Method, device and storage medium for processing radar altimeter data

Abstract

The application discloses a radar altimeter data processing method and device and a storage medium, and relates to the technical field of radar altimeter data processing. The method comprises the following steps: processing all echo signals and target detection to obtain potential target points; dividing M pulse echo data in a frame into multiple data segments with different accumulation lengths; calculating the coherent accumulation energy of each potential target point in each data segment; judging and filtering out non-coherent interference target points according to the change rule of the coherent accumulation energy of the potential target points among the data segments; sorting the remaining target points after filtering according to energy, and selecting the Q target points with the highest energy for tracking; comparing the energy of each target tracking chain in historical frames, and taking the distance value corresponding to the target tracking chain with the highest energy as the final output value of the radar altimeter. The application greatly suppresses data jumping from the data source, and solves the problem of lacking power supply fluctuation and electromagnetic interference suppression capability.

Description

TECHNICAL FIELD

[0001] The present application belongs to the technical field of radar altimeter data processing, and particularly relates to a radar altimeter data processing method, device and storage medium. BACKGROUND

[0002] Low-altitude aircraft, especially unmanned aerial vehicles, are crucial to precise and stable perception of flight height when performing tasks such as surveying and mapping, logistics, inspection and urban environment flight. As a core height-determining sensor, radar altimeter directly measures the relative height between the aircraft and the ground by transmitting and receiving electromagnetic waves to the ground, and its data accuracy is directly related to the stability of flight control and flight safety.

[0003] However, in actual application, the output data of the existing radar altimeter is easily disturbed by various internal and external uncertainty factors, resulting in problems such as jump in height-determining data and out-of-range errors. Once the erroneous height information is adopted by the flight control system, it is easy to cause the aircraft to misoperate, such as abnormal climbing or diving, and ultimately lead to loss of control or even crash of the aircraft. The current radar altimeter mainly faces the following challenges in the application process:

[0004] (1) The complex electromagnetic environment inside the aircraft constitutes a serious interference source.

[0005] The aircraft integrates various communication modules, sensors and power conversion devices, and the non-periodic strong electromagnetic interference generated by the motor during operation may cause the radar to misjudge the interference signal as a target point in the signal spectrum, resulting in abnormal mutation of the output height information. Although the radar hardware design usually takes certain electromagnetic protection measures, the interference spectrum, intensity and type of different aircrafts are greatly different, and a single hardware solution is difficult to effectively suppress all potential uncertainty interference.

[0006] (2) The unstable on-board power supply system of the aircraft introduces power supply interference.

[0007] The aircraft is powered by a battery, and when performing drastic attitude adjustment or sudden acceleration action, the power supply system will withstand large fluctuations, which may produce instantaneous under-voltage, over-voltage or power supply pulse. If these power supply interferences are not effectively suppressed, they will directly affect the normal operation of the radar hardware system, causing the output data to jump, and the existing radar altimeter generally lacks special detection and suppression capability for such power fluctuation interference at the data processing level.

[0008] (3) The complex and variable ground environment poses a severe test to radar height measurement.

[0009] When the aircraft is operating in uneven terrain or urban environment, there may be sudden high-rise buildings such as high-rise buildings and mountains on the ground. The radar altimeter transmits a beam with a certain width, and when flying over such areas, its beam may cover the ground and the building at the same time, causing the radar altimeter to detect between the true ground height and the building height, and the output data changes dramatically in a large range.

[0010] The Chinese patent document with the application number CN201610497842.1 discloses a data processing method for unmanned aerial vehicle radar altimeter, which mainly judges the output by comparing the size relationship of adjacent period height values and performing statistical classification. In the scenario of correct sudden change of ground height, the core means of this method relies on comparing adjacent frame target values, and the method is relatively single. Especially when facing long-duration fixed electromagnetic interference or power interference, since the interference signal exists stably in each frame of data, this method may incorrectly identify the interference point as an effective target point and output, which lacks reliability, causing the aircraft to fail to respond in time and causing collision risk.

[0011] (4) The echo characteristics (RCS) of the radar detection target have the problem of flickering and weak signal-to-noise ratio (SNR).

[0012] The radar cross section (RCS) of the ground target varies due to differences in size, shape, and material, and exhibits a flickering characteristic over time, resulting in unstable target echo energy received by the radar from frame to frame. At the same time, when the aircraft is flying at ultra-high range or the ground target reflection coefficient is too low, the target signal-to-noise ratio will suddenly or continuously weaken. If the existing technology simply relies on the echo energy size of a single frame or a small number of consecutive frames to distinguish between RCS flickering and signal-to-noise ratio weakening caused by ultra-range, it lacks effective historical management and trend analysis of target echo energy, and is prone to misjudgment of low-energy noise points as target points or weak targets as invalid signals, resulting in abnormal output of height information.

[0013] In summary, the existing radar altimeter technology has obvious shortcomings, mainly reflected in: a) Lack of effective suppression mechanism for power fluctuation interference at the data processing level; b) Difficulty in dealing with complex and variable uncertain electromagnetic interference on the aircraft platform; c) Lack of response capability to sudden changes in real ground environment height, prone to misjudgment; d) Lack of robust management method for target RCS flickering and signal-to-noise ratio weakening. Therefore, there is an urgent need for a radar altimeter data processing method and system that can comprehensively cope with the above-mentioned multiple interference sources and has stronger robustness and environmental adaptability. SUMMARY

[0014] The application aims to provide a radar altimeter data processing method, device and storage medium, and aims to solve the problems of height determination data jumping and over-range caused by various internal and external uncertainties of the radar altimeter in a low-altitude aircraft.

[0015] The application solves the above technical problems by the following technical scheme: a radar altimeter data processing method, comprising:

[0016] Receiving echo signals corresponding to each transmitted linear modulation pulse;

[0017] Processing and target detection are performed on all echo signals to obtain potential target points; wherein all echo signals refer to echo signals received by the radar altimeter when a frame of modulation signals composed of M linear modulation pulses is sequentially transmitted;

[0018] M pulse echo data in a frame are divided into data segments of different accumulation lengths; wherein single pulse echo data refers to digital signal data obtained by mixing and sampling echo signals corresponding to a single linear modulation pulse;

[0019] For each potential target point, the coherent accumulation energy of the potential target point in each data segment is calculated;

[0020] According to the change rule of the coherent accumulation energy of the potential target point between the data segments, non-coherent interference target points are discriminated and filtered out;

[0021] The remaining target points after filtering are sorted according to energy, and the Q target points with the highest energy are selected for tracking to form Q target tracking chains;

[0022] The energy of each target tracking chain in the historical frame is compared, and the distance value corresponding to the target tracking chain with the highest energy is taken as the final output value of the radar altimeter.

[0023] In this embodiment, based on the essential characteristics that the echo of the real target is homologous to the transmitted signal and the interference is not homologous to the transmitted signal, by dividing data segments of different accumulation lengths and checking whether the target point energy conforms to the coherent accumulation gain rule, all interferences not conforming to the coherent accumulation gain rule are fundamentally recognized and filtered out. No matter whether these interferences come from the electromagnetic environment or power fluctuations, as long as the non-coherent nature is unchanged, they can be effectively eliminated, and the data jumping problem caused by electromagnetic interference and power interference is directly solved from the data source.

[0024] The application simultaneously tracks Q targets with the highest energy, establishes and maintains an independent target tracking chain for each target, and compares the energy of each target tracking chain in historical multiple frames. When the terrain changes, the energy of the high-rise target tracking chain will remain stable at a high level, and after multiple frame accumulation, the historical average energy will exceed that of the flat ground target chain, so the output is switched to the target chain representing a higher height. This method can quickly respond to real environment changes and avoid false switching caused by single frame flickering, solving the problem of insufficient response to real ground environment height changes.

[0025] When the real ground target temporarily decreases in energy due to RCS flickering, since the application is based on the historical energy of the target tracking chain, the average energy is still likely to remain the highest, thereby maintaining the output, and the target will not be lost due to temporary energy decline. This greatly enhances the tracking robustness of the flickering target and effectively distinguishes between temporary RCS flickering and permanent signal-to-noise ratio disappearance.

[0026] Further, all echo signals are processed and target detection, specifically including:

[0027] Mixing the echo signal corresponding to the linear modulation pulse being transmitted at the current time and the linear modulation pulse having been transmitted at the previous time, to obtain an analog intermediate frequency signal;

[0028] Sampling the analog intermediate frequency signal within the duration of the linear modulation pulse being transmitted at the current time, to obtain K sampling points;

[0029] Arranging all the sampling points collected in a frame of modulation signal in order to form a KxM original data matrix;

[0030] Performing two-dimensional fast Fourier transform on the original data matrix to obtain a KxM distance-Doppler matrix;

[0031] Performing target detection on the distance-Doppler matrix to obtain the potential target point.

[0032] In this embodiment, the original analog echo signal is converted into a digitized matrix containing target distance and velocity information, and the potential target point is preliminarily detected, providing input for subsequent processing.

[0033] Further, the coherent accumulation energy of the potential target point in each data segment is calculated, specifically including:

[0034] According to the position of the potential target point in the distance-Doppler matrix, the corresponding distance gate is determined;

[0035] For each data segment, the complex signal value corresponding to the distance gate where the potential target point is located is extracted from the data segment;

[0036] vector adding all complex signal values in the data segment;

[0037] calculating square of modulus of the complex vector adding result, taking the calculation result as coherent accumulation energy of the potential target point in the data segment.

[0038] Further, according to variation rule of coherent accumulation energy of the potential target point between data segments, non-coherent interference target points are distinguished and filtered out, specifically including:

[0039] calculating coherent accumulation energy gain of the potential target point between data segments with longer accumulation length and data segments with shorter accumulation length;

[0040] comparing the coherent accumulation energy gain with a preset threshold interval determined based on coherent accumulation gain theory;

[0041] if the coherent accumulation energy gain meets requirement of the preset threshold interval, the potential target point is determined as coherent signals of the same origin and is reserved, otherwise, the potential target point is determined as non-coherent interference signals and is filtered out.

[0042] Further, the number of the data segments is at least 3, and accumulation lengths are in multiple relationship; the step of distinguishing and filtering out includes:

[0043] calculating coherent accumulation energy gain of the potential target point between every two adjacent data segments with accumulation lengths in multiple relationship, to obtain multiple coherent accumulation energy gains;

[0044] counting number of the multiple coherent accumulation energy gains falling into the preset threshold interval;

[0045] if the counted number reaches or exceeds a preset qualified number threshold, the potential target point is determined as coherent signals of the same origin and is reserved; otherwise, the potential target point is determined as non-coherent interference signals and is filtered out.

[0046] Further, nearest neighbor algorithm is used to perform cross-frame track clustering on the Q target points, to form Q target tracking chains, specifically including:

[0047] a target tracking chain is established and maintained for each tracked target, to record state information of the target in continuous multiple frames;

[0048] according to state information of each target tracking chain in a previous frame, a predicted state of the target tracking chain in a current frame is predicted;

[0049] association cost between state information of the Q targets in the current frame and the predicted states of the Q target tracking chains is calculated;

[0050] For each target tracking chain, find the target point with the minimum association cost to its predicted state from the Q targets in the current frame;

[0051] Update the state information of the successfully associated target point into the corresponding target tracking chain, forming the target tracking chain of the current frame.

[0052] In this embodiment, the nearest neighbor algorithm performs cross-frame point cluster to form a target chain, upgrading the data processing from static and passive screening to dynamic and active tracking, greatly enhancing the robustness to instantaneous interference and target RCS flicker, effectively suppressing high data jumps, and providing smooth and stable height output.

[0053] Further, the method further comprises a target attention mechanism, which specifically comprises the following steps:

[0054] Real-time monitoring of the signal-to-noise ratio of the target tracking chain corresponding to the final output value;

[0055] When the signal-to-noise ratio is lower than the first preset threshold, triggering the target attention mechanism;

[0056] If the target attention mechanism is triggered within p consecutive frames, it is determined that the effective detection range of the radar altimeter has been exceeded;

[0057] When it is determined that the effective detection range is exceeded, the confidence of the radar altimeter output height value is set to zero, and the output is kept at the valid height value of the last frame before the target attention mechanism is triggered.

[0058] In this embodiment, the target attention mechanism is used to actively monitor whether the radar altimeter reaches the detection range, and actively admit failure and alarm when it reaches the detection range, avoiding the guided aircraft from making wrong decisions; the continuous p-frame judgment avoids false positives caused by instantaneous fluctuations in the signal; and the data retention strategy provides the flight control system with critical decision-making and switching time.

[0059] Further, the method further comprises an environment detection and determination step, specifically comprising:

[0060] When the target tracking chain corresponding to the final output value has a number mutation, triggering the environment detection;

[0061] Calculate the root mean square average of the energy of all FFT points in the distance dimension of the current frame;

[0062] Calculate the difference between the root mean square average and the historical background energy;

[0063] If the difference exceeds the second preset threshold, it is determined that there is a real environment change, and a new height value is confirmed and output; if the difference does not exceed the second preset threshold, it is determined that there is an error jump, and the new height value is rejected.

[0064] The historical background energy is a statistical value of the root mean square average of the most recent L frames before the mutation occurs.

[0065] In the embodiment, through cross-validation of local mutation and global background, real environment mutation and error jump caused by interference are accurately distinguished, and then timely and correct response to ground environment change is made, correct decision of the aircraft is guided, and collision risk is effectively avoided.

[0066] Based on the same concept, the application also provides an electronic device, comprising a memory, a processor and a computer program or instructions stored in the memory, wherein the processor executes the computer program or instructions to realize the radar altimeter data processing method as described above.

[0067] Based on the same concept, the application also provides a computer readable storage medium, which stores a computer program or instructions, and the computer program or instructions are executed by a processor to realize the radar altimeter data processing method as described above.

[0068] Compared with the prior art, the application has the beneficial effects that:

[0069] The method of the application does not rely on threshold judgment of interference intensity, but discriminates and filters by checking whether the growth rule of signal energy with the increase of accumulated pulse number conforms to coherent accumulation theory, can uniformly and effectively filter various internal (such as power supply fluctuation) and external (such as electromagnetic) interferences of different sources from radar transmitted signals, greatly suppresses data jump from data source, and solves the problem of lack of power supply fluctuation and electromagnetic interference suppression capability.

[0070] The application upgrades data processing from static and single-frame point screening to dynamic and cross-frame target tracking chain, improves the tolerance capability to target RCS flicker and temporary interference, and improves the smoothness and stability of output height value.

[0071] The application proposes a target attention mechanism, which can intelligently distinguish transient signal fading and permanent signal loss (such as over-range) by monitoring signal-to-noise ratio and introducing delay decision logic; after confirming failure, through the strategy of "zero confidence + height value retention", a "failure-safety" degradation mode is realized. This not only prevents random noise data output at the performance boundary from causing aircraft misoperation, but also wins valuable buffer time for the flight control system to switch to the backup sensor, greatly enhancing the overall safety of the system.

[0072] The application adopts the energy root mean square average method to detect and determine the environment, ignores a few isolated peak values, captures the macro characteristics of the background reflection intensity of the entire detection scene, provides a reliable basis for determining whether the environment has undergone a fundamental change, solves the problem of distinguishing between real environment mutation and false data jump, and ensures the accuracy and intelligence of decision-making in a complex environment. BRIEF DESCRIPTION OF DRAWINGS

[0073] In order to more clearly illustrate the technical solutions of the present application, the following will briefly introduce the drawings needed to be used in the embodiments. Obviously, the drawings described below are only one embodiment of the present application, and other drawings can be obtained by those skilled in the art without creative labor.

[0074] Figure 1 is a flow chart of the radar altimeter data processing method in the embodiment of the present application;

[0075] Figure 2 is a schematic diagram of the frequency modulation continuous wave radar waveform modulation in the embodiment of the present application;

[0076] Figure 3 is an angular reflection target distance dimension spectrum diagram of the radar altimeter in the strong reflection body background in the embodiment of the present application;

[0077] Figure 4 is an angular reflection target distance dimension spectrum diagram of the radar altimeter in the weak reflection body background in the embodiment of the present application. DETAILED DESCRIPTION

[0078] The technical solutions in the present application will be described clearly and completely in combination with the drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, not all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor are within the scope of protection of the present application.

[0079] The technical solutions of the present application will be described in detail in combination with specific embodiments. The following specific embodiments can be combined with each other, and the same or similar concepts or processes may not be described in some embodiments.

[0080] Embodiment one

[0081] As shown in the figure, the radar altimeter data processing method provided by the embodiment includes the following steps: Figure 1

[0082] Step 1: receiving the echo signal corresponding to each transmitted linear modulation pulse.

[0083] The radar altimeter transmits a linear modulation pulse, and receives the echo signal corresponding to the linear modulation pulse.​Figure 2 The modulation method generates a frame of modulated signal. For example... Figure 2 As shown, a single frame of modulated signal consists of M linear modulation pulses (i.e., M chirp signals). The fast time corresponds to the duration of a single chirp signal, and the slow time corresponds to the duration of a frame of modulated signal. For example, a frame of modulated signal consists of 128 linear modulation pulses.

[0084] Radar altimeter according to Figure 2 M linear modulation pulses are transmitted sequentially, and the echo signals corresponding to each transmitted linear modulation pulse are received sequentially to obtain all the echo signals of the current frame, that is, the echo signals corresponding to the M transmitted linear modulation pulses.

[0085] Step 2: Process all echo signals and perform target detection to obtain potential target points.

[0086] In a specific embodiment of the present invention, processing and target detection of all echo signals specifically includes:

[0087] Step 2.1: Mix the echo signal corresponding to the linear modulation pulse being transmitted at the current moment with the linear modulation pulse transmitted at the previous moment to obtain the analog intermediate frequency signal;

[0088] Step 2.2: During the duration of the linear modulation pulse being transmitted at the current moment (i.e., one fast time), use an analog-to-digital converter to sample the analog intermediate frequency signal to obtain K sampling points (e.g., 256 points are collected in one fast time).

[0089] Step 2.3: Arrange all the sampling points acquired within a frame of modulated signal in order to form a K×M original data matrix;

[0090] Step 2.4: Perform a two-dimensional fast Fourier transform on the original data matrix to obtain a K×M dimensional distance-Doppler matrix;

[0091] Step 2.5: Perform target detection on the range-Doppler matrix to obtain all potential target points, and number all potential target points.

[0092] In this embodiment, performing a two-dimensional fast Fourier transform on the original data matrix includes:

[0093] Perform a Fast Fourier Transform (FFT) on each column (i.e., each independent Chirp) of the original data matrix to obtain an intermediate matrix;

[0094] Perform a Fast Fourier Transform (FFT) on each row of the intermediate matrix (i.e., each distance gate) to obtain the distance-Doppler matrix.

[0095] The distance dimension FFT transforms the signal from the fast time domain to the range frequency domain, and each point (i.e., a range gate) after the FFT corresponds to a specific range cell, and the amplitude of the point reflects whether there is a target at the range and the signal strength thereof. The velocity dimension FFT transforms the signal from the slow time domain to the Doppler frequency domain, and each point after the FFT corresponds to a specific velocity cell, and the amplitude of the point reflects whether there is a target at the velocity and range combination.

[0096] In step 2.5, a cell average constant false alarm rate (CA-CFA) detection method is used to detect targets in the range-Doppler matrix, and the detection threshold is set to 5 dB. The cell average constant false alarm rate detection method is a constant false alarm rate detection method with the background noise power as the threshold. Since the threshold is the background noise power, a constant false alarm probability can be ensured. Setting the threshold to a low 5 dB means that the detection sensitivity is high, ensuring that as many real targets (even weak targets) as possible can be initially detected, and then entering the subsequent judgment process, thereby improving the detection capability of the radar for weak targets.

[0097] Step 3: Divide the M pulse echo data in a frame into data segments of different accumulation lengths.

[0098] The single pulse echo data refers to the digital signal data obtained after mixing and sampling the echo signal corresponding to a single linear modulation pulse, i.e., the data obtained in step 2.3. By segmenting, an accumulation length gradient is created to observe the growth rule of signal energy with the increase of the number of accumulated pulses, which is the basis for subsequent discrimination.

[0099] For example, when M = 128, the 128 pulse echo data are sequentially divided into data segments with accumulation lengths of 16, 32, 64, and 128. The data segment A with an accumulation length of 16 includes the first 16 pulse echo data, the data segment B with an accumulation length of 32 includes the first 32 pulse echo data, the data segment C with an accumulation length of 64 includes the first 64 pulse echo data, and the data segment D with an accumulation length of 128 includes all the pulse echo data.

[0100] Step 4: For each potential target point, calculate the coherent accumulation energy thereof in each data segment.

[0101] In the specific embodiment of the present application, for each potential target point, the coherent accumulation energy thereof in each data segment is calculated, specifically including:

[0102] Step 4.1: According to the position of the potential target point in the range-Doppler matrix, determine the corresponding range gate thereof;

[0103] Step 4.2: For each data segment, extract the complex signal value (including amplitude and phase information) corresponding to the distance gate where the potential target point is located from the data segment;

[0104] Step 4.3: Perform complex vector addition on all complex signal values in the data segment, i.e. coherent accumulation.

[0105] Step 4.4: Calculate the square of the modulus of the complex vector addition result, and use the calculation result as the coherent accumulation energy of the potential target point in the data segment.

[0106] Step 5: According to the variation law of the coherent accumulation energy of the potential target point between data segments, discriminate and filter out non-coherent interference target points.

[0107] In the specific embodiment of the present application, according to the variation law of the coherent accumulation energy of the potential target point between data segments, discrimination and filtering out of non-coherent interference target points are specifically included:

[0108] Step 5.1: Calculate the coherent accumulation energy gain of the potential target point between the data segment with longer accumulation length and the data segment with shorter accumulation length.

[0109] Step 5.2: Compare the coherent accumulation energy gain with the preset threshold interval determined based on the coherent accumulation gain theory.

[0110] If the coherent accumulation energy gain is within the preset threshold interval, it is determined that the potential target point is a coherent signal of the same origin and is retained, otherwise it is determined to be a non-coherent interference signal and is filtered out.

[0111] In the specific embodiment of the present application, the number of data segments is at least 3, and the accumulation lengths are in a multiple relationship; the discrimination and filtering out step includes:

[0112] Calculate the coherent accumulation energy gain of the potential target point between each two adjacent data segments with accumulation lengths in a multiple relationship, to obtain a plurality of coherent accumulation energy gains; count the number of coherent accumulation energy gains falling within the preset threshold interval; if the counted number reaches or exceeds a preset qualified number threshold, it is determined that the potential target point is a coherent signal of the same origin and is retained; otherwise, it is determined that the potential target point is a non-coherent interference signal and is filtered out.

[0113] In this embodiment, the preset threshold interval is [3-δ, 3+δ] dB, δ represents a tolerance value, which can be set to 0.5 dB or 1 dB. The qualified number threshold is the number of coherent accumulation energy gains, i.e. all coherent accumulation energy gains fall within the preset threshold interval, then it is determined that the corresponding potential target point is a coherent signal of the same origin and is retained.

[0114] The coherent accumulation gain formula is: G d =10logN, wherein, Gd represents the coherent accumulation gain, and N represents the number of accumulated pulses.

[0115] For example, for the aforementioned four divided data segments: data segment A with an accumulation length of 16, data segment B with an accumulation length of 32, data segment C with an accumulation length of 64, and data segment D with an accumulation length of 128, we have:

[0116] From data segment A to data segment B, the number of pulses doubles, and the theoretical coherent accumulation energy gain is 10log(32) - 10log(16) ≈ 3dB, so the coherent accumulation energy of data segment B should be about 3dB higher than that of data segment A.

[0117] From data segment B to data segment C, the number of pulses doubles, and the theoretical coherent accumulation energy gain is 10log(64) - 10log(32) ≈ 3dB, so the coherent accumulation energy of data segment C should be about 3dB higher than that of data segment B.

[0118] From data segment C to data segment D, the number of pulses doubles, and the theoretical coherent accumulation energy gain is 10log(128) - 10log(64) ≈ 3dB, so the coherent accumulation energy of data segment D should be about 3dB higher than that of data segment C.

[0119] It can be seen that the energy value of the real target (homogeneous coherent signal) will increase in a stable, approximately 3dB per doubling step, with the increase in the number of accumulated pulses.

[0120] The phase of the interference signal is random, different from the radar transmission signal. Therefore, the accumulation across Chirp is non-coherent accumulation (energy addition, not vector addition). When non-coherent accumulation, the signal power and noise power are linearly superimposed, the signal power becomes 2 times (increases by 3dB), and the noise power also becomes 2 times (increases by 3dB). Therefore, the signal-to-noise ratio remains unchanged. That is, the ratio of the coherent accumulation energy of data segment B to the coherent accumulation energy of data segment A, the ratio of the coherent accumulation energy of data segment C to the coherent accumulation energy of data segment B, and the ratio of the coherent accumulation energy of data segment D to the coherent accumulation energy of data segment C are all close to 1 (0dB gain), and there is no stable 3dB gain step. Its energy curve will be a relatively flat, irregular growth line.

[0121] Therefore, for each potential target point, checking the growth relationship between its coherent accumulation energy in data segment A, data segment B, data segment C, and data segment D can determine whether the potential target point is an interference signal.

[0122] The present application can accurately distinguish real targets from interference by whether the energy of each potential target point on the spectrum on the multiple accumulated pulses has a 3dB change, without the need for complex calculation logic to support.

[0123] Step 6: The target points remaining after filtering are sorted by energy, and the Q target points with the highest energy are selected for tracking to form Q target tracking chains.

[0124] The target points remaining after filtering are sorted by energy, where the energy refers to the signal power of the distance-velocity cell corresponding to the target point in the distance-Doppler matrix. The Q target points (for example, 3 target points) are numbered, and the energy value of each target in each frame of data after starting tracking is recorded. The nearest neighbor algorithm is used to cluster the Q target points across frames to form Q target tracking chains. In the specific embodiments of the present application, the nearest neighbor algorithm clusters the Q target points across frames, specifically including:

[0125] Step 6.1: A target tracking chain is established and maintained for each tracked target to record the state information of the target in consecutive multiple frames;

[0126] Step 6.2: The predicted state of each target tracking chain in the current frame is predicted according to the state information of the target tracking chain in the previous frame;

[0127] Step 6.3: The association cost between the state information of the Q targets in the current frame and the predicted state of the Q target tracking chains is calculated;

[0128] Step 6.4: For each target tracking chain, find the target point with the minimum association cost with the predicted state from the Q targets in the current frame to associate;

[0129] Step 6.5: The state information of the successfully associated target point is updated to the corresponding target tracking chain to form the target tracking chain of the current frame.

[0130] In this embodiment, the predicted state of the target tracking chain in the current frame can be realized by using existing methods such as Kalman filter. The nearest neighbor algorithm is used to cluster the data across frames to form a target chain, which upgrades the data processing from static and passive screening to dynamic and active tracking, greatly enhances the robustness to instantaneous interference and target RCS flicker, effectively suppresses the high data jump, and provides smooth and stable height output.

[0131] Step 7: Compare the energy of each target tracking chain in the historical frame, and take the distance value corresponding to the target tracking chain with the highest energy as the final output value of the radar altimeter.

[0132] In the specific embodiments of the present application, the processing method further includes a target attention mechanism, and the specific steps include:

[0133] Step 7.1: Real-time monitoring of the signal-to-noise ratio of the target tracking chain corresponding to the final output value;

[0134] Step 7.2: Triggering the target attention mechanism when the signal-to-noise ratio is lower than a first preset threshold (e.g., 10 dB);

[0135] Step 7.3: If the target attention mechanism is triggered in consecutive p frames (e.g., 2 frames), it is determined that the effective detection range of the radar altimeter has been exceeded;

[0136] Step 7.4: When it is determined that the effective detection range has been exceeded, the confidence of the radar altimeter output height value is set to zero, and the output is kept at the valid height value of the last frame before the target attention mechanism is triggered.

[0137] The target attention mechanism is similar to a state flag. When consecutive p frames (e.g., 2 frames) trigger the target attention mechanism, it indicates that attention needs to be paid to the validity of the radar altimeter output height value. At this time, the confidence of the radar altimeter output height value is set to zero, and the valid height value of the last frame before the target attention mechanism is triggered is kept.

[0138] The radar altimeter data processing of the present application is performed in consecutive frames, and real-time monitoring is performed on whether the number of the target tracking chain corresponding to the final output value has a mutation. The target tracking chain corresponding to the final output value is one of Q target tracking chains. The number of the target tracking chain corresponding to the final output value has a mutation, which means that in consecutive frames, the number of the target tracking chain that is the final output has changed, and another target tracking chain in the Q target tracking chains outputs a height value. There are two reasons for this mutation: one is a real and correct environmental change, and the other is an error and dangerous data jump. Therefore, it is necessary to determine the real reason for the mutation and perform environmental detection and determination. In the specific embodiments of the present application, environmental detection and determination specifically includes:

[0139] Triggering environmental detection when the number of the target tracking chain corresponding to the final output value has a mutation;

[0140] Calculating the root mean square average of the energy of all FFT points in the distance dimension of the current frame;

[0141] Calculating the difference between the root mean square average and the historical background energy;

[0142] If the difference exceeds a second preset threshold, it is determined that there is a real environmental change, and a new height value is confirmed and output; if the difference does not exceed the second preset threshold, it is determined that there is an error jump, and the new height value is rejected.

[0143] In step 2, when K is 256 and M is 128, the size of the range-Doppler matrix is 256x128, and the number of range dimension FFT points is 256. After performing range dimension FFT on the echo signal of one Chirp signal of the radar, the signal power corresponding to each range gate in the spectrum is the FFT point energy. The specific formula of the root mean square average of all range dimension FFT point energies in the current frame is: wherein, represents the root mean square average of all range dimension FFT point energies; K represents the number of range dimension FFT points; , , represents the energy value of each range gate.

[0144] The root mean square average has the advantage of low sensitivity to sudden outliers, and is more suitable for evaluating the overall characteristics of the data, and has a good inhibitory effect on the value of the interference pulse. By calculating the root mean square average of all range dimension FFT point energies, a few isolated peaks are ignored, and the macroscopic characteristics of the background reflection intensity of the entire detection scene are captured, thereby providing a reliable basis for judging whether the environment has changed fundamentally.

[0145] The historical background energy is a reference value or a benchmark value, and the historical background energy represents a representative overall energy level of the reflection environment of the radar before the number mutation of the target tracking chain occurs. In the embodiment, the historical background energy can be set as the root mean square average of the last frame before the mutation occurs , or the average or median of the root mean square average of the last L frames before the mutation occurs . In the embodiment, the value of the second preset threshold is in the range of 5dB to 15dB.

[0146] Figure 3 and Figure 4 respectively show the range dimension spectrum of the corner reflection target under the strong reflector background and the weak reflector background, wherein mark 1 is the corner reflector target, mark 2 is the environmental noise floor, Y is the distance of the target (unit: meter), and Z is the energy value of the target (unit: dB). From the positions of mark 1 in Figure 3 and Figure 4 , it can be known that the distances of the two targets from the radar altimeter are consistent, and from mark 2, it can be known that the noise floors in the two environments differ by about 12dB.

[0147] The root mean square average reflects the overall noise performance, and the energy root mean square average in the strong reflector background can better reflect the change of the radar altimeter background, for example, the target detected by the radar altimeter suddenly changes from a land background to a high-rise building, and through the environment detection and judgment of the application, the reflection background of the radar altimeter can be accurately judged, the real environment mutation and the error jump caused by interference can be accurately distinguished, and then the ground environment change can be responded in time and correctly.

[0148] Embodiment two

[0149] The embodiment of the application also provides an electronic device, which comprises a memory, a processor and a computer program or instructions stored in the memory, and the processor executes the computer program or instructions to realize the radar altimeter data processing method in the embodiment one of the application.

[0150] Although not shown, the electronic device comprises a processor, which can perform various appropriate operations and processes according to programs and / or data stored in a read-only memory (ROM) or loaded from a storage part into a random access memory (RAM). The processor can be a multi-core processor, or can comprise a plurality of processors. In some embodiments, the processor can comprise a general-purpose main processor and one or more special-purpose coprocessors, such as a central processing unit, a graphics processing unit (GPU), a neural network processing unit (NPU), a digital signal processor (DSP), etc. In the RAM, various programs and data required for device operation are also stored. The processor, the ROM and the RAM are connected to each other through a bus. An input / output (I / O) interface is also connected to the bus.

[0151] The above processor and memory are used together to execute programs / instructions stored in the memory, and the programs / instructions are executed by a computer to realize the methods, steps or functions described in the above embodiments.

[0152] Although not shown, the embodiment of the application also provides a computer readable storage medium, which stores a computer program or instructions, and the computer program or instructions are executed by a processor to realize the radar altimeter data processing method in the embodiment one of the application.

[0153] Read only memory (ROM), electrically programmable read only memory (EPROM), electrically erasable programmable read only memory (EEPROM), flash memory or other memory technologies, compact disc read only memory (CD-ROM), digital versatile discs (DVD) or other optical storage, magnetic cassettes, magnetic tape, magnetic disk storage or other magnetic storage devices, or any other medium which can be used to store information which can be accessed by a computing device. According to the definition used herein, computer readable medium does not include transitory media, such as modulated data signals and carrier waves.

[0154] The above disclosure is merely a specific implementation of the present application, but the protection scope of the present application is not limited thereto, and any person skilled in the art can easily think of changes or modifications within the technical scope disclosed by the present application, which should be covered within the protection scope of the present application.

Claims

1. A radar altimeter data processing method, characterized in that, The processing method includes: Receive the echo signal corresponding to each transmitted linearly modulated pulse; All echo signals are processed and targets are detected to obtain potential target points; where all echo signals refer to the echo signals received when the radar altimeter sequentially transmits a frame of modulated signals consisting of M linearly modulated pulses. The M pulse echo data within a frame are divided into multiple data segments with different accumulation lengths; where, single pulse echo data refers to the digital signal data obtained by mixing and sampling the echo signal corresponding to a single linear modulation pulse. For each potential target point, calculate its coherent cumulative energy within each data segment; Based on the variation pattern of coherent accumulated energy of potential target points between data segments, identify and filter out incoherent interfering target points; The remaining target points after filtering are sorted by energy, and the Q target points with the highest energy are selected for tracking, forming Q target tracking chains; By comparing the energy of each target tracking chain in historical frames, the distance value corresponding to the target tracking chain with the highest energy is taken as the final output value of the radar altimeter.

2. The radar altimeter data processing method according to claim 1, characterized in that, All echo signals are processed and targets are detected, specifically including: The echo signal corresponding to the linear modulation pulse being transmitted at the current moment is mixed with the linear modulation pulse transmitted at the previous moment to obtain an analog intermediate frequency signal. During the duration of the linear modulation pulse being transmitted at the current moment, the analog intermediate frequency signal is sampled to obtain K sampling points; All sampling points acquired within a frame of modulated signal are arranged in order to form a K×M original data matrix; A two-dimensional fast Fourier transform is performed on the original data matrix to obtain a K×M dimensional distance-Doppler matrix; Target detection is performed on the distance-Doppler matrix to obtain the potential target points.

3. The radar altimeter data processing method according to claim 1, characterized in that, Calculating the coherent accumulation energy of potential target points in each data segment specifically includes: Determine the corresponding range gate based on the position of the potential target point in the range-Doppler matrix; For each data segment, extract the complex signal value corresponding to the distance gate of the potential target point from the data segment; Perform complex vector summation on all complex signal values ​​within the data segment; The square of the modulus of the sum of complex vectors is calculated, and this result is used as the coherent accumulated energy of the potential target point within the data segment.

4. The radar altimeter data processing method according to claim 1, characterized in that, Based on the variation pattern of coherent accumulated energy of potential target points across data segments, incoherent interfering target points are identified and filtered out, specifically including: Calculate the coherent accumulation energy gain of potential target points between data segments with longer accumulation lengths and data segments with shorter accumulation lengths; The coherent accumulation energy gain is compared with a preset threshold range determined based on the coherent accumulation gain theory; If the coherent accumulated energy gain meets the requirements of the preset threshold range, the potential target point is determined to be a coherent signal from the same source and is retained; otherwise, it is determined to be an incoherent interference signal and is filtered out.

5. The radar altimeter data processing method according to claim 4, characterized in that, The number of data segments is at least three, and the accumulated lengths are multiples of each other; the discrimination and filtering steps include: Calculate the coherent accumulation energy gain between adjacent data segments where the accumulation length is a multiple of each other for the potential target point, and obtain multiple coherent accumulation energy gains; Count the number of coherent accumulated energy gains that fall within a preset threshold range; If the number of counts reaches or exceeds the preset qualified quantity threshold, the potential target point is determined to be a coherent signal from the same source and is retained; otherwise, the potential target point is determined to be an incoherent interference signal and is filtered out.

6. The radar altimeter data processing method according to claim 1, characterized in that, The nearest neighbor algorithm is used to cluster Q target points across frames, forming Q target tracking chains, specifically including: A target tracking chain is established and maintained for each tracked target to record the target's state information in multiple consecutive frames; Based on the state information of each target tracking chain in the previous frame, predict its predicted state in the current frame. Calculate the association cost between the state information of Q targets in the current frame and the predicted state of Q target tracking chains; For each target tracking chain, find the target point with the minimum association cost with its predicted state from the Q targets in the current frame and associate it with it; The status information of successfully associated target points is updated to the corresponding target tracking chain to form the target tracking chain for the current frame.

7. The radar altimeter data processing method according to any one of claims 1 to 6, characterized in that, The method also includes a target attention mechanism, the specific steps of which include: Real-time monitoring of the signal-to-noise ratio of the target tracking chain corresponding to the final output value; When the signal-to-noise ratio is lower than a first preset threshold, the target attention mechanism is triggered; If the target attention mechanism is triggered within p consecutive frames, it is determined that the effective detection range of the radar altimeter has been exceeded. When it is determined that the target is outside the effective detection range, the confidence level of the height value output by the radar altimeter is set to zero, and the output remains the effective height value of the last frame before the target attention mechanism is triggered.

8. The radar altimeter data processing method according to any one of claims 1 to 6, characterized in that, The method also includes an environmental detection and judgment step, specifically including: When the target tracking chain corresponding to the final output value undergoes a sudden numbering change, environmental detection is triggered. Calculate the root mean square average of the energy of all FFT points in the distance dimension of the current frame; Calculate the difference between the root mean square value and the historical background energy; If the difference exceeds the second preset threshold, it is determined to be a real environmental change, and a new height value is confirmed and output; if the difference does not exceed the second preset threshold, it is determined to be an erroneous jump, and the new height value is rejected. The historical background energy is the statistical value of the root mean square average of the most recent L frames before the mutation occurred.

9. An electronic device comprising a memory, a processor, and a computer program or instructions stored in the memory, characterized in that, The processor executes the computer program or instructions to implement the radar altimeter data processing method as described in any one of claims 1 to 8.

10. A computer-readable storage medium having a computer program or instructions stored thereon, characterized in that, When the computer program or instructions are executed by the processor, they implement the radar altimeter data processing method as described in any one of claims 1 to 8.

Images