three parameters: radial velocity, distance, and azimuth angle. For example, a trace can be represented as (v1, R1, θ1), where v1 is the radial velocity, R1 is the distance, and θ1 is the azimuth angle. The original set of potential target traces contains all possible potential target information, but there may be false traces and noise interference, which need to be further processed.

[0056] Step S1238: Perform trace clustering processing on the original set of potential target traces in the three-dimensional space of distance-azimuth-velocity, and merge potential target traces with close spatial positions and similar velocity characteristics into the same potential target trace cluster.

[0057] The density-based clustering algorithm (such as DBSCAN) is used to perform clustering processing on the original set of potential target traces. In the three-dimensional space of distance, azimuth angle, and velocity, appropriate neighborhood radius and minimum number of points thresholds are set, and traces with close distances, similar azimuth angles, and similar velocities are aggregated into a cluster. For example, the neighborhood radius is set to 5 meters in the distance dimension, 2 degrees in the azimuth angle dimension, 1 m / s in the velocity dimension, and the minimum number of points threshold is set to 3. Through clustering processing, traces belonging to the same target are merged into a potential target trace cluster, reducing the influence of false traces and improving the accuracy of target detection.

[0058] Step S1239: Extract the statistical distribution center of the azimuth angle estimates of all points in each potential target point cluster as the comprehensive azimuth angle feature of the potential target point cluster, and extract the statistical distribution center of the radial velocity estimates of all points in the potential target point cluster as the comprehensive radial velocity feature of the potential target point cluster.

[0059] For each potential target point cluster, the mean of the azimuth estimates of all points within the cluster is calculated as the comprehensive azimuth feature, and the mean of the radial velocity estimates is calculated as the comprehensive radial velocity feature. For example, if a point cluster contains three points with azimuth estimates of 30 degrees, 32 degrees, and 31 degrees, the comprehensive azimuth feature is (30+32+31) / 3 = 31 degrees; and the radial velocity estimates are 2 m / s, 3 m / s, and 2.5 m / s, respectively, the comprehensive radial velocity feature is (2+3+2.5) / 3 = 2.5 m / s. The comprehensive azimuth and comprehensive radial velocity features can represent the overall azimuth and velocity characteristics of the potential target point cluster.

[0060] Step S12310: Integrate the comprehensive azimuth features and comprehensive radial velocity features of all potential target point clusters to generate an azimuth feature vector characterizing the azimuth distribution characteristics of potential moving targets contained in the preliminary radar echo data set and a radial velocity feature vector characterizing the radial velocity distribution characteristics of potential moving targets contained in the preliminary radar echo data set.

[0061] The combined azimuth features of all potential target clusters are combined into one vector, namely the azimuth feature vector; the combined radial velocity features are combined into another vector, namely the radial velocity feature vector. For example, if there are 5 potential target clusters with combined azimuth features of 30 degrees, 60 degrees, 90 degrees, 120 degrees, and 150 degrees, then the azimuth feature vector is [30, 60, 90, 120, 150]; and the combined radial velocity features are 2 m / s, 3 m / s, -1 m / s, 4 m / s, and -2 m / s, then the radial velocity feature vector is [2, 3, -1, 4, -2]. These two feature vectors comprehensively reflect the azimuth and radial velocity distribution of potential moving targets in the preliminary radar echo data set.

[0062] Step S124: Based on the azimuth distribution characteristics and radial velocity distribution characteristics of the potential moving targets, construct a set of potential target threat airspace that needs to be focused on in subsequent detection cycles.

[0063] The azimuth and radial velocity distribution characteristics of potential moving targets reflect their position and motion characteristics. Based on these characteristics, the threat level of the targets can be assessed, and a set of potential target threat airspaces requiring focused attention can be constructed. For each cluster of potential target points, based on its comprehensive azimuth and radial velocity characteristics, its trajectory and possible airspace range in subsequent detection cycles are predicted. Combined with threat assessment rules, its threat level is determined, thereby delineating the airspace areas of focus.

[0064] Step S1241: Based on the sign and absolute value of each radial velocity feature component in the radial velocity distribution characteristics of the potential moving target, determine the radial motion tendency attribute and radial motion intensity attribute of the corresponding potential target point cluster relative to the radar platform.

[0065] The sign of the radial velocity characteristic component indicates the target's direction of motion relative to the radar platform; a positive sign indicates the target is moving away from the radar, and a negative sign indicates the target is moving closer to the radar. The absolute value indicates the magnitude of the radial motion; the larger the absolute value, the more intense the radial motion. For example, a radial velocity characteristic component of 3 m / s indicates the target is moving away from the radar with moderate radial motion intensity; a radial velocity characteristic component of -5 m / s indicates the target is moving closer to the radar with high radial motion intensity. By determining the radial motion tendency and intensity attributes, the target's threat potential can be preliminarily assessed. Targets moving closer to the radar with intense motion generally have a higher threat level.

[0066] Step S1242: Based on the value of each azimuth feature component in the azimuth feature vector of the potential moving target azimuth distribution characteristics, and combined with the statistical trend of the real-time rotating azimuth sequence, calculate the azimuth change rate parameter of the corresponding potential target point cluster relative to the rotation center of the radar platform.

[0067] Each component in the azimuth feature vector represents the composite azimuth of a cluster of potential target points. By combining the statistical trends of the real-time rotational azimuth sequence, such as the mean rotational angular velocity, the rate of change of the target's azimuth relative to the radar platform's rotation center is calculated. For example, if the target's composite azimuth is θ_target and the mean rotational angular velocity of the radar platform is ω_avg, then the azimuth rate of change parameter is dθ / dt = ω_avg - dθ_target / dt, where dθ_target / dt is the rate of change of azimuth caused by the target's own motion. By calculating the azimuth rate of change parameter, the target's azimuth change in subsequent detection cycles can be predicted, determining whether the target will enter the radar's blind zone or critical area.

[0068] Step S1243: Integrate the radial motion tendency attribute of the corresponding potential target cluster relative to the radar platform, the radial motion intensity attribute of the corresponding potential target cluster relative to the radar platform, and the azimuth angle change rate parameter of the corresponding potential target cluster relative to the rotation center of the radar platform, and calculate the real-time threat level score of each potential target cluster using a preset threat assessment rule base.

[0069] The pre-built threat assessment rule base includes multiple assessment indicators and corresponding weights, such as radial motion tendency weight, radial motion intensity weight, and azimuth rate of change weight. The radial motion tendency attribute, radial motion intensity attribute, and azimuth rate of change parameter are quantified into scores, and then weighted and summed according to their weights to obtain the real-time threat level score for each potential target cluster. For example, if the radial motion tendency is close to radar (score 80), the radial motion intensity is high (score 90), and the azimuth rate of change is fast (score 85), with weights of 0.4, 0.3, and 0.3 respectively, then the real-time threat level score is 80×0.4 + 90×0.3 + 85×0.3 = 84.5. The higher the threat level score, the greater the threat level of the target.

[0070] Step S1244: Sort and filter all potential target point clusters according to the real-time threat level score of each potential target point cluster, and select potential target point clusters whose real-time threat level score exceeds the preset threat level threshold to form a potential target point cluster subset.

[0071] The preset threat level threshold is set according to the actual application scenario, for example, 70 points. All potential target clusters are sorted from high to low according to their real-time threat level scores, and clusters with scores above 70 points are selected to form a subset of potential target clusters. This subset of potential target clusters contains potential targets with a high threat level, which need to be focused on in subsequent detection cycles. By sorting and filtering, radar resources can be concentrated on detecting high-threat targets, improving detection efficiency and reliability.

[0072] Step S1245: For each potential target point cluster in the subset of potential target point clusters, based on the comprehensive azimuth and comprehensive radial velocity characteristics of the potential target point cluster, predict the azimuth range and distance range that the potential target point cluster may appear in subsequent detection cycles.

[0073] Based on the comprehensive azimuth and radial velocity characteristics of potential target clusters, combined with the radar platform's rotation prediction information, the azimuth and range intervals of the target in subsequent detection cycles are predicted. For example, if the current comprehensive azimuth is θ0, the comprehensive radial velocity is v0, the radar platform predicts the rotational angular velocity as ω_pred, and the duration of the subsequent detection cycle is T, then the azimuth interval range is [θ0 + (ω_pred - dθ / dt) × T - Δθ, θ0 + (ω_pred - dθ / dt) × T + Δθ], where Δθ is the prediction error range; the range interval range is [R0 + v0 × T - ΔR, R0 + v0 × T + ΔR], where R0 is the current range, and ΔR is the range prediction error range. By predicting the azimuth and range intervals, the possible airspace location of the target in subsequent detection cycles can be determined.

[0074] Step S1246: Based on the azimuth interval range and range range that each potential target point cluster may appear in subsequent detection cycles, define a sector-shaped airspace description block constrained by the azimuth interval boundary and the range interval boundary in the radar polar coordinate system.

[0075] In the radar polar coordinate system, the radar platform is the origin, the azimuth is the polar angle, and the range is the polar radius. Based on the predicted azimuth interval [θ_start, θ_end] and range interval [R_start, R_end], a sector-shaped airspace description block is defined. The boundary of this sector-shaped airspace description block is determined by θ_start, θ_end, R_start, and R_end, and it contains all possible airspace locations where the target may appear. The sector-shaped airspace description block accurately reflects the target's distribution range in the polar coordinate system.

[0076] Step S1247: For each sector spatial description block, attach a real-time threat level score of the source potential target point trace cluster corresponding to the sector spatial description block as the threat level identifier of the sector spatial description block.

[0077] The real-time threat level score of the source potential target point cluster corresponding to each sector of the airspace description block is used as the threat level identifier for that sector of the airspace description block. For example, if the real-time threat level score of the source potential target point cluster is 85, then the threat level identifier for that sector of the airspace description block is 85. The threat level identifier can intuitively reflect the importance of the airspace.

[0078] Step S1248: Summarize all sector-shaped airspace description blocks with added threat level identifiers to construct a structured database of potential target threat airspace sets that need to be focused on in subsequent detection cycles.

[0079] All sector-shaped airspace description blocks with attached threat level identifiers are stored in the database in descending order of threat level, forming a potential target threat airspace set database. This database contains information such as the azimuth range, distance range, and threat level identifier for each sector-shaped airspace description block. The structured database facilitates rapid querying and access, improving the efficiency of subsequent processing.

[0080] Step S125: Integrate the predicted rotational angular velocity value, the predicted rotational azimuth angle value, and the potential target threat airspace set to generate a set of candidate key detection airspaces that the radar beam should prioritize covering in subsequent detection cycles.

[0081] The predicted rotational angular velocity and predicted rotational azimuth reflect the radar platform's motion state during subsequent detection cycles, while the potential target threat airspace set includes airspace areas requiring priority attention. By fusing these two pieces of information, a set of candidate key detection airspace areas that the radar beam should prioritize covering during subsequent detection cycles is determined. For example, based on the predicted rotational trajectory, the airspace range that the radar can cover during subsequent detection cycles is determined, and then this range is intersected with the potential target threat airspace set to obtain the candidate key detection airspace set. These candidate airspace areas are the regions where the radar needs to prioritize resource allocation for detection to ensure effective tracking and detection of high-threat targets.

[0082] Step S126: Based on the predicted rotational angular velocity value and the predicted rotational azimuth angle value, calculate the theoretical maximum airspace boundary that the radar platform can cover by completing a single-cycle rotational scan in subsequent detection cycles.

[0083] Given the predicted rotational angular velocity ω_pred (radians / second) and the subsequent detection period duration T (seconds), the rotation angle of the radar platform within the subsequent detection period is Δθ_total = ω_pred × T (radians). The predicted rotational azimuth angle is θ_start_pred (radians). Therefore, the starting azimuth angle of the theoretical maximum airspace boundary is θ_start_pred, and the ending azimuth angle is θ_start_pred + Δθ_total. If the ending azimuth angle exceeds 360 degrees, 360 degrees are subtracted to ensure it remains within the 0-360 degree range. The theoretical maximum airspace boundary reflects the maximum airspace range that the radar can cover under ideal conditions.

[0084] Step S127: Based on the candidate key detection airspace set and the theoretical maximum airspace boundary, perform airspace resource allocation optimization processing to obtain the theoretical dwell time ratio parameter of each candidate key detection airspace allocated in the subsequent detection cycle.

[0085] Each airspace in the candidate key detection airspace set has a corresponding threat level identifier, and the theoretical maximum airspace boundary limits the airspace range that the radar can cover. The goal of airspace resource allocation optimization is to allocate a corresponding dwell time proportion to the candidate key detection airspaces within the theoretical maximum airspace boundary, based on their threat levels. A weighted allocation method is used, with airspaces with higher threat levels receiving a larger dwell time proportion. For example, if the threat level score of a candidate key detection airspace is s_i, and the total threat level score is S = Σs_i, then the theoretical dwell time proportion parameter for the i-th airspace is p_i = s_i / S. This method ensures that high-threat airspaces receive more radar resources, improving detection effectiveness.

[0086] Step S128: Based on the theoretical dwell time ratio parameter and the predicted rotation angular velocity value, reversely calculate the azimuth interval width parameter occupied by each candidate key detection airspace in the radar platform rotation coordinate system.

[0087] Given that the duration of the subsequent detection cycle is T and the theoretical dwell time ratio parameter is p_i, the dwell time of the i-th candidate key detection airspace is t_i = p_i × T. The predicted rotational angular velocity is ω_pred, and the azimuth interval width parameter Δθ_i = ω_pred × t_i. For example, if the theoretical dwell time ratio parameter is 0.2, the predicted rotational angular velocity is 0.1 radians / second, and the duration of the subsequent detection cycle is 10 seconds, then the dwell time t_i = 0.2 × 10 = 2 seconds, and the azimuth interval width parameter Δθ_i = 0.1 × 2 = 0.2 radians (approximately 11.5 degrees). The azimuth interval width parameter determines the coverage area of ​​each candidate key detection airspace in the radar rotating coordinate system.

[0088] Step S129: Integrate the azimuth interval width parameter with the threat level identifier corresponding to each airspace in the potential target threat airspace set to generate a structured radar detection airspace dynamic division scheme description file.

[0089] The azimuth interval width parameters and corresponding threat level identifiers for each candidate key detection airspace are integrated into a structured description file. This description file contains information such as airspace number, initial azimuth value, azimuth width, and threat level identifier. For example, the description information for an airspace might be: number 1, initial azimuth value 30 degrees, azimuth width 10 degrees, threat level identifier 85. The structured description file facilitates reading and parsing by the radar beam scheduling decision module.

[0090] Step S1210: Convert the radar detection airspace dynamic partitioning scheme description file into a radar detection airspace dynamic partitioning scheme data interface for subsequent detection cycles that can be directly called by the radar beam scheduling decision module.

[0091] The structured description file of the dynamic airspace partitioning scheme for radar detection is converted into a data interface format that the radar beam scheduling decision module can recognize and access. This data interface uses standardized data structures, such as arrays of structures, where each structure contains various parameters of the airspace. Through this data interface, the radar beam scheduling decision module can easily obtain airspace partitioning information for subsequent detection cycles, thereby generating corresponding beam scheduling commands. This data interface conversion ensures smooth data interaction between different modules, improving the overall efficiency of the system.

[0092] Step S130: Based on the radar detection airspace dynamic division scheme and the rotation motion state description parameters, generate a set of radar beam parameter scheduling instructions for controlling the radar platform to work in subsequent detection cycles.

[0093] The dynamic airspace partitioning scheme for radar detection identifies the key airspace areas requiring coverage during subsequent detection cycles, while rotational motion state description parameters provide motion information for the radar platform. Combining these two pieces of information, a set of radar beam parameter scheduling instructions is generated. This set includes various parameters controlling the radar's transmit and receive links, such as beam pointing, dwell time, and pulse repetition frequency. Through these beam parameter scheduling instructions, radar resources are optimally allocated, ensuring effective detection of key airspace areas.

[0094] Step S131: Analyze the radar detection airspace dynamic division scheme, and extract the azimuth interval width parameter of each candidate key detection airspace defined in the scheme under the radar platform rotating coordinate system and the threat level identifier corresponding to each candidate key detection airspace.

[0095] The data interface of the dynamic partitioning scheme for radar detection airspace is parsed to extract the azimuth interval width parameter and threat level identifier for each candidate key detection airspace. For example, after parsing, the azimuth interval width of candidate key detection airspace 1 is 10 degrees and the threat level identifier is 85; the azimuth interval width of candidate key detection airspace 2 is 8 degrees and the threat level identifier is 75, etc. These parameters are the basis for generating beam scheduling instructions and determine the allocation of detection resources for each airspace.

[0096] Step S132: Extract the predicted rotational angular velocity value of the radar platform at the start of the subsequent detection cycle from the rotational motion state description parameters.

[0097] The rotational motion state description parameters include the predicted rotational angular velocity value of the radar platform at the start of subsequent detection cycles, which can be directly extracted by parsing this parameter. For example, the predicted rotational angular velocity value is 0.1 radians / second. This predicted rotational angular velocity value is used to calculate the dwell time and azimuth coverage of the radar beam in each airspace.

[0098] Step S133: Based on the azimuth interval width parameter and the predicted rotation angular velocity value, calculate the maximum theoretical time length allowed for the radar beam to scan and cover each candidate key detection airspace in subsequent detection cycles.

[0099] The time it takes for a radar beam to cover a candidate key detection area is related to the azimuth interval width and the predicted rotational angular velocity. The maximum theoretical time length t_max_i = Δθ_i / ω_pred, where Δθ_i is the azimuth interval width parameter and ω_pred is the predicted rotational angular velocity value. For example, if the azimuth interval width is 0.2 radians and the predicted rotational angular velocity is 0.1 radians / second, then the maximum theoretical time length t_max_i = 0.2 / 0.1 = 2 seconds. This maximum theoretical time length is the longest time the radar beam can remain in that airspace.

[0100] Step S134: Based on the threat level identifier corresponding to each candidate key detection airspace, assign a beam energy allocation priority coefficient corresponding to the threat level identifier to each candidate key detection airspace.

[0101] The beam energy allocation priority coefficient is directly proportional to the threat level indicator; the higher the threat level indicator, the larger the priority coefficient. For example, the priority coefficient k_i is set to k_i = s_i / s_max, where s_i is the threat level indicator and s_max is the maximum threat level indicator. If a candidate key detection airspace has a threat level indicator of 85 and a maximum threat level indicator of 100, then the priority coefficient k_i = 0.85. The priority coefficient is used to adjust the dwell time of each airspace, allowing high-threat airspaces to receive more energy allocation.

[0102] Step S135: Based on the maximum theoretical time length and the beam energy allocation priority coefficient corresponding to each candidate key detection airspace, calculate the actual radar beam dwell time length to be allocated to each candidate key detection airspace.

[0103] The actual dwell time t_i = t_max_i × k_i × α, where α is an adjustment coefficient used to ensure that the total dwell time across all airspaces does not exceed the duration of the subsequent detection cycle T. The adjustment coefficient α is determined through iterative calculation, such that Σt_i = T. For example, if the maximum theoretical dwell time is 2 seconds, the priority coefficient is 0.85, and the adjustment coefficient is 0.9, then the actual dwell time t_i = 2 × 0.85 × 0.9 = 1.53 seconds. Using this method, the actual dwell time is allocated according to the threat level while meeting time constraints.

[0104] Step S136: Based on the actual radar beam dwell time length to be allocated to each candidate key detection airspace, and in conjunction with the minimum pulse repetition interval parameter of the radar system, calculate the cumulative number of radar transmitted pulses allocated to each candidate key detection airspace.

[0105] If the minimum pulse repetition interval of the radar system is T_prf_min (seconds), then the cumulative number of transmitted pulses for each candidate key detection airspace is N_i = t_i / T_prf_min. For example, if the actual dwell time is 1.53 seconds and the minimum pulse repetition interval is 100 microseconds (0.0001 seconds), then the cumulative number of transmitted pulses is N_i = 1.53 / 0.0001 = 15300. The cumulative number of transmitted pulses determines the radar's detection energy in that airspace; the higher the number, the higher the detection signal-to-noise ratio and the greater the target detection probability.

[0106] Step S137: Based on the actual dwell time of the radar beam to be allocated to each candidate key detection airspace and the predicted rotation angular velocity value, calculate the actual change in the azimuth angle of the radar platform during the scanning of each candidate key detection airspace by the radar beam.

[0107] The actual change in angle is Δθ_actual_i = ω_pred × t_i. For example, if the actual dwell time is 1.53 seconds and the predicted rotational angular velocity is 0.1 radians / second, then the actual change in angle is Δθ_actual_i = 0.1 × 1.53 = 0.153 radians (approximately 8.77 degrees). This actual change in angle reflects the rotation angle of the radar platform during the dwell time and is used to determine the start and end azimuth angles of beam scanning.

[0108] Step S138: Based on the predicted rotation azimuth angle value and the actual change angle, determine the radar beam scanning start azimuth angle and radar beam scanning end azimuth angle for each candidate key detection airspace.

[0109] The radar beam scan starts with an azimuth angle θ_start_i = θ_start_pred + Σ_{j=1}^{i-1}Δθ_actual_j, where θ_start_pred is the predicted rotation azimuth angle, and Σ_{j=1}^{i-1}Δθ_actual_j is the sum of the actual angle changes of the previous i-1 candidate key detection airspaces. The scan ends with an azimuth angle θ_end_i = θ_start_i + Δθ_actual_i. For example, if the predicted rotation azimuth angle is 30 degrees, the actual angle change of the previous airspace is 8 degrees, and the actual angle change of the current airspace is 8.77 degrees, then the scan starts with an azimuth angle θ_start_i = 30 + 8 = 38 degrees, and the scan ends with an azimuth angle θ_end_i = 38 + 8.77 = 46.77 degrees. By determining the scan start and end azimuth angles, it is ensured that the radar beam can accurately cover the candidate key detection airspaces.

[0110] Step S139: Based on the radar beam dwell time and the cumulative number of radar transmitted pulses, calculate the pulse repetition frequency and pulse width parameters that the radar system should use during the scanning period corresponding to each candidate key detection airspace.

[0111] The pulse repetition frequency (PRF_i) is calculated as N_i / t_i, and the pulse width (τ_i) is set according to the radar system's power and range resolution requirements. Generally, a wider pulse width is chosen to increase transmission energy while meeting range resolution requirements. For example, if the cumulative number of transmitted pulses is 15300 and the dwell time is 1.53 seconds, then the pulse repetition frequency (PRF_i) is 15300 / 1.53 = 10000 Hz. The pulse width is set to 100 nanoseconds to ensure that the range resolution requirements are met. The pulse repetition frequency and pulse width parameters directly affect the radar's detection performance, such as maximum detection range, range resolution, and Doppler resolution.

[0112] Step S1310: Integrate the radar beam scan start azimuth angle, the radar beam scan end azimuth angle, the pulse repetition frequency parameter and the pulse width parameter, and generate a serialized radar beam parameter scheduling instruction set in chronological order.

[0113] The radar beam scanning start azimuth, end azimuth, pulse repetition frequency, and pulse width parameters for each candidate key detection airspace are combined into a beam parameter scheduling command. These commands are then arranged in chronological order according to the scanning time, forming a serialized set of radar beam parameter scheduling commands. Each command contains all the parameters controlling the radar's operation within a specific time period, and the radar system executes the detection tasks sequentially according to the command order. This serialized command set ensures orderly radar beam scheduling, enabling efficient detection of key airspace.

[0114] Step S140: Based on the radar beam parameter scheduling instruction set, control the transmission and reception links of the radar platform, execute target detection operations in the airspace defined by the radar detection airspace dynamic division scheme, and collect the radar echo data set corresponding to the subsequent detection cycle.

[0115] The radar beam parameter scheduling instruction set contains detailed parameters controlling the operation of the radar transmit and receive links. According to these instructions, the radar platform's transmitter and receiver operate according to the set parameters, probing the designated airspace and receiving echo signals. After signal processing, the radar echo data set corresponding to subsequent detection cycles is obtained.

[0116] Step S141: Based on the radar beam scanning start azimuth angle contained in the first radar beam parameter scheduling instruction in the radar beam parameter scheduling instruction set, drive the radar servo control system to adjust the radar antenna beam pointing to the radar beam scanning start azimuth angle.

[0117] The radar servo control system receives the initial azimuth angle from the first command and drives the radar antenna to rotate via a control motor, adjusting the beam pointing to that azimuth angle. During the adjustment process, the servo control system monitors the actual azimuth angle of the antenna in real time through encoder feedback and uses a PID control algorithm to achieve precise alignment, ensuring that the beam pointing error is less than 0.1 degrees. For example, if the initial azimuth angle of the first command is 38 degrees, the servo control system drives the antenna to rotate to 38 degrees, completing the beam pointing adjustment.

[0118] Step S142: After the radar antenna beam reaches the radar beam scanning start azimuth angle, configure the radio frequency signal generation module of the radar transmitter and the signal sampling module of the radar receiver according to the pulse repetition frequency parameter and pulse width parameter contained in the first radar beam parameter scheduling instruction in the radar beam parameter scheduling instruction set.

[0119] Once the radar antenna beam reaches the initial azimuth angle, the beam control module sends pulse repetition frequency and pulse width parameters to the transmitter's RF signal generation module, configuring the signal generator to produce a compliant RF pulse signal. Simultaneously, it sends parameters such as sampling frequency and quantization bits to the receiver's signal sampling module to ensure synchronization between sampling and transmitted pulses. For example, if the pulse repetition frequency is 10000Hz and the pulse width is 100 nanoseconds, the RF signal generation module generates the pulse signal according to these parameters, and the sampling module samples at a sampling frequency of 1GHz.

[0120] Step S143: Start the radar transmitter to transmit a series of radar detection pulse signals according to the configured pulse repetition frequency parameters and pulse width parameters, and simultaneously start the radar receiver to receive the backscattered echo signal from the first candidate key detection airspace defined by the radar detection airspace dynamic division scheme.

[0121] After configuration, the transmitter transmits radar detection pulse signals according to the set pulse repetition frequency and pulse width. Simultaneously, the receiver starts up and receives echo signals from the first candidate key detection airspace. Synchronization between transmission and reception is achieved through a system clock, ensuring that the echo signal corresponding to each transmitted pulse is accurately received. After being transmitted by the antenna, the radar detection pulse signal propagates through space and encounters a target, causing backscattering. The scattered signal is received by the antenna and sent to the receiver.

[0122] Step S144: Perform down-conversion and analog-to-digital conversion on the received backscattered echo signal to generate the original digital echo data block corresponding to the first candidate key detection airspace.

[0123] The received backscattered echo signal is first down-converted to an intermediate frequency (IF) signal, and then converted from an analog-to-digital converter (ADC) to a digital signal. The down-conversion process uses a superheterodyne receiver, mixing the echo signal with a local oscillator signal generated by a local oscillator to obtain the IF signal. The ADC uses a 12-bit converter, and the sampling frequency is set according to the IF signal bandwidth to ensure distortion-free sampling. The generated raw digital echo data blocks are arranged according to fast and slow time dimensions, forming a two-dimensional array to store the echo data of the first candidate key detection area.

[0124] Step S145: Continuously monitor the real-time feedback value of the radar platform's rotation azimuth angle. When the real-time feedback value of the radar platform's rotation azimuth angle reaches the radar beam scanning end azimuth angle contained in the first radar beam parameter scheduling instruction in the radar beam parameter scheduling instruction set, immediately stop the radar signal transmission and reception operation for the first candidate key detection airspace.

[0125] During the detection process, the radar servo control system monitors the rotation azimuth angle in real time and compares the feedback value with the scan end azimuth angle in the command. When the real-time feedback value reaches the scan end azimuth angle, the beam control module sends a stop command, the transmitter stops transmitting pulse signals, and the receiver stops receiving echo signals. For example, if the scan end azimuth angle is 46.77 degrees, when the real-time azimuth angle reaches 46.77 degrees, the detection of the first candidate key detection airspace is stopped.

[0126] Step S146: According to the order of the instructions in the radar beam parameter scheduling instruction set, the operations of driving the radar antenna beam pointing adjustment, configuring radar transmission and reception parameters, transmitting and receiving radar signals, and generating raw digital echo data blocks are repeated in sequence until the scanning and detection of all candidate key detection airspace is completed.

[0127] After completing the detection of the first candidate key detection airspace, the same operations are performed on the next candidate key detection airspace in the order of the instruction set: adjusting beam pointing, configuring parameters, transmitting and receiving signals, and generating data blocks. This process continues until all candidate key detection airspaces have been scanned and detected. During this process, the detection parameters for each airspace are set according to the corresponding instructions to ensure the accuracy and specificity of the detection.

[0128] Step S147: After completing the scanning and detection of all candidate key detection airspaces, the generated raw digital echo data blocks corresponding to each candidate key detection airspace are spliced ​​and time-marked according to the detection time sequence.

[0129] The raw digital echo data blocks from all candidate key detection airspaces are stitched together in chronological order of detection to form a complete raw digital echo data sequence. Simultaneously, a corresponding timestamp is added to each data block to record the start and end times of the detection. This stitching and timestamp processing facilitates subsequent analysis and processing of the echo data, enabling accurate correlation between the echo data and the radar's rotation status.

[0130] Step S148: Perform pulse compression and moving target display filtering on the complete original digital echo data sequence after splicing and time stamping to suppress ground clutter and improve the target signal-to-noise ratio.

[0131] Pulse compression processing uses matched filtering to compress wide pulses into narrow pulses, improving range resolution. Moving target display filtering uses an MTI filter to suppress ground clutter while preserving moving target signals. Pulse compression processing uses a filter matched to the transmitted pulse to perform convolution operations on the echo data; the MTI filter employs a recursive filter structure, eliminating fixed clutter by subtracting the echo signals of adjacent pulses. After these processing steps, the signal-to-noise ratio of the target echo signal is significantly improved, and clutter interference is effectively suppressed.

[0132] Step S149: Output the high-resolution radar echo data sequence after pulse compression and moving target indication filtering, as the radar echo data set corresponding to the subsequent detection period.

[0133] The high-resolution radar echo data sequence, after pulse compression and moving target indication filtering, contains clear target echo information and is output as the radar echo data set for subsequent detection cycles. This radar echo data set will be used to update the radar platform's rotational motion state parameters and dynamic airspace partitioning scheme, achieving closed-loop iteration of the detection process.

[0134] Step S150: Based on the radar echo data set and the rotational motion state description parameters, update the expected rotational motion state parameters of the radar platform at the start of the next detection cycle and the dynamic partitioning scheme of the radar detection airspace.

[0135] The radar echo dataset contains the target's latest position and motion information, while the rotational motion state description parameters reflect the current motion state of the radar platform. By combining these two pieces of information, the expected rotational motion state parameters for the next detection cycle and the dynamic partitioning scheme of the radar detection airspace are updated, enabling the radar to adapt to changes in target and its own motion and continuously optimize the detection strategy.

[0136] Step S151: Perform constant false alarm rate (CFAR) detection processing on the radar echo data set, extract target points that exceed the detection threshold, and record the distance value, Doppler velocity value, and azimuth value corresponding to each target point.

[0137] A constant false alarm rate (CFAR) detection algorithm is used to process the radar echo data set. The CFAR probability is set to a fixed value (e.g., 10^-6), and the detection threshold is adaptively adjusted according to the noise level. The echo signal amplitude of each range cell and Doppler cell is compared with the detection threshold; signals exceeding the threshold are considered target points. The range, Doppler velocity, and azimuth angle of each target point are recorded to form a target point list. CFAR detection maintains a stable CFAR probability under different noise environments, improving the reliability of target detection.

[0138] Step S152: Based on the azimuth angle value corresponding to each target point and the timestamp information during the acquisition process of the radar echo data set, fit and calculate the azimuth angle change trajectory of each target point in the rotating coordinate system of the radar platform.

[0139] Each target point has a corresponding azimuth value and timestamp. By performing curve fitting on the above data, the trajectory of azimuth change over time is obtained. The least squares method is used to fit the azimuth-time curve to obtain the trend equation of azimuth change. For example, the fitted azimuth change trajectory is θ(t) = a × t + b, where a and b are fitting coefficients. The azimuth change trajectory reflects the target's motion in the radar rotating coordinate system.

[0140] Step S153: Based on the azimuth change trajectory of each target point in the rotating coordinate system of the radar platform, and combined with the actual rotational motion state record of the radar platform in subsequent detection cycles, calculate the absolute velocity vector and absolute direction of motion of each target point in the inertial coordinate system.

[0141] The radar platform's actual rotational motion record during subsequent detection cycles includes the changes in angular velocity and azimuth angle over time. The target's azimuth angle in the radar's rotating coordinate system changes at a rate of dθ_radar / dt, and the radar platform's rotational angular velocity is ω_radar. Therefore, the target's azimuth angle in the inertial coordinate system changes at a rate of dθ_inertial / dt = dθ_radar / dt + ω_radar. Combining this with the target's range value R, the target's tangential velocity v_tangential = R × dθ_inertial / dt can be calculated. The target's radial velocity v_radial is obtained from the Doppler velocity value. The magnitude of the absolute velocity vector is sqrt(v_radial² + v_tangential²), and its direction is arctan(v_tangential / v_radial). Through this method, the target's absolute velocity vector and direction in the inertial coordinate system are obtained by inversion.

[0142] Step S154: Based on the absolute velocity vector and absolute direction of motion of each target point in the inertial coordinate system, predict the position coordinates and velocity vector of each target point in the inertial coordinate system at the beginning of the next detection cycle.

[0143] Based on the target's current position coordinates, absolute velocity vector, and direction of motion, a uniform linear motion model is used to predict the position coordinates and velocity vector at the start of the next detection cycle. The predicted position coordinates are the current position coordinates plus the product of the velocity vector and the time interval, while the velocity vector remains unchanged. For example, if the current position coordinates are (x0, y0), the velocity vector is (vx, vy), and the time interval is T, then the predicted position coordinates are (x0 + vx × T, y0 + vy × T), and the velocity vector remains (vx, vy).

[0144] Step S155: Transform the position coordinates of each target point in the inertial coordinate system at the beginning of the next detection cycle to the rotating coordinate system of the radar platform, and obtain the predicted relative azimuth and predicted relative distance of each target point relative to the radar platform at the beginning of the next detection cycle.

[0145] Through coordinate transformation, the predicted position coordinates in the inertial coordinate system are converted to polar coordinates in the rotating coordinate system of the radar platform. Let the predicted position coordinates in the inertial coordinate system be (x, y), and the position of the radar platform in the inertial coordinate system be (x_radar, y_radar). Then the relative coordinates are (Δx = x - x_radar, Δy = y - y_radar). The predicted relative distance R_relative = sqrt(Δx² + Δy²), and the predicted relative azimuth θ_relative = arctan²(Δy, Δx). Coordinate transformation ensures an accurate representation of the target position in the radar coordinate system, facilitating subsequent airspace delineation.

[0146] Step S156: Update the threat level identifier and airspace boundary parameters of the corresponding airspace in the potential target threat airspace set according to the predicted relative azimuth angle and the predicted relative distance.

[0147] The predicted relative azimuth and predicted relative distance are matched with airspaces in the potential target threat airspace set to find the corresponding airspace. Based on changes in the predicted position and velocity of the target, the threat level label and airspace boundary parameters of that airspace are adjusted. For example, if the target approaches the radar and its velocity increases, the threat level label is increased, and the airspace boundary parameters are expanded accordingly. The updated potential target threat airspace set more accurately reflects the current state and threat level of the targets.

[0148] Step S157: Based on the actual rotational motion state record of the radar platform in subsequent detection cycles, correct the internal parameters of the radar platform rotational motion state prediction model, and use the corrected radar platform rotational motion state prediction model to re-predict the predicted rotational angular velocity and predicted rotational azimuth angle of the radar platform at the beginning of the next detection cycle.

[0149] The radar platform rotation motion state prediction model may contain prediction errors. Therefore, the model parameters are corrected using actual rotation motion state records. Methods such as least squares or Kalman filtering are employed to adjust parameters such as the model's state transition matrix and observation matrix based on the residuals between the actual measured and predicted values. The corrected model then re-predicts the rotational angular velocity and azimuth angle values ​​at the start of the next detection cycle, improving prediction accuracy and ensuring the accuracy of airspace division and beam scheduling.

[0150] Step S1571: The actual rotational motion state record of the radar platform in subsequent detection cycles includes the sequence of actual rotational angular velocity measurements and the sequence of actual rotational azimuth measurements at multiple sampling moments in subsequent detection cycles.

[0151] The actual rotational motion state is recorded by the radar platform's inertial measurement unit and rotary encoder during subsequent detection cycles. It includes actual rotational angular velocity measurements and actual rotational azimuth measurements at multiple sampling moments, forming two one-dimensional sequences. For example, if the sampling frequency is 200Hz and the subsequent detection cycle is 10 seconds, then each sequence contains 2000 sampling points.

[0152] Step S1572: The azimuth angle value corresponding to the target point extracted after constant false alarm rate detection and processing of the radar echo data set and the timestamp information during the acquisition process are combined with the actual rotation azimuth angle measurement value sequence, and the theoretical azimuth angle prediction value of the radar platform beam pointing corresponding to each target point is calculated by time synchronization comparison.

[0153] The azimuth angle of the target point is collected when the radar beam is pointing at the target. Combined with the timestamp information, the azimuth angle measurement value at the corresponding time in the actual rotation azimuth angle measurement value sequence can be found and used as the theoretical azimuth angle prediction value. For example, if the acquisition timestamp of the target point is t, then the measurement value θ_actual(t) at time t can be found from the actual rotation azimuth angle measurement value sequence and used as the theoretical azimuth angle prediction value.

[0154] Step S1573: Calculate the azimuth prediction residual sequence between the theoretical azimuth prediction value of the radar platform beam pointing corresponding to each target point and the actual measurement value at the corresponding time in the actual rotation azimuth measurement value sequence.

[0155] The azimuth prediction residual δθ_i = θ_pred_i - θ_actual_i, where θ_pred_i is the theoretical azimuth prediction value and θ_actual_i is the actual measured value. The residuals of all target points are combined into an azimuth prediction residual sequence, which reflects the error of the prediction model.

[0156] Step S1574: Perform statistical analysis on the azimuth prediction residual sequence, and calculate the mean drift and standard deviation fluctuation of the azimuth prediction residual sequence as a quantitative indicator to evaluate the current prediction performance of the radar platform rotation motion state prediction model.

[0157] The mean shift μδθ = (1 / N)Σδθ_i, where N is the length of the residual sequence; the standard deviation fluctuation σδθ = sqrt((1 / N)Σ(δθ_i-μδθ)²). The mean shift reflects the systematic error of the prediction model, while the standard deviation fluctuation reflects the dispersion of the prediction error. These two indicators quantify the predictive performance of the model.

[0158] Step S1575: Based on the mean drift and standard deviation fluctuation of the azimuth prediction residual sequence, derive the angular velocity compensation correction amount used to correct the angular velocity prediction value of the next cycle.

[0159] The angular velocity compensation correction Δω = k1 × μδθ + k2 × σδθ, where k1 and k2 are correction coefficients determined empirically or through an adaptive algorithm. For example, if k1 = 0.5, k2 = 0.3, and the mean drift is 0.02 radians and the standard deviation fluctuation is 0.01 radians, then the angular velocity compensation correction Δω = 0.5 × 0.02 + 0.3 × 0.01 = 0.013 radians / second. The angular velocity compensation correction is used to adjust the angular velocity output of the prediction model.

[0160] Step S1576: Based on the mean drift and standard deviation fluctuation of the azimuth prediction residual sequence, derive the azimuth compensation correction amount used to correct the azimuth prediction value for the next period.

[0161] The azimuth compensation correction Δθ = m1 × μδθ + m2 × σδθ, where m1 and m2 are correction coefficients. For example, if m1 = 1.0, m2 = 0.2, the mean shift is 0.02 radians, and the standard deviation fluctuation is 0.01 radians, then the azimuth compensation correction Δθ = 1.0 × 0.02 + 0.2 × 0.01 = 0.022 radians. The azimuth compensation correction is used to adjust the azimuth output of the prediction model.

[0162] Step S1577: Input the angular velocity compensation correction amount and the azimuth angle compensation correction amount into the parameter adaptive update interface of the radar platform rotation motion state prediction model, and incrementally adjust the state transition matrix parameters and observation matrix parameters inside the radar platform rotation motion state prediction model according to the preset learning rate coefficient.

[0163] The learning rate coefficient is set to α (e.g., 0.1), and the adjustment amounts for the state transition matrix parameter A and the observation matrix parameter H are ΔA = α × Δω × A and ΔH = α × Δθ × H, respectively. Through incremental adjustments, the model parameters gradually adapt to the actual motion state, reducing prediction errors. The parameter adaptive update interface enables online learning and optimization of the model.

[0164] Step S1578: Using the radar platform rotation motion state prediction model with adjusted parameters, the actual rotation angular velocity measurement value and the actual rotation azimuth angle measurement value of the radar platform at the end of the subsequent detection cycle are used as new initial state inputs to perform forward inference calculation of the radar platform rotation motion state prediction model.

[0165] The parameter-adjusted prediction model uses the actual rotational angular velocity and azimuth at the end of the subsequent detection cycle as the initial state, and calculates the predicted value at the start of the next detection cycle through forward inference. The forward inference process is based on the model's state equations and observation equations, taking into account the rotational dynamics of the radar platform.

[0166] Step S1579: Obtain the intermediate values ​​of angular velocity and azimuth at the start of the next detection cycle from the radar platform rotation motion state prediction model after parameter adjustment.

[0167] The intermediate values ​​for the model's angular velocity prediction are ω_mid and the intermediate value for the azimuth prediction are θ_mid. These intermediate values ​​have not yet been compensated or corrected.

[0168] Step S15710: The predicted intermediate value of the azimuth angle is fused with the azimuth angle compensation correction amount to obtain the predicted rotation azimuth angle value of the radar platform at the start of the next detection cycle.

[0169] The final predicted azimuth angle value θ_pred_final = θ_mid + Δθ. For example, if the midpoint of the predicted azimuth angle is 30 degrees and the azimuth angle compensation correction is 0.022 radians (approximately 1.26 degrees), then the final predicted azimuth angle value is 31.26 degrees.

[0170] Step S15711: The intermediate value of the angular velocity prediction is fused with the angular velocity compensation correction amount to obtain the final re-predicted angular velocity value of the radar platform at the start of the next detection cycle.

[0171] The final predicted rotational angular velocity value ω_pred_final = ω_mid + Δω. For example, if the median predicted angular velocity is 0.1 radians / second and the angular velocity compensation correction is 0.013 radians / second, then the final predicted rotational angular velocity value is 0.113 radians / second.

[0172] Step S158: Integrate the updated set of potential target threat airspace with the re-predicted rotational angular velocity and rotational azimuth values ​​to generate an updated dynamic partitioning scheme for radar detection airspace for the next detection cycle.

[0173] The updated potential target threat airspace set contains the latest threat information for the targets, and the re-predicted rotation angular velocity and azimuth angle values ​​reflect the latest motion state of the radar platform. By fusing the above information and following a method similar to step S120, an updated dynamic radar detection airspace partitioning scheme for the next detection cycle is generated, ensuring that the radar can continuously and effectively detect key targets.

[0174] Step S1581: The updated potential target threat airspace set includes the updated threat level identifier and updated airspace boundary parameters for each sector airspace description block.

[0175] The threat level identifier and airspace boundary parameters of each sector airspace description block in the updated potential target threat airspace set have been adjusted based on the predicted location and motion status of the target, more accurately reflecting the current target threat situation.

[0176] Step S1582: Based on the predicted rotational angular velocity value and the predicted rotational azimuth angle value, recalculate the updated theoretical maximum airspace boundary that the radar platform can cover in the next detection cycle after completing a single-cycle rotational scan.

[0177] Similar to step S126, the updated theoretical maximum airspace boundary is calculated based on the re-predicted rotational angular velocity and azimuth values.

[0178] Step S1583: Extract the updated threat level identifiers of all sector airspace description blocks from the updated potential target threat airspace set, and sort them in descending order according to the numerical value of the threat level identifiers to generate a threat level priority queue.

[0179] The updated threat level identifiers are sorted from high to low to form a threat level priority queue, ensuring that high-threat airspace is given priority in resource allocation.

[0180] Step S1584: For a preset number of sector airspace description blocks sorted in the threat level priority queue, calculate the required initial azimuth interval width parameters of the sector airspace description block in the radar platform rotating coordinate system based on the updated airspace boundary parameters of the sector airspace description block.

[0181] The preset number is set based on radar resources and detection requirements, for example, selecting the top 10 high-threat airspaces. Based on the updated boundary parameters of each airspace, the initial azimuth interval width requirement parameter is calculated as the initial basis for resource allocation.

[0182] Step S1585: Based on the initial azimuth interval width requirement parameter and the predicted rotation angular velocity value, preliminarily estimate the minimum beam dwell time requirement parameter required to cover each sector spatial description block.

[0183] The minimum beam dwell time requirement parameter t_min_i = Δθ_demand_i / ω_pred_final, where Δθ_demand_i is the initial azimuth interval width requirement parameter. This initial azimuth interval width requirement parameter ensures the minimum dwell time for the radar to cover the airspace.

[0184] Step S1586: Compare the sum of the minimum beam dwell time requirement parameters with the total time budget available for key detection by the radar platform in the next detection cycle. If the total requirement exceeds the total budget, reduce the allocation time of the lower-ranked sector airspace description blocks in the order of threat level priority queue until the total requirement meets the total budget constraint.

[0185] The total time budget is the duration T of the next detection cycle. If Σt_min_i > T, the dwell time is reduced starting from the airspace with the lowest threat level, with a certain percentage (e.g., 10%) reduced each time, until Σt_min_i ≤ T. This method optimizes resource allocation under time budget constraints.

[0186] Step S1587: Based on the beam dwell time parameter finally assigned to each sector airspace description block and the predicted rotation angular velocity value, recalculate the updated azimuth interval width parameter that each sector airspace description block is actually assigned to occupy in the radar platform rotation coordinate system.

[0187] Similar to step S128, the updated azimuth interval width parameter Δθ_updated_i = ω_pred_final × t_final_i, where t_final_i is the final assigned dwell time.

[0188] Step S1588: Integrate the updated azimuth interval width parameters of all sector airspace description blocks with their corresponding updated threat level identifiers to generate an updated structured radar detection airspace dynamic partitioning scheme description file.

[0189] The updated azimuth interval width parameter and threat level identifier are integrated into the structured description file, which includes information such as airspace number, starting azimuth, width, and threat level.

[0190] Step S1589: Convert the updated structured radar detection airspace dynamic partitioning scheme description file into an updated radar detection airspace dynamic partitioning scheme data interface for the next detection cycle, which can be directly called by the radar beam scheduling decision module in the next round.

[0191] The structured description file is converted into a data interface for the radar beam scheduling decision module to call, thereby generating the next round of beam scheduling instructions.

[0192] Step S159: The re-predicted rotational angular velocity value and the predicted rotational azimuth angle value are used as new expected rotational motion state parameters. The updated radar detection airspace dynamic partitioning scheme is used as the new radar detection airspace dynamic partitioning scheme and input into the next round of radar beam parameter scheduling instruction generation step, thereby realizing the continuous iterative operation of the closed-loop beam scheduling detection process based on radar platform rotation dynamics.

[0193] The updated expected rotational motion state parameters and the dynamic partitioning scheme of the radar detection airspace are input into step S130 to begin the next round of beam parameter scheduling command generation, forming a closed-loop iterative process. Through continuous iteration, the radar can adapt to changes in target and its own motion, optimize detection strategies, and improve detection performance.

[0194] In one exemplary embodiment, a rotating radar detection system based on intelligent beam scheduling is provided. This system can be a terminal, server, etc., and its internal structure diagram can be as follows: Figure 2 As shown, the intelligent beam-spacing-based rotating radar detection system includes a processor, memory, input / output interface, communication interface, display unit, and input device. The processor, memory, and input / output interface are connected via a system bus, and the communication interface, display unit, and input device are also connected to the system bus via the input / output interface. The processor provides computational and control capabilities. The memory includes a non-volatile storage medium and internal memory. The non-volatile storage medium stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage medium. The input / output interface is used for exchanging information between the processor and external devices. The communication interface is used for wired or wireless communication with external terminals; wireless communication can be achieved through Wi-Fi, mobile cellular networks, near-field communication, or other technologies. When the computer program is executed by the processor, it implements a rotating radar detection method based on intelligent beam-spacing. The display unit is used to form a visually visible image and can be a display screen, projection device, or virtual reality imaging device. The display screen can be an LCD screen or an e-ink screen. The input device can be a touch layer covering the display screen, or a button, trackball, or touchpad set on the housing of a rotating radar detection system based on intelligent beam scheduling, or an external keyboard, touchpad, or mouse, etc.

[0195] It should be noted that, in order to simplify the description of the present invention and thus help to understand one or more embodiments of the invention, multiple features may sometimes be grouped into one embodiment, drawing or description thereof in the foregoing description of the embodiments of the present invention.

Claims

1. A rotating radar detection method based on intelligent beam scheduling, characterized in that, The method includes: Acquire the rotational motion state description parameters of the radar platform during the current detection period and the preliminary radar echo data set corresponding to the current detection period; Based on the rotational motion state description parameters and the preliminary radar echo data set, a dynamic radar detection airspace partitioning scheme is generated for subsequent detection cycles. Based on the radar detection airspace dynamic division scheme and the rotation motion state description parameters, a set of radar beam parameter scheduling instructions is generated to control the operation of the radar platform in subsequent detection cycles. Based on the radar beam parameter scheduling instruction set, the transmission and reception links of the radar platform are controlled to perform target detection operations in the airspace defined by the radar detection airspace dynamic division scheme, and radar echo data sets corresponding to subsequent detection cycles are collected. Based on the radar echo data set and the rotational motion state description parameters, update the expected rotational motion state parameters of the radar platform at the start of the next detection cycle and the dynamic partitioning scheme of the radar detection airspace.

2. The rotating radar detection method based on intelligent beam scheduling according to claim 1, characterized in that, The step of generating a dynamic radar detection airspace partitioning scheme for subsequent detection cycles based on the rotational motion state description parameters and the preliminary radar echo data set includes: The real-time rotational angular velocity sequence and the real-time rotational azimuth sequence of the radar platform in the current detection period are extracted from the rotational motion state description parameters. Based on the real-time rotational angular velocity sequence and the real-time rotational azimuth sequence, calculate the predicted rotational angular velocity value of the radar platform at the beginning of the subsequent detection cycle and the predicted rotational azimuth value of the radar platform at the beginning of the subsequent detection cycle. The preliminary radar echo data set is processed to extract moving target features, thereby obtaining the azimuth distribution features and radial velocity distribution features of potential moving targets contained in the preliminary radar echo data set. Based on the azimuth distribution characteristics and radial velocity distribution characteristics of the potential moving targets, a set of potential target threat airspace that needs to be focused on in subsequent detection cycles is constructed. By integrating the predicted rotational angular velocity value, the predicted rotational azimuth angle value, and the potential target threat airspace set, a candidate key detection airspace set that the radar beam should prioritize covering in subsequent detection cycles is generated. Based on the predicted rotational angular velocity and the predicted rotational azimuth, the theoretical maximum airspace boundary that the radar platform can cover in a single rotational scan during subsequent detection cycles is calculated. Based on the candidate key detection airspace set and the theoretical maximum airspace boundary, airspace resource allocation optimization is performed to obtain the theoretical dwell time ratio parameter of each candidate key detection airspace in the subsequent detection cycle. Based on the theoretical dwell time ratio parameter and the predicted rotation angular velocity value, the azimuth interval width parameter occupied by each candidate key detection airspace in the radar platform rotation coordinate system is calculated in reverse. By integrating the azimuth interval width parameter with the threat level identifier corresponding to each airspace in the potential target threat airspace set, a structured radar detection airspace dynamic partitioning scheme description file is generated; The description file of the radar detection airspace dynamic partitioning scheme is converted into a data interface for the radar detection airspace dynamic partitioning scheme for subsequent detection cycles, which can be directly called by the radar beam scheduling decision module.

3. The rotating radar detection method based on intelligent beam scheduling according to claim 2, characterized in that, The process of extracting moving target features from the preliminary radar echo data set to obtain the azimuth distribution features and radial velocity distribution features of potential moving targets contained in the preliminary radar echo data set includes: The initial radar echo data set is divided into range gate units to obtain multiple radar echo range gate data segments arranged according to the range dimension; Doppler spectrum analysis was performed on each radar echo range gate data segment to obtain the Doppler power spectral density distribution curve corresponding to each radar echo range gate data segment; Significant spectral peaks exceeding the noise floor threshold are detected from the Doppler power spectral density distribution curves corresponding to each radar echo range gate data segment, and the Doppler frequency and power amplitude values ​​corresponding to each significant spectral peak are recorded. Based on the Doppler frequency value corresponding to each significant spectral peak point, and combined with the carrier wavelength parameter of the radar transmitted signal, the estimated radial velocity of the potential target corresponding to each significant spectral peak point is calculated. Based on the range gate number corresponding to the radar echo range gate data segment to which each significant spectral peak point belongs, and in conjunction with the radar range resolution parameters, the potential target range estimate corresponding to each significant spectral peak point is calculated. Using the real-time rotating azimuth sequence, the instantaneous beam pointing azimuth of the radar platform corresponding to each significant spectral peak point is determined through interpolation matching, and the instantaneous beam pointing azimuth of the radar platform is used as the azimuth estimate of the corresponding potential target. Aggregate the potential target radial velocity estimates, potential target distance estimates, and potential target azimuth estimates corresponding to all significant spectral peak points to form an original set of potential target points; The original set of potential target points is subjected to point clustering processing based on three-dimensional space of distance, orientation, and velocity, and potential target points that are close in spatial location and have similar velocity characteristics are merged into the same potential target point cluster; Extract the statistical distribution center of the azimuth angle estimates of all points within each potential target point cluster as the comprehensive azimuth angle feature of the potential target point cluster, and extract the statistical distribution center of the radial velocity estimates of all points within the potential target point cluster as the comprehensive radial velocity feature of the potential target point cluster. By integrating the comprehensive azimuth and radial velocity features of all potential target point clusters, an azimuth feature vector characterizing the azimuth distribution of potential moving targets contained in the preliminary radar echo data set and a radial velocity feature vector characterizing the radial velocity distribution of potential moving targets contained in the preliminary radar echo data set are generated.

4. The rotating radar detection method based on intelligent beam scheduling according to claim 2, characterized in that, The step involves constructing a set of potential target threat airspace that requires close attention during subsequent detection cycles based on the azimuth distribution characteristics and radial velocity distribution characteristics of the potential moving targets. This set includes: Based on the sign and absolute value of each radial velocity feature component in the radial velocity distribution characteristics of the potential moving target, the radial motion tendency attribute and radial motion intensity attribute of the corresponding potential target point cluster relative to the radar platform are determined. Based on the value of each azimuth feature component in the azimuth feature vector of the potential moving target azimuth distribution characteristics, and combined with the statistical trend of the real-time rotating azimuth sequence, the azimuth change rate parameter of the corresponding potential target point cluster relative to the rotation center of the radar platform is calculated. By integrating the radial motion tendency attribute of the corresponding potential target cluster relative to the radar platform, the radial motion intensity attribute of the corresponding potential target cluster relative to the radar platform, and the azimuth angle change rate parameter of the corresponding potential target cluster relative to the rotation center of the radar platform, the real-time threat level score of each potential target cluster is calculated through a pre-set threat assessment rule base. All potential target point clusters are sorted and filtered based on the real-time threat level score of each potential target point cluster. Potential target point clusters with real-time threat level scores exceeding a preset threat level threshold are selected to form a subset of potential target point clusters. For each potential target point cluster in the subset of potential target point clusters, based on the comprehensive azimuth and comprehensive radial velocity characteristics of the potential target point cluster, the range of azimuth and range of distance that the potential target point cluster may appear in subsequent detection cycles are predicted. Based on the azimuth interval range and range range that each potential target cluster may appear in during subsequent detection cycles, a sector-shaped airspace description block is defined in the radar polar coordinate system, constrained by the azimuth interval boundary and the range interval boundary. For each sector-shaped spatial description block, a real-time threat level score is attached to generate the source potential target point trace cluster corresponding to that sector-shaped spatial description block, which serves as the threat level identifier for that sector-shaped spatial description block. All sector-shaped airspace description blocks with added threat level identifiers are aggregated to form a structured database of potential target threat airspace sets that need to be focused on in subsequent detection cycles.

5. The rotating radar detection method based on intelligent beam scheduling according to claim 1, characterized in that, The step of generating a set of radar beam parameter scheduling instructions for controlling the radar platform's operation in subsequent detection cycles, based on the radar detection airspace dynamic partitioning scheme and the rotational motion state description parameters, includes: The radar detection airspace dynamic division scheme is analyzed, and the azimuth interval width parameter of each candidate key detection airspace defined in the scheme and the threat level identifier corresponding to each candidate key detection airspace are extracted in the radar platform rotating coordinate system. The predicted rotational angular velocity value of the radar platform at the start of the subsequent detection cycle is extracted from the rotational motion state description parameters; Based on the azimuth interval width parameter and the predicted rotation angular velocity value, calculate the maximum theoretical time length allowed for the radar beam to scan and cover each candidate key detection airspace in subsequent detection cycles; Based on the threat level identifier corresponding to each candidate key detection airspace, assign a beam energy allocation priority coefficient that corresponds to the threat level identifier to each candidate key detection airspace; Based on the maximum theoretical time length and the beam energy allocation priority coefficient corresponding to each candidate key detection airspace, calculate the actual radar beam dwell time length that should be allocated to each candidate key detection airspace. Based on the actual radar beam dwell time length to be allocated to each candidate key detection airspace, and combined with the minimum pulse repetition interval parameter of the radar system, calculate the cumulative number of radar transmitted pulses allocated to each candidate key detection airspace. Based on the actual radar beam dwell time length to be allocated to each candidate key detection airspace and the predicted rotation angular velocity value, calculate the actual change in the radar platform rotation azimuth angle during the scanning of each candidate key detection airspace by the radar beam. Based on the predicted rotation azimuth angle value and the actual change angle, the radar beam scanning start azimuth angle and radar beam scanning end azimuth angle for each candidate key detection airspace are determined. Based on the radar beam dwell time and the cumulative number of radar transmitted pulses, calculate the pulse repetition frequency and pulse width parameters that the radar system should use during the scanning period corresponding to each candidate key detection airspace. The radar beam scan start azimuth angle, the radar beam scan end azimuth angle, the pulse repetition frequency parameter, and the pulse width parameter are integrated to generate a serialized set of radar beam parameter scheduling instructions in chronological order.

6. The rotating radar detection method based on intelligent beam scheduling according to claim 1, characterized in that, The system controls the transmission and reception links of the radar platform based on the radar beam parameter scheduling command set, performs target detection operations on the airspace defined by the radar detection airspace dynamic partitioning scheme, and collects radar echo data sets corresponding to subsequent detection cycles, including: According to the radar beam scanning start azimuth angle contained in the first radar beam parameter scheduling instruction in the radar beam parameter scheduling instruction set, drive the radar servo control system to adjust the radar antenna beam pointing to the radar beam scanning start azimuth angle. After the radar antenna beam reaches the radar beam scanning start azimuth angle, the radio frequency signal generation module of the radar transmitter and the signal sampling module of the radar receiver are configured according to the pulse repetition frequency parameter and pulse width parameter contained in the first radar beam parameter scheduling instruction in the radar beam parameter scheduling instruction set. The radar transmitter is started to transmit a series of radar detection pulse signals according to the configured pulse repetition frequency parameters and pulse width parameters, and the radar receiver is started simultaneously to receive the backscattered echo signal from the first candidate key detection airspace defined by the radar detection airspace dynamic division scheme. The received backscattered echo signal is downconverted and analog-to-digital converted to generate the original digital echo data block corresponding to the first candidate key detection airspace. The real-time feedback value of the radar platform rotation azimuth angle is continuously monitored. When the real-time feedback value of the radar platform rotation azimuth angle reaches the radar beam scanning end azimuth angle contained in the first radar beam parameter scheduling instruction in the radar beam parameter scheduling instruction set, the radar signal transmission and reception operation for the first candidate key detection airspace is immediately stopped. According to the order of the instructions in the radar beam parameter scheduling instruction set, the operations of driving the radar antenna beam pointing adjustment, configuring radar transmission and reception parameters, transmitting and receiving radar signals, and generating raw digital echo data blocks are repeated in sequence until the scanning and detection of all candidate key detection airspace is completed. After completing the scanning and detection of all candidate key detection airspaces, the generated raw digital echo data blocks corresponding to each candidate key detection airspace are spliced ​​and time-marked according to the detection time sequence. Pulse compression and moving target display filtering are performed on the complete raw digital echo data sequence after splicing and time stamping to suppress ground clutter and improve the target signal-to-noise ratio. The high-resolution radar echo data sequence, after pulse compression and moving target indication filtering, is output as the radar echo data set for subsequent detection cycles.

7. The rotating radar detection method based on intelligent beam scheduling according to claim 2, characterized in that, The step of updating the expected rotational motion state parameters of the radar platform at the start of the next detection cycle and the dynamic partitioning scheme of the radar detection airspace based on the radar echo data set and the rotational motion state description parameters includes: The radar echo data set is processed by constant false alarm rate detection to extract target points that exceed the detection threshold, and the distance value, Doppler velocity value and azimuth value corresponding to each target point are recorded. Based on the azimuth value corresponding to each target point and the timestamp information during the acquisition of the radar echo data set, the trajectory of the azimuth change of each target point in the rotating coordinate system of the radar platform is fitted and calculated. Based on the azimuth change trajectory of each target point in the rotating coordinate system of the radar platform, and combined with the actual rotational motion state record of the radar platform in subsequent detection cycles, the absolute motion velocity vector and absolute motion direction of each target point in the inertial coordinate system are calculated by inversion. Based on the absolute velocity vector and absolute direction of motion of each target point in the inertial coordinate system, predict the position coordinates and velocity vector of each target point in the inertial coordinate system at the beginning of the next detection cycle. The position coordinates of each target point in the inertial coordinate system at the beginning of the next detection cycle are transformed to the rotating coordinate system of the radar platform to obtain the predicted relative azimuth and predicted relative distance of each target point relative to the radar platform at the beginning of the next detection cycle. Based on the predicted relative azimuth and the predicted relative distance, update the threat level identifier and airspace boundary parameters of the corresponding airspace in the potential target threat airspace set; Based on the actual rotational motion state records of the radar platform in subsequent detection cycles, the internal parameters of the radar platform rotational motion state prediction model are corrected, and the corrected radar platform rotational motion state prediction model is used to re-predict the predicted rotational angular velocity and predicted rotational azimuth angle of the radar platform at the beginning of the next detection cycle. By integrating the updated set of potential target threat airspace with the re-predicted predicted rotational angular velocity and predicted rotational azimuth values, an updated dynamic partitioning scheme for radar detection airspace is generated for the next detection cycle. The re-predicted rotational angular velocity and rotational azimuth angle are used as new expected rotational motion state parameters. The updated radar detection airspace dynamic partitioning scheme is used as the new radar detection airspace dynamic partitioning scheme and input into the next round of radar beam parameter scheduling instruction generation step, thereby realizing the continuous iterative operation of the closed-loop beam scheduling detection process based on radar platform rotational dynamics.

8. The rotating radar detection method based on intelligent beam scheduling according to claim 7, characterized in that, The step of correcting the internal parameters of the radar platform rotation motion state prediction model based on the actual rotation motion state records of the radar platform in subsequent detection cycles, and using the corrected radar platform rotation motion state prediction model to re-predict the predicted rotation angular velocity and predicted rotation azimuth angle of the radar platform at the beginning of the next detection cycle includes: The actual rotational motion state record of the radar platform in subsequent detection cycles includes the sequence of actual rotational angular velocity measurements and the sequence of actual rotational azimuth measurements at multiple sampling moments in subsequent detection cycles. The azimuth angle value corresponding to the target point and its timestamp information during the acquisition process are extracted from the constant false alarm rate detection and processing of the radar echo data set. Combined with the actual rotation azimuth angle measurement value sequence, the theoretical azimuth angle prediction value of the radar platform beam pointing corresponding to each target point is calculated by time synchronization comparison. Calculate the azimuth prediction residual sequence between the theoretical azimuth prediction value of the radar platform beam pointing corresponding to each target point and the actual measurement value at the corresponding time in the actual rotation azimuth measurement value sequence; The azimuth prediction residual sequence is statistically analyzed and processed to calculate the mean drift and standard deviation fluctuation of the azimuth prediction residual sequence, which serve as quantitative indicators for evaluating the current prediction performance of the radar platform rotation motion state prediction model. Based on the mean drift and standard deviation fluctuation of the azimuth prediction residual sequence, the angular velocity compensation correction amount used to correct the angular velocity prediction value of the next cycle is derived. Based on the mean drift and standard deviation fluctuation of the azimuth prediction residual sequence, the azimuth compensation correction amount used to correct the azimuth prediction value of the next period is derived. The angular velocity compensation correction and the azimuth angle compensation correction are input into the parameter adaptive update interface of the radar platform rotation motion state prediction model, and the state transition matrix parameters and observation matrix parameters inside the radar platform rotation motion state prediction model are incrementally adjusted according to the preset learning rate coefficient. Using the radar platform rotation motion state prediction model with adjusted parameters, the actual rotation angular velocity measurement value and the actual rotation azimuth angle measurement value of the radar platform at the end of the subsequent detection cycle are used as new initial state inputs to perform forward inference calculation of the radar platform rotation motion state prediction model. Obtain the median predicted angular velocity and azimuth angle at the start of the next detection cycle from the radar platform rotation motion state prediction model after parameter adjustment. The predicted intermediate value of the azimuth angle is fused with the azimuth angle compensation correction amount to obtain the predicted rotation azimuth angle value of the radar platform at the start of the next detection cycle. The intermediate value of the predicted azimuth angle is fused with the angular velocity compensation correction to obtain the final predicted rotational azimuth angle value of the radar platform at the start of the next detection cycle.

9. The rotating radar detection method based on intelligent beam scheduling according to claim 8, characterized in that, The fused and updated set of potential target threat airspace, together with the re-predicted rotational angular velocity and rotational azimuth values, generates an updated dynamic radar detection airspace partitioning scheme for the next detection cycle, including: The updated set of potential target threat airspace includes the updated threat level identifier and updated airspace boundary parameters for each sector airspace description block; Based on the predicted rotational angular velocity and the predicted rotational azimuth, the updated theoretical maximum airspace boundary that the radar platform can cover in the next detection cycle after completing a single rotational scan is recalculated. From the updated set of potential target threat airspace, extract the updated threat level identifiers of all sector airspace description blocks, and sort them in descending order according to the numerical value of the threat level identifiers to generate a threat level priority queue. For a preset number of sector-shaped airspace description blocks sorted in the threat level priority queue, the required initial azimuth interval width parameters of the sector-shaped airspace description block in the rotating coordinate system of the radar platform are calculated based on the updated airspace boundary parameters of the sector-shaped airspace description block. Based on the initial azimuth interval width requirement parameter and the predicted rotation angular velocity value, the minimum beam dwell time requirement parameter required to cover each sector spatial description block is initially estimated. The sum of the minimum beam dwell time requirement parameters is compared with the total time budget available for key detection by the radar platform in the next detection cycle. If the total requirement exceeds the total budget, the allocation time of the lower-ranked sector airspace description blocks is reduced in order of threat level priority queue until the total requirement meets the total budget constraint. Based on the beam dwell time parameter finally assigned to each sector airspace description block and the predicted rotation angular velocity value, the updated azimuth interval width parameter that each sector airspace description block is actually assigned to occupy in the radar platform rotation coordinate system is recalculated. The updated azimuth interval width parameters of all sector airspace description blocks and their corresponding updated threat level identifiers are integrated to generate an updated structured radar detection airspace dynamic partitioning scheme description file. The updated structured description file of the dynamic partitioning scheme for radar detection airspace is transformed into a data interface for the updated dynamic partitioning scheme for radar detection airspace in the next detection cycle, which can be directly called by the radar beam scheduling decision module in the next round.

10. A rotating radar detection system based on intelligent beam scheduling, characterized in that, include: processor; A machine-readable storage medium for storing machine-executable instructions of the processor; The processor is configured to execute the rotating radar detection method based on intelligent beam scheduling according to any one of claims 1 to 9 by executing the machine-executable instructions.

Images