A multi-spectrum illumination based automobile millimeter wave radar profile imaging method and system

Abstract

The application belongs to the technical field of radar imaging, and discloses a kind of automobile millimeter wave radar profile imaging method and system based on multispectral illumination, comprising: S1: transmit N pulse Pulse with pulse interval frequency interval Δf;S2: set minimum processing pulse number N m , N m It can be set to about half of N pulse ;S3: set pulse number increment Δn, if not considering calculation amount, Δn can be set to 1, Δn can also be set between 1 to 10;S4: multispectral illumination number initialization n d =1, start from first time;S5: to the 1st to N m +(n d -1)×Δn pulse is carried out phase correlation processing, and the obtained pulse echo data matrix is carried out two-dimensional FFT to obtain high resolution RD spectrum;S6: obtain high resolution wideband synthesis bandwidth (N m +(n d -1)×Δn)·Δf of synthesis result;S7: judge whether pulse number is processed, namely N m +(n d -1)×Δn<N pulse , when the formula is established, jump to the fifth step;If not, then next step;S8: the target imaging result of each spectrum illumination obtained by S6 is superimposed to obtain target multispectral illumination imaging result.

Description

Technical Field

[0001] This invention belongs to the field of radar imaging technology, and particularly relates to a method and system for contour imaging of automotive millimeter-wave radar based on multispectral illumination. It can significantly improve the quantity and quality of point cloud output from millimeter-wave automotive radar, laying a very good foundation for sensor detection in autonomous vehicles. Background Technology

[0002] Automotive millimeter-wave radar sensors can meet the requirements of vehicles' adaptability to all weather conditions. Millimeter-wave radar has already played, or will play, an increasingly important role in assisted driving and future autonomous driving, offering significant advantages over other sensors. In particular, the 4D millimeter-wave radar introduced in recent years can provide target height information and richer scene information, giving it the potential for long-term sustainable development.

[0003] However, the biggest drawback of millimeter-wave radar is its inability to image the contours of targets. LiDAR, through beam scanning, can detect the boundaries of objects in a scene, accurately determining the precise location of targets. However, even the most advanced millimeter-wave radar products currently available internationally cannot fully meet the demands of autonomous driving, primarily because millimeter-wave radar cannot provide sufficiently high angular resolution. This is the only inherent limitation of millimeter-wave radar, despite its numerous advantages such as high range resolution, good environmental adaptability, long detection range, and high velocity measurement accuracy.

[0004] Therefore, the inability of automotive millimeter-wave radar to perform contour imaging of targets and its insufficient angular resolution to support high-quality point cloud generation are the biggest obstacles to its current application. In fact, these shortcomings stem from the design of millimeter-wave radar itself. The angular resolution of automotive millimeter-wave radar is the same as that of general radar, and the direct influencing factor on angular resolution is the array aperture or array size. A larger array results in a narrower beam, leading to better angular resolution; conversely, a smaller array results in a wider beam, making it difficult to distinguish targets within that beam angularly, resulting in poorer angular resolution. Because the installation location of automotive radar is very limited, constraining the aperture size of millimeter-wave radar, and considering cost factors, a smaller array size cannot achieve the angular resolution of lidar.

[0005] In the field of automotive active safety, millimeter-wave radar sensors have become the industry's recognized mainstream choice due to their ability to operate in all weather conditions, unaffected by light, fog, sandstorms, or other adverse weather conditions. This has resulted in huge market demand and makes it a primary research and development focus for automotive electronics manufacturers. Such applications are already widespread in European, American, and Japanese automobiles, where almost all vehicles are now equipped with millimeter-wave radar sensors, including collision avoidance radar and blind spot detection radar.

[0006] Transmit a frequency-modulated continuous wave (FMCW) signal with multiple cycles, each sweep cycle being T. chirp It can simultaneously measure the range and radial velocity of each target in a multi-target scenario, and the emitted waveform is as follows: Figure 1 As shown.

[0007] This waveform system is currently the most widely used, and it can simultaneously measure the distance and velocity of multiple targets. This waveform has been applied to automotive radar, skywave ultra-vision radar for ship detection, and tsunami detection.

[0008] The pulse repetition period is an important parameter, a short-time T chirp The frequency f of the echo signal after de-skewing B The frequency is negative. After the echo signal is mixed, the baseband signal can be measured using FFT to determine the difference frequency signal f. B The target distance R can be measured by the difference frequency signal f. B Given a relative delay τ, the target distance R is...

[0009]

[0010] Where c represents the speed of light, and T chirp and f SW This indicates the repetition period and signal bandwidth.

[0011] Traditional millimeter-wave radar processing typically involves transmitting signals with relatively small bandwidth. This is because for medium- to long-range targets, a larger transmitted signal bandwidth would increase the requirements for the analog-to-digital converter (AD) sampling rate. At the same time, an increased sampling rate would also increase computational complexity, making it impossible for some currently available commercial devices to meet the demands.

[0012] By employing an inter-pulse transmission carrier frequency that increments or decrements, the range resolution can be significantly improved while maintaining the existing device characteristics, largely transforming the angular resolution problem of automotive millimeter-wave radar into a range resolution problem. For example, a typical mid-to-long-range radar has a signal bandwidth of 200MHz and a range resolution of 0.75m. If two targets differ in distance by 0.1m but are close in angle, they cannot be effectively distinguished using traditional methods. However, with synthetic bandwidth, even if the two targets have identical angles, they can still be effectively distinguished. However, this synthetic bandwidth processing mode can only perform a single broadband imaging of the current target, still not achieving the effect of dense point cloud imaging.

[0013] Based on the above analysis, the problems and defects of the existing technology are as follows: the existing synthetic bandwidth processing mode can only perform broadband imaging of the current target once, and still cannot achieve the effect of contour dense point cloud imaging. Summary of the Invention

[0014] To address the problems existing in the prior art, this invention provides a method and system for contour imaging of automobiles using multispectral illumination millimeter-wave radar.

[0015] This invention is implemented as follows: a method for contour imaging of automobiles using multispectral illumination millimeter-wave radar, comprising:

[0016] S1: N, where the frequency interval between transmitted pulses is Δf. pulse One pulse;

[0017] S2: Set the minimum number of processing pulses N m N m It can be set to N pulse About half;

[0018] S3: Set the pulse count increment Δn. If the computational load is not considered, Δn can be set to 1, or Δn can be set between 1 and 10.

[0019] S4: Initialize the number of multispectral irradiations n d =1, starting from the first time;

[0020] S5: For numbers 1 to N m +(n d -1)×Δn pulses are coherently processed, and a two-dimensional FFT is performed on the obtained pulse echo data matrix to obtain a high-resolution RD spectrum;

[0021] S6: The high-resolution broadband composite bandwidth is obtained as (N) m +(n d The result of the synthesis of -1)×Δn)·Δf;

[0022] S7: Determine if the number of pulses has been processed, i.e., N. m +(n d -1)×Δn <N pulse If the expression is true, proceed to step 5; otherwise, proceed to the next step.

[0023] S8: Superimpose the target imaging results obtained from each spectral illumination in S6 to obtain the target multispectral illumination imaging results.

[0024] Furthermore, S6 includes:

[0025] Based on the target distance cell number, the distance interval of this distance segment is determined, and the calculation method is as follows:

[0026] r i,min = (i-1)·ΔR

[0027] r i,max =i·ΔR

[0028] Where ΔR=c / (2B) is the distance resolution corresponding to the bandwidth B of a single pulse transmission signal.

[0029] Furthermore, based on the boundaries of the distance cells, the corresponding velocity interval boundaries are calculated, and the velocity channel numbers are determined:

[0030]

[0031]

[0032] Where f start r is the frequency of the first pulse within the current velocity-dimensional FFT processing CPI, Δf is the frequency of the frequency step between pulses, and T is the pulse repetition interval. i,min r is the lower bound of the distance corresponding to the i-th distance unit. i,max It is the upper bound of the distance corresponding to the i-th distance unit.

[0033] Furthermore, based on the calculated speed range boundaries, the speed channel number is determined:

[0034] Based on the calculated velocity values, the corresponding Doppler channel numbers are calculated. Assuming the channel numbers start from 0 and go up to M-1, the channel numbers corresponding to the minimum and maximum values ​​of this velocity range are as follows:

[0035]

[0036]

[0037] in This indicates rounding down to the nearest integer.

[0038] Furthermore, the corresponding high-resolution range image within the velocity range is calculated:

[0039] Based on the velocity interval number q corresponding to the current distance unit. i,min and q i,max Calculate the corresponding high-resolution range image output within this velocity range.

[0040]

[0041] Where, q i,j Indicates that between q i,min and q i,max The speed channel number value between them.

[0042] Furthermore, the high-resolution range image output is corrected when velocity blur occurs:

[0043] The above formula calculates the distance value r. i,j Furthermore, this distance value is the distance value when the Doppler velocity caused by distance is not blurred. If velocity blurring has already occurred, the actual distance value is corrected to...

[0044]

[0045] Where n amb This represents the multiplicity of the velocity channel blurring; it is 1 when the channel is unblurred. For the current distance cell, this can be determined by checking if the current distance cell's value is greater than the maximum unblurred distance, i.e.:

[0046] n amb =floor(r) i / r v,max )

[0047] When the calculated values ​​of the lower and upper boundaries of a distance cell differ, the distance cell needs to be split into two parts, i.e.:

[0048] floor(r i,min / r v,max )≠floor(r i,max / r v,max )

[0049] When it is established, the following calculation is performed first:

[0050] n amb =floor(r) i,min / r v,max )

[0051] The search speed channel is used, and when the speed channel returns from M-1 to 0, the corresponding fuzzy multiplicity is incremented by 1, that is:

[0052] n amb =namb +1

[0053] Since the calculated velocity values ​​change in intervals, they can be represented as a matrix, and the corresponding distance dimension numbering should be:

[0054] r syn,num =q i,j +(n amb -1)·M

[0055] A higher-dimensional, high-resolution RD spectrum was obtained.

[0056] Furthermore, DOA estimation is performed on the high-resolution range spectrum to obtain the current spectral illumination imaging results of the target:

[0057] For each receiver matrix, calculate the corresponding high-resolution RD spectrum matrix, and perform DOA estimation on the first column of the matrix, which is the zero-velocity channel of different distance dimensions. The estimation can be performed using DBF or MUSIC methods. At this time, the clutter distribution can be accurately described in two-dimensional space.

[0058] Another object of the present invention is to provide a multispectral illumination-based automotive millimeter-wave radar contour imaging system that applies the aforementioned multispectral illumination-based automotive millimeter-wave radar contour imaging method, comprising:

[0059] Pulse transmission module: used to transmit N pulses with a frequency interval of Δf between them. pulse One pulse;

[0060] Minimum Pulse Count Setting Module: Used to set the minimum number of processing pulses N. m Typically N m It can be set to N pulse About half;

[0061] Pulse count increment module: Used to set the pulse count increment Δn. If the amount of calculation is not considered, Δn can be set to 1. In practical applications, Δn can be set between 1 and 10.

[0062] Irradiation frequency initialization module: used to initialize the number of multispectral irradiations n d =1, starting from the first time;

[0063] Coherent processing module: used for processing parameters from the 1st to the Nth coherent lines. m +(n d -1)×Δn pulses are coherently processed, and a two-dimensional FFT is performed on the obtained pulse echo data matrix to obtain a high-resolution RD spectrum;

[0064] Synthesis Result Acquisition Module: Used to obtain high-resolution broadband synthesis bandwidth of (N) m +(n dThe result of the synthesis of -1)×Δn)·Δf;

[0065] Imaging result acquisition module: used to superimpose the target imaging results of each spectral illumination to obtain the target multispectral illumination imaging result.

[0066] Another object of the present invention is to provide a computer device including a memory and a processor, wherein the memory stores a computer program, and when the computer program is executed by the processor, the processor performs the steps of the described method for contour imaging of automobiles based on multispectral illumination millimeter-wave radar.

[0067] Another object of the present invention is to provide a computer-readable storage medium storing a computer program, which, when executed by a processor, causes the processor to perform the steps of the described method for contour imaging of automobiles based on multispectral illumination millimeter-wave radar.

[0068] Another objective of this invention is to provide an information data processing terminal for implementing the aforementioned multispectral illumination-based automotive millimeter-wave radar contour imaging system.

[0069] Based on the above technical solutions and the technical problems solved, the advantages and positive effects of the technical solution to be protected by this invention are as follows:

[0070] First, regarding the technical problems existing in the aforementioned prior art, the creative technical effects resulting from solving these problems are described in detail below:

[0071] (1) This invention proposes a contour imaging method for automobile millimeter-wave radar based on multispectral illumination, that is, selecting a portion of all pulses for broadband synthesis instead of performing a single operation of selecting all broadband synthesis. Through multiple synthesis algorithms under various bandwidths, the target in the actual scene can be segmented according to different resolutions, which greatly increases the number of point cloud images.

[0072] (2) For ease of description, the frequency agility between transmitted pulses in this invention is increased at equal intervals, or at unequal intervals, and this method is still applicable.

[0073] (3) In this invention, continuous processing is performed from the first pulse to the last pulse. In practical applications, any pulse can be randomly selected for broadband synthesis. By controlling the number of selected pulses, the bandwidth of the synthesized signal can be controlled, thus achieving the effect of broadband illumination.

[0074] Secondly, this invention proposes a contour imaging method for automobile millimeter-wave radar based on multispectral illumination. Compared with the traditional method of non-frequency agility, which selects all pulses for processing at once, or the method of frequency agility, which also selects all pulses for processing, this invention selects only a portion of all pulses for broadband synthesis, instead of performing a single broadband synthesis operation. In this way, through multiple synthesis algorithms under various bandwidths, the target in the actual scene can be segmented according to different resolutions. Since the scattering points of the extended target will exhibit great flicker when illuminated by radar at different resolutions, it is by utilizing this flicker that the number of point cloud images can be greatly increased, and the target contour imaging can be achieved in the end.

[0075] Third, the expected benefits and commercial value of the technical solution of this invention after transformation are as follows:

[0076] After the technical solution of this invention is transformed, it is expected to realize a car radar product that can truly image targets, overturning the traditional millimeter-wave car radar that only focuses on the single perspective of angle resolution, and providing better sensor detection equipment for autonomous driving.

[0077] Fourth, the multispectral illumination-based automotive millimeter-wave radar contour imaging method has indeed brought about significant technological progress, specifically in the following aspects:

[0078] 1. Improved resolution: By using multispectral illumination and synthetic broadband techniques, this method enables high-resolution imaging. This is crucial for detecting small, fast-moving targets, such as pedestrians and bicycles.

[0079] 2. Enhanced target recognition capability: This method can improve the system's ability to recognize the shape and characteristics of targets by synthesizing imaging results from multiple spectra.

[0080] 3. Improved system robustness: Under multispectral illumination, even if the signal of a certain spectrum is interfered with or lost, it will not have a significant impact on the overall imaging results.

[0081] 4. Increased detection range: By using millimeter-wave pulses, this method can achieve long-distance detection, which is very important for scenarios such as cars driving on highways.

[0082] 5. Enhanced system real-time performance: This method can achieve rapid imaging by processing multiple pulses in parallel, thus meeting real-time requirements.

[0083] In summary, this multispectral illumination-based automotive millimeter-wave radar contour imaging method, with its advantages of high resolution, high target recognition capability, and wide detection range, provides an efficient and reliable perception solution for autonomous vehicles and intelligent transportation systems, and has high practical value. Attached Figure Description

[0084] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the embodiments of the present invention will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0085] Figure 1 This is a classic FMCW transmission sequence diagram provided in an embodiment of the present invention;

[0086] Figure 2 The present invention provides an FMCW transmission sequence and range-Doppler spectrum with continuously increasing carrier frequency; wherein, (a) is the transmission waveform, and (b) is the RD spectrum of an ideal point target;

[0087] Figure 3 These are schematic diagrams of several target scenarios provided in embodiments of the present invention; wherein, (a) an ideal point target, (b) two ideal point targets, and (c) continuous clutter;

[0088] Figure 4 The following is an RD spectrum of a conventional linear frequency modulated signal provided in an embodiment of the present invention; wherein, (a) is the transmitted waveform, (b) is the range-Doppler spectrum of an ideal point target, (c) is the range-Doppler spectrum of two ideal point targets, and (d) is the range-Doppler spectrum of continuous clutter.

[0089] Figure 5 These are the RD spectra of a conventional linear frequency modulated signal provided in this embodiment of the invention; wherein, (a) is the transmitted waveform, (b) is the RD spectrum of an ideal point target, (c) is the RD spectrum of two point targets, and (d) is the RD spectrum of a continuous clutter target;

[0090] Figure 6 This is a schematic diagram of spatial propagation of signals at different frequencies provided in an embodiment of the present invention.

[0091] Figure 7 This is a schematic diagram of continuous single-antenna transmission provided in an embodiment of the present invention;

[0092] Figure 8 This is a schematic diagram of a two-dimensional RD spectrum in a real-world scenario provided by an embodiment of the present invention;

[0093] Figure 9 This is a rendering of a car target under multispectral illumination in a car radar detection scenario provided by an embodiment of the present invention;

[0094] Figure 10 This is a flowchart of the vehicle millimeter-wave radar contour imaging method based on multispectral illumination provided in an embodiment of the present invention;

[0095] Figure 11 This is a structural diagram of a multispectral illumination-based automotive millimeter-wave radar contour imaging system provided in an embodiment of the present invention. Detailed Implementation

[0096] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0097] There are several specific implementation methods for the above-described multispectral illumination-based automotive millimeter-wave radar contour imaging method. The following are two examples:

[0098] Example 1:

[0099] The above method is applied to a millimeter-wave-based autonomous vehicle radar system, and the specific implementation scheme is as follows:

[0100] S1: First, the system transmits 100 pulses with a frequency interval of 50MHz.

[0101] S2: Set the minimum number of pulses to be processed to 50.

[0102] S3: Set the pulse count increment to 1.

[0103] S4: Initialize the number of multispectral irradiations to 1, starting from the first irradiation.

[0104] S5: Perform coherent processing on pulses 1 to 50, and perform two-dimensional FFT on the obtained pulse echo data matrix to obtain a high-resolution RD spectrum.

[0105] S6: Obtain high-resolution broadband synthesis results with a bandwidth of 5 GHz.

[0106] S7: Determine if the pulse count has been processed. If not, return to S5 to process the next group of pulses.

[0107] S8: Superimpose all target imaging results to obtain the final multispectral illumination imaging result.

[0108] Example 2:

[0109] The above method is applied to a millimeter-wave-based intelligent transportation system, and the specific implementation scheme is as follows:

[0110] S1: The system transmits 200 pulses at a frequency interval of 60MHz.

[0111] S2: Set the minimum number of processing pulses to 100.

[0112] S3: Set the pulse count increment to 2.

[0113] S4: Initialize the number of multispectral irradiations to 1, starting from the first irradiation.

[0114] S5: Perform coherent processing on the first to 100th pulses, and perform two-dimensional FFT on the obtained pulse echo data matrix to obtain the high-resolution RD spectrum.

[0115] S6: Obtain high-resolution broadband synthesis results with a bandwidth of 12 GHz.

[0116] S7: Determine if the pulse count has been processed. If not, return to S5 to process the next group of pulses.

[0117] S8: Superimpose all target imaging results to obtain the final multispectral illumination imaging result.

[0118] In these two embodiments, the first is for a radar system for autonomous vehicles, and the second is for an intelligent transportation system. Note that the specific implementation may vary depending on the specific needs and limitations of the system, such as the number of pulses, frequency intervals, and the minimum number of pulses to process.

[0119] To address the problems existing in the prior art, the present invention provides a method and system for contour imaging of automobiles using multispectral illumination millimeter-wave radar. The present invention will be described in detail below with reference to the accompanying drawings.

[0120] There are several specific implementation methods for the above-described multispectral illumination-based automotive millimeter-wave radar contour imaging method. The following are two examples:

[0121] Example 1:

[0122] The above method is applied to a millimeter-wave-based autonomous vehicle radar system, and the specific implementation scheme is as follows:

[0123] 1.S1: First, the system transmits 100 pulses with a frequency interval of 50MHz.

[0124] 2.S2: Set the minimum number of processing pulses to 50.

[0125] 3.S3: Set the pulse count increment to 1.

[0126] 4.S4: Initialize the number of multispectral irradiations to 1, starting from the first time.

[0127] 5.S5: Perform coherent processing on pulses 1 to 50, and perform two-dimensional FFT on the obtained pulse echo data matrix to obtain the high-resolution RD spectrum.

[0128] 6.S6: Obtain high-resolution broadband synthesis results with a bandwidth of 5GHz.

[0129] 7.S7: Determine if the number of pulses has been processed. If not, return to S5 to process the next group of pulses.

[0130] 8.S8: Superimpose all target imaging results to obtain the final multispectral illumination imaging result.

[0131] Example 2:

[0132] The above method is applied to a millimeter-wave-based intelligent transportation system, and the specific implementation scheme is as follows:

[0133] 1.S1: The system transmits 200 pulses at a frequency interval of 60MHz.

[0134] 2.S2: Set the minimum number of processing pulses to 100.

[0135] 3.S3: Set the pulse count increment to 2.

[0136] 4.S4: Initialize the number of multispectral irradiations to 1, starting from the first time.

[0137] 5.S5: Perform coherent processing on the first to 100th pulses, and perform a two-dimensional FFT on the resulting pulse echo data matrix to obtain a high-resolution RD spectrum.

[0138] 6.S6: Obtain high-resolution broadband synthesis results with a bandwidth of 12GHz.

[0139] 7.S7: Determine if the number of pulses has been processed. If not, return to S5 to process the next group of pulses.

[0140] 8.S8: Superimpose all target imaging results to obtain the final multispectral illumination imaging result.

[0141] In these two embodiments, the first is for a radar system for autonomous vehicles, and the second is for an intelligent transportation system. Note that the specific implementation may vary depending on the specific needs and limitations of the system, such as the number of pulses, frequency intervals, and the minimum number of pulses to process.

[0142] The automotive millimeter-wave radar transmits multiple pulses of linear frequency modulated (LFM) signals. Each pulse is sampled and subjected to a range-dimensional Fourier transform (FFT) to obtain the range-dimensional result. Similar processing is then performed on each pulse, and corresponding to a specific range cell, the echoes from multiple pulses are subjected to a velocity-dimensional FFT, i.e., a second FFT process. This yields a range-Doppler matrix.

[0143] To distinguish multiple targets, the synthetic bandwidth method used in radar can be employed to improve range resolution. For example, the carrier frequency of each transmitted pulse signal gradually increases, thus greatly expanding the overall transmission bandwidth. This method has several advantages: the bandwidth of a single pulse transmission is not large, and the signal bandwidth after radar echo de-skewing is also small, so a low sampling rate can be maintained, reducing the requirements for sampling and processing devices; multiple pulses are combined in a wideband at the receiver, achieving extremely high bandwidth. For example, in automotive radar applications, the synthetic bandwidth can reach up to 4 GHz, at which point the range resolution will reach 3.75 cm. This means that two targets that are 3.75 cm apart can be distinguished as two separate targets in the range dimension, regardless of whether they are angularly close.

[0144] Synthetic broadband methods can be divided into time-domain synthetic bandwidth method, frequency-domain synthetic bandwidth method, and spatial-domain synthetic bandwidth method.

[0145] The time-domain composite bandwidth method involves splicing the LFM echo signals of each sub-pulse in the time domain. The resulting signal is an LFM signal whose bandwidth and duration are the sum of the individual sub-pulses. The basic steps are: first, upsampling the sub-pulse signals using an interpolation-like process; then, performing spectrum shifting and phase compensation in the time domain, and delay compensation in the frequency domain; finally, splicing the signals in the time domain.

[0146] The frequency domain synthesis bandwidth method involves splicing the LFM echo signals of individual sub-pulses in the frequency domain. The basic steps are: first, perform a fast-time FFT on the sub-pulse signals; then, perform frequency domain spectrum shifting; and finally, perform spectrum splicing.

[0147] The spatial synthesized bandwidth method utilizes the separation of echo channels of signals transmitted by different antennas in space and then synthesizes the bandwidth.

[0148] The following are some typical test scenarios, where the radar range resolution is ΔR. Figure 3 (a) In the scene, an ideal point target is contained within a distance cell of distance r1; Figure 3 (b) In the scene, there are two ideal point targets within a distance cell of distance r1; Figure 3 (c) In the scenario, there is a continuous ground clutter within a distance cell of distance r1.

[0149] When the transmitted signal is a conventional linear frequency modulated signal, a two-dimensional FFT transform using the range dimension and the pulse dimension yields a range-Doppler matrix, commonly referred to as the RD spectrum, or range-Doppler spectrum, such as... Figure 4 As shown.

[0150] When each transmitted pulse signal has a different frequency range, such as when each pulse jumps according to a frequency Δf, the resulting RD spectrum will vary for different radar detection scenarios. Figure 5 As shown.

[0151] As can be seen from the figure, for an ideal stationary point target, the range cell in the RD spectrum still corresponds to the same range cell as the conventional linear frequency modulated signal, but the Doppler velocity is no longer 0. Similarly, for two point targets, two point targets appear in the RD spectrum, and the velocities of both are not 0. For a continuous clutter target, a continuous velocity range is corresponding to the Doppler cell.

[0152] This is because, corresponding to transmissions at different frequencies, the frequency variation between pulses causes a regular phase change between them. Performing FFT processing in the pulse dimension allows this continuous phase change to be reflected as a certain velocity value. Assume the precise distance to point target a is r. a The precise distance to point target b is r. b The starting distance of the continuous clutter is r1, and the distance length is ΔR, which means it covers the current distance cell.

[0153] First, taking target a as an example, the diagram shows the propagation of signals of different frequencies in the air as follows: Figure 6 As shown.

[0154] When the signal carrier frequency is f0, the phase obtained through two-way propagation in space is... The following can be calculated:

[0155]

[0156] Where c is the speed of light, and λ0 is the wavelength corresponding to frequency f0. Similarly, when the next pulse uses carrier frequency f0 + Δf, i.e., after increasing the frequency, the resulting phase is:

[0157]

[0158] As can be seen from equation (2), the frequency Δf is increased, and an additional phase term is added to the phase term. This phase is related to two factors: the target distance and the frequency increment.

[0159] When the pulse time interval is T, FFT processing is performed in the pulse dimension, and the continuous phase difference generates the velocity value.

[0160]

[0161] It is worth noting that the first starting pulse in the FFT process above starts at f0. If the starting frequency is f... startThe corresponding velocity value should be:

[0162]

[0163] In practice, calculating the target velocity value using FFT presents issues related to velocity resolution and the maximum unambiguous velocity range. Assuming the number of pulses processed by coherent processing is M, if only positive velocities are considered, the maximum unambiguous velocity range is:

[0164]

[0165] The speed resolution is:

[0166]

[0167] The number of velocity units spanned by one distance unit is:

[0168]

[0169] The above formula can also be written as:

[0170]

[0171] The frequency step size of each pulse transition is Δf. When M pulses are transmitted, the equivalent bandwidth is M·Δf, and the equivalent range resolution is:

[0172]

[0173] In other words, by varying the frequency between pulses, the original range cell is subdivided into Q resolvable range cells. A calculation example is given below; assuming T remains constant at 100 μs, the maximum unambiguous velocity v... max The velocity is 19.4805 m / s, and the velocity resolution Δv is 0.1522 m / s. According to typical automotive radar parameters, let the signal bandwidth B of each pulse be 200 MHz, and the initial range cell be 0.75 m. Then the velocity span of one range cell is 0.9740 m / s, which will span 7 velocity cells in the velocity dimension.

[0174] Considering only one antenna for transmission, the pulse repetition period T is 50µs.

[0175] Continuous clutter in a range cell can span multiple velocity cells, reaching the range value of the first maximum unambiguous velocity:

[0176]

[0177] Therefore, for actual clutter echo scenarios, its two-dimensional RD spectrum will be as follows: Figure 8 As shown.

[0178] Since the velocity difference between distance units is not exactly an integer multiple of the velocity resolution, there is an overlapping region between the extended velocity units of different distance units in the above figure, which depends on the actual parameter settings.

[0179] like Figure 9 As shown, a typical vehicle is 5 meters long, so it can be regarded as a continuous extended target. At a distance resolution of 0.75 meters, it will still occupy 6-8 distance units.

[0180] At this time, the transmitted signal adopts the following method: Figure 7 The waveform shown depicts an increasing transmission frequency between each pulse. Assume the total number of transmitted pulses is N. pulse The minimum number of pulses processed is N. m ,like Figure 10 As shown, the processing steps are as follows:

[0181] (1) Step 1: N with a frequency interval of Δf between transmitted pulses pulse One pulse;

[0182] (2) Second step: Set the minimum number of processing pulses N m Typically N m It can be set to N pulse About half;

[0183] (3) Third step: Set the pulse number increment Δn. If the amount of calculation is not considered, Δn can be set to 1. In practical applications, Δn can be set between 1 and 10.

[0184] (4) Fourth step: Initialize n times of multispectral irradiation d =1, starting from the first time;

[0185] (5) Fifth step: For the first to Nth m +(n d -1)×Δn pulses are coherently processed, and a two-dimensional FFT is performed on the obtained pulse echo data matrix to obtain a high-resolution RD spectrum;

[0186] (6) Step 6: Obtain the high-resolution broadband composite bandwidth as (N m +(n d The result of the synthesis of -1)×Δn)·Δf;

[0187] Specifically, it may include the following sub-steps:

[0188] ① Based on the target distance cell number, determine the distance interval of this distance segment. The calculation method is as follows:

[0189] r i,min = (i-1)·ΔR (11)

[0190] ri,max =i·ΔR (12)

[0191] Where ΔR=c / (2B) is the distance resolution corresponding to the bandwidth B of a single pulse transmission signal.

[0192] ② Based on the boundary of the distance cell, calculate the corresponding velocity interval boundary and determine the velocity channel number.

[0193]

[0194]

[0195] Where f start Δf is the frequency of the first pulse within the current velocity-dimensional FFT processing CPI, Δf is the frequency of the frequency step between pulses, and T is the pulse repetition interval.

[0196] ③ Determine the speed channel number based on the calculated speed range boundaries.

[0197] Based on the calculated velocity values, the corresponding Doppler channel numbers are calculated. Assuming the channel numbers start from 0 and go up to M-1, the channel numbers corresponding to the minimum and maximum values ​​of this velocity range are as follows:

[0198]

[0199]

[0200] in This indicates rounding down to the nearest integer.

[0201] ④ Calculate the high-resolution range image corresponding to the velocity range.

[0202] Based on the velocity interval number q corresponding to the current distance unit. i,min and q i,max Calculate the corresponding high-resolution range image output within this velocity range.

[0203]

[0204] Where, q i,j Indicates that between q i,min and q i,max The speed channel number value between them.

[0205] ⑤ High-resolution range image output when correcting speed blur

[0206] The above formula calculates the distance value r. i,j Furthermore, this distance value is the distance value when the Doppler velocity caused by distance is not blurred. If velocity blurring has already occurred, the actual distance value is corrected to...

[0207]

[0208] Where n amb This represents the multiplicity of the velocity channel blurring; it is 1 when the channel is unblurred. For the current distance cell, this can be determined by checking if the current distance cell's value is greater than the maximum unblurred distance.

[0209] n amb =floor(r) i / r v,max (19)

[0210] When the calculated values ​​of the lower and upper boundaries of a distance cell differ, the distance cell needs to be split into two parts.

[0211] floor(r i,min / r v,max )≠floor(r i,max / r v,max (20)

[0212] When it is established, the first calculation is performed.

[0213] n amb =floor(r) i,min / r v,max ) (twenty one)

[0214] The search speed channel is used, and when the speed channel returns from M-1 to 0, the corresponding fuzzy multiplicity is incremented by 1.

[0215] n amb =n amb +1 (22)

[0216] The velocity values ​​calculated according to equation (23) change in intervals, so they should be represented as a matrix. The corresponding distance dimension number should be...

[0217] r syn,num =q i,j +(n amb -1)·M (23)

[0218] At this point, a higher-dimensional, high-resolution RD spectrum can be obtained.

[0219] ⑥ Perform DOA estimation on the high-resolution range spectrum to obtain the current spectral illumination imaging results of the target.

[0220] For each receiver matrix, calculate the corresponding high-resolution RD spectrum matrix, and perform DOA estimation on the first column of the matrix, which is the zero-velocity channel of different range dimensions. This estimation can be performed using methods such as DBF or MUSIC. At this point, the clutter distribution can be accurately described in two-dimensional space.

[0221] (7) Step 7: Determine whether the number of pulses has been processed, i.e., N m +(n d -1)×Δn <N pulse If the expression is true, proceed to step five; otherwise, proceed to the next step.

[0222] (8) Step 8: Superimpose the target imaging results obtained from each spectrum illumination obtained in sub-step 6 of step 6 to obtain the target multispectral illumination imaging results.

[0223] like Figure 11 As shown, the multispectral illumination-based automotive millimeter-wave radar contour imaging system provided in this embodiment of the invention includes:

[0224] Pulse transmission module: used to transmit N pulses with a frequency interval of Δf between them. pulse One pulse;

[0225] Minimum Pulse Count Setting Module: Used to set the minimum number of processing pulses N. m Typically N m It can be set to N pulse About half;

[0226] Pulse count increment module: Used to set the pulse count increment Δn. If the amount of calculation is not considered, Δn can be set to 1. In practical applications, Δn can be set between 1 and 10.

[0227] Irradiation frequency initialization module: used to initialize the number of multispectral irradiations n d =1, starting from the first time;

[0228] Coherent processing module: used for processing parameters from the 1st to the Nth coherent lines. m +(n d -1)×Δn pulses are coherently processed, and a two-dimensional FFT is performed on the obtained pulse echo data matrix to obtain a high-resolution RD spectrum;

[0229] Synthesis Result Acquisition Module: Used to obtain high-resolution broadband synthesis bandwidth of (N) m +(n d The result of the synthesis of -1)×Δn)·Δf;

[0230] Imaging result acquisition module: used to superimpose the target imaging results of each spectral illumination to obtain the target multispectral illumination imaging result.

[0231] An application embodiment of the present invention provides a computer device, which includes a memory and a processor. The memory stores a computer program, and when the computer program is executed by the processor, the processor performs the steps of a multispectral illumination-based automotive millimeter-wave radar contour imaging method.

[0232] An application embodiment of the present invention provides a computer-readable storage medium storing a computer program that, when executed by a processor, causes the processor to perform the steps of a multispectral illumination-based automotive millimeter-wave radar contour imaging method.

[0233] An application embodiment of the present invention provides an information data processing terminal, which is used to realize a multispectral illumination-based automotive millimeter-wave radar contour imaging system.

[0234] II. Application Examples. To demonstrate the inventiveness and technical value of the technical solution of this invention, this section provides application examples of the technical solution of the claims on specific products or related technologies.

[0235] III. Evidence of the Relevant Effects of the Embodiments. The embodiments of the present invention have achieved some positive effects during research and development or use, and indeed possess significant advantages compared to existing technologies. The following description, in conjunction with data, charts, and other materials from the experimental process, illustrates these advantages.

[0236] The automotive millimeter-wave radar developed through this patented solution can achieve contour imaging of stationary and moving objects. For example, the detection of vehicles on the road can capture the shape characteristics of the vehicles, which facilitates subsequent target recognition and scene generation, providing more comprehensive and clear perception results for autonomous driving.

[0237] It should be noted that embodiments of the present invention can be implemented in hardware, software, or a combination of both. The hardware portion can be implemented using dedicated logic; the software portion can be stored in memory and executed by a suitable instruction execution system, such as a microprocessor or dedicated-design hardware. Those skilled in the art will understand that the above-described devices and methods can be implemented using computer-executable instructions and / or included in processor control code, for example, such code provided on a carrier medium such as a disk, CD, or DVD-ROM, a programmable memory such as read-only memory (firmware), or a data carrier such as an optical or electronic signal carrier. The devices and modules of the present invention can be implemented by hardware circuitry such as very large-scale integrated circuits or gate arrays, semiconductors such as logic chips, transistors, or programmable hardware devices such as field-programmable gate arrays, programmable logic devices, etc., or by software executed by various types of processors, or by a combination of the above-described hardware circuitry and software, such as firmware.

[0238] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any modifications, equivalent substitutions, and improvements made by those skilled in the art within the scope of the technology disclosed in the present invention, and within the spirit and principles of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. A method for contour imaging of automobiles using millimeter-wave radar based on multispectral illumination, characterized in that, include: S1: The frequency interval between transmitted pulses is... of One pulse; S2: Set the minimum number of processing pulses , Set as Half of; S3: Set pulse count increment , Set between 1 and 10; S4: Multispectral irradiation frequency initialization =1, starting from the first time; S5: For the first to Each pulse is coherently processed, and a two-dimensional FFT is performed on the resulting pulse echo data matrix to obtain a high-resolution RD spectrum. S6: The high-resolution broadband composite bandwidth is obtained as follows: The synthesized results include: determining the distance interval of the distance segment based on the target distance cell number. r i,min , r i,max ]; based on the distance interval boundary of the distance cell and , Calculate the corresponding velocity range boundaries, where The frequency corresponding to the first pulse within the CPI currently undergoing velocity-dimensional FFT processing. The frequency of the inter-pulse frequency step is determined; the velocity channel number is determined based on the calculated velocity interval boundary; the high-resolution range image corresponding to the velocity interval is calculated; the high-resolution range image output when velocity ambiguity is corrected, and a high-dimensional high-resolution RD spectrum is obtained; DOA estimation is performed on the high-resolution range spectrum to obtain the current spectrum illumination imaging result of the target. S7: Determine whether the pulse count has been processed, that is... If the expression is true, proceed to S5; otherwise, proceed to S8. S8: Superimpose the target imaging results obtained from each spectral illumination in S6 to obtain the target multispectral illumination imaging results.

2. The vehicle millimeter-wave radar contour imaging method based on multispectral illumination as described in claim 1, characterized in that, The distance interval of the distance segment is determined based on the target distance cell number. r i,min , r i,max The calculation method is as follows: in, The distance resolution is the bandwidth B of a single pulse transmitted signal.

3. The vehicle millimeter-wave radar contour imaging method based on multispectral illumination as described in claim 1, characterized in that, The distance interval boundary based on the distance unit and , Calculate the corresponding speed range boundaries and determine the speed channel number: T is the pulse repetition interval.

4. The vehicle millimeter-wave radar contour imaging method based on multispectral illumination as described in claim 1, characterized in that, Based on the calculated velocity value, the corresponding Doppler channel number is calculated, assuming the channel numbering starts from 0 and continues to 0. Then the channel numbers corresponding to the minimum and maximum values ​​of this speed range are respectively: in This indicates rounding down to the nearest integer.

5. The vehicle millimeter-wave radar contour imaging method based on multispectral illumination as described in claim 1, characterized in that, Based on the velocity range number corresponding to the current distance unit. and Calculate the corresponding high-resolution range image output within this velocity range: in, Indicates the distance value. Indicates being between and The speed channel number value between them.

6. The vehicle millimeter-wave radar contour imaging method based on multispectral illumination as described in claim 2, characterized in that, The high-resolution range image output when the speed blur is corrected then the true range value is corrected to... in This represents the multiplicity of the velocity channel blurring; when unblurred, this value is 1. For the current distance cell, this can be determined by checking if the current distance cell's value is greater than the maximum unblurred distance, i.e.: When the calculated values ​​of the lower and upper boundaries of a distance cell differ, the distance cell needs to be split into two parts, i.e.: When it is established, the following calculation is performed first: The search speed channel is used, and when the speed channel returns from M-1 to 0, the corresponding fuzzy multiplicity is incremented by 1, that is: Since the calculated velocity values ​​change in intervals, they can be represented as a matrix, and the corresponding distance dimension numbering should be: High-resolution RD spectra with high dimensionality were obtained.

7. The vehicle millimeter-wave radar contour imaging method based on multispectral illumination as described in claim 1, characterized in that, The step of performing DOA estimation on the high-resolution range spectrum to obtain the current spectral illumination imaging result of the target includes: For each receiver matrix, the corresponding high-resolution RD spectrum matrix is ​​calculated, and the DOA estimation is performed on the first column of the matrix, which is the zero-velocity channel of different distance dimensions. The DOA estimation is performed using the DBF or MUSIC method, which can accurately describe the clutter distribution in two-dimensional space.

8. A multispectral illumination-based vehicle millimeter-wave radar contour imaging system applying the multispectral illumination-based vehicle contour imaging method as described in any one of claims 1 to 7, characterized in that, include: Pulse transmission module: used for transmitting pulses with a frequency interval of [missing value]. of One pulse; Minimum Pulse Count Setting Module: Used to set the minimum number of pulses processed. , Set as Half of; Pulse Count Increment Module: Used to set the pulse count increment. , Set between 1 and 10; Irradiation count initialization module: Used for initializing the number of multispectral irradiations. =1, starting from the first time; Coherent processing module: used for processing the first to... Each pulse is coherently processed, and a two-dimensional FFT is performed on the resulting pulse echo data matrix to obtain a high-resolution RD spectrum. Synthesis Result Acquisition Module: Used to obtain the high-resolution broadband synthesis bandwidth. The results of the synthesis; Imaging result acquisition module: used to superimpose the target imaging results of each spectral illumination to obtain the target multispectral illumination imaging result.

9. A computer device comprising a memory and a processor, the memory storing a computer program, which, when executed by the processor, causes the processor to perform the steps of the multispectral illumination-based automotive millimeter-wave radar contour imaging method as described in any one of claims 1 to 7.

10. An information data processing terminal, the information data processing terminal being used to implement the automotive millimeter-wave radar contour imaging system based on multispectral illumination as described in claim 8.

Images