An environment transfer function identification method, system and radar for error compensation of a vehicle-mounted radar system

Abstract

This invention discloses an environmental transfer function (EPF) identification method, system, and radar for error compensation in vehicle-mounted radar systems. By constructing an EPF degradation criterion and fusing radar spatial spectrum and vehicle motion state data for real-time identification, this invention achieves automated and highly reliable determination of the ideal environment (far field, straight road, horizontal, unobstructed, no strong reflections, moving targets) required for real-vehicle calibration. An error compensation enable command is triggered only when all criteria are met to capture high-quality echo data, thereby incorporating three major systematic deviations—bumper penetration effect, ground multipath effect, and non-far field wavefront curvature effect—into the calibration process in one step. Actual road tests show that this technical solution reduces the pitch angle measurement error rate of vehicle-mounted radar by an order of magnitude compared to traditional dark box calibration schemes, improving angle measurement accuracy and system robustness in complex road environments.

Description

Technical Field

[0001] This invention belongs to the field of vehicle-mounted millimeter-wave radar technology, specifically relating to an environmental transfer function identification method, system, and radar for error compensation in vehicle-mounted radar systems. Background Technology

[0002] Automotive millimeter-wave radar is one of the core sensors in advanced driver assistance systems (ADAS) and autonomous driving systems. Its pitch angle measurement capability is crucial for identifying targets in front of the vehicle (such as cars, trucks, road obstacles, and ground depressions). Pitch angle measurement is usually achieved based on the phase difference between signals received by multiple antennas, and its accuracy is highly dependent on the calibration of the amplitude and phase consistency between the radar's various receiving channels.

[0003] Currently, the industry standard practice is to perform channel calibration in a darkroom on the production line before the radar leaves the factory. However, after extensive testing and research, the inventors of this invention discovered that radars calibrated in the darkroom still exhibit a persistently high pitch angle measurement error rate after being deployed on actual roads. The reason for this lies in the following fundamental discrepancies between the darkroom calibration environment and the actual driving environment, which cannot be eliminated on the production line: The lack of ground multipath effect and the absence of ground reflection inside the dark box, while the ground in a real road acts as a huge reflection boundary, will change the amplitude and phase distribution of the radar antenna aperture, especially significantly affecting the characteristics of the elevation dimension.

[0004] The lack of integrated effect from the bumper means that the radar, after actual installation, must be mounted behind bumpers of different models and materials. The bumper, acting as a medium, introduces additional amplitude and phase errors related to angle and frequency. Dark box calibration is performed during the radar production process, making it impossible to predict the bumper model to be installed on the future vehicle.

[0005] Inconsistent near-field and far-field conditions mean that the size of the anechoic chamber on the production line is typically only about 2 meters, while to meet far-field conditions (usually requiring more than 6 meters), the actual calibration target must be located more than 150 meters away. Although the target simulated by the anechoic chamber can adjust the echo delay, it cannot change the wavefront curvature. This leads to a systematic deviation in the path difference calculation when the near-field calibration coefficients are directly applied to far-field angle measurement.

[0006] In existing technologies, some solutions propose eliminating the aforementioned deviations through real-vehicle calibration, which involves collecting vehicle echoes from real roads under specific scenarios for calibration. This approach imposes stringent requirements on the scenario (e.g., a straight road at least 150 meters away, unobstructed, non-sloping, free of strong reflective objects, and with the target vehicle in motion). However, how to automatically, in real-time, and reliably identify scenarios that meet these stringent conditions and distinguish them from those that do not, in order to trigger or prohibit the collection of calibration data, is a critical problem that urgently needs to be solved in the industry, and currently, there is no effective technical solution.

[0007] Therefore, in view of the above-mentioned technical defects, there is an urgent need for an environmental transfer function identification method, system and radar for error compensation of vehicle radar systems. Summary of the Invention

[0008] To overcome the shortcomings and difficulties of the existing technology, the purpose of this invention is to propose an environmental transfer function identification method, system and radar for vehicle radar system error compensation, so as to automatically and accurately identify the ideal environment that can be used for real vehicle calibration, thereby providing a reliable triggering mechanism for vehicle radar system error compensation.

[0009] The first objective of this invention is to provide an environmental transfer function identification method for error compensation in vehicle-mounted radar systems; The second objective of this invention is to provide an environmental transfer function identification system for error compensation in vehicle-mounted radar systems; A third objective of this invention is to provide a vehicle-mounted radar; The first objective of this invention is achieved by the method comprising the following steps: Create, generate, and acquire radar detection spatial spectrum data and vehicle motion state data respectively; Based on the preset environmental transfer function degradation criterion, the radar detection spatial spectrum data and vehicle motion state data are fused and processed, and it is identified whether the transfer function of the current driving environment is equivalent to the ideal far-field free space transfer function. When the identification results are equivalent, a corresponding error compensation enable command is generated; wherein, the error compensation enable command is used to trigger the acquisition of radar echo data at the current moment in order to extract the system error compensation coefficient.

[0010] Furthermore, the degradation criterion for the environment transfer function includes a spatial geometric consistency criterion; the verification of the spatial geometric consistency criterion includes: The roadside point cloud is segmented into linear fits, and the direction vector dispersion of each fitted line segment is calculated. When the dispersion is lower than a preset straight road segment threshold, the criterion is satisfied.

[0011] Furthermore, the degradation criterion for the environmental transfer function includes an electromagnetic boundary sparsity criterion; the verification of the electromagnetic boundary sparsity criterion includes: Using the target track frame as a reference, detect whether there are other track frames blocking the line of sight from the radar phase center to the track frame, and detect whether there are discrete reflectors with radar cross sections exceeding a preset clutter threshold in the target's lateral neighborhood; when there are no obstructions and no high cross section reflectors in the lateral neighborhood, the criterion is satisfied.

[0012] Furthermore, the degradation criterion for the environmental transfer function includes a terrain levelness criterion; the verification of the terrain levelness criterion includes: The vehicle pitch angle time series data reported by the vehicle body inertial measurement unit is obtained. When all sample values ​​of the time series data are within the preset horizontal threshold range, the criterion is determined to be met.

[0013] Furthermore, the degradation criterion of the environment transfer function includes a target motion saliency criterion; the verification of the target motion saliency criterion includes: Based on the vehicle's speed and the target's radial speed, the target's absolute speed is calculated. When the absolute speed exceeds a preset dynamic target threshold, the criterion is satisfied.

[0014] Furthermore, the degradation criterion for the environment transfer function also includes a far-field distance criterion; the verification of the far-field distance criterion includes: Extract the radial distance of the target. If the radial distance is greater than or equal to a preset far-field boundary threshold, the criterion is satisfied.

[0015] Furthermore, the preset environment transfer function degradation criterion is a multi-criteria logical AND operation; when any criterion is not satisfied, the identification result is not equivalent, and the error compensation enable instruction is reset.

[0016] The second objective of the present invention is achieved as follows: the system includes a data acquisition module, a transfer function identification engine, and an enable gating module; The data acquisition module is configured to acquire radar detection spatial spectrum data and vehicle motion state data; The transfer function identification engine is configured to identify whether the transfer function of the current driving environment is equivalent to the ideal far-field free space transfer function by fusing the radar detection spatial spectrum data and vehicle motion state data based on the environmental transfer function degradation criterion. The enable gating module is configured to generate an error compensation enable command when the identification result is equivalent, so as to trigger the acquisition of radar echo data at the current moment.

[0017] Furthermore, the transfer function identification engine includes a spatial geometry analysis unit, which determines spatial geometric consistency by quantitatively evaluating the dispersion of the direction vector of the roadside point cloud.

[0018] The third objective of this invention is achieved as follows: the radar includes an antenna array, a transceiver channel, a digital signal processor, and a non-volatile storage medium; when the digital signal processor executes the computer program stored in the non-volatile storage medium, it implements the environmental transfer function identification method for error compensation of a vehicle-mounted radar system as described above.

[0019] This invention transforms the stringent scenario conditions required for real-vehicle calibration into a multi-dimensional "environmental transfer function degradation criterion" that can be calculated in real time and performs logical AND judgment. This invention achieves automated and highly reliable online identification of the ideal calibration environment, thereby enabling the precise triggering of system error compensation data capture in real driving scenarios. It calibrates the three major systematic deviations of bumper, ground multipath and non-far-field effects in one go, ultimately reducing the pitch angle measurement error rate of the vehicle radar by an order of magnitude compared to the traditional dark box calibration scheme, and improving the angle measurement accuracy and robustness of the radar in complex road environments. Attached Figure Description

[0020] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying 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.

[0021] Figure 1 This is a schematic diagram of the process steps of an environmental transfer function identification method for error compensation in a vehicle-mounted radar system according to the present invention. Figure 2 This is an embodiment of the environmental transfer function identification method for error compensation in vehicle-mounted radar systems according to the present invention; Figure 3 This is a second embodiment of the environmental transfer function identification method for error compensation in vehicle-mounted radar systems according to the present invention; Figure 4 This is a third embodiment of the present invention, which is a method for identifying the environmental transfer function for error compensation in a vehicle-mounted radar system. Figure 5 This is the fourth embodiment of the environmental transfer function identification method for error compensation in vehicle-mounted radar systems according to the present invention; Figure 6 This is the fifth embodiment of the present invention, which is a method for identifying the environmental transfer function for error compensation in a vehicle-mounted radar system. Figure 7 This is the sixth embodiment of the present invention, which is a method for identifying the environmental transfer function for error compensation in a vehicle-mounted radar system. Figure 8 This is a schematic diagram of an environmental transfer function identification system architecture for error compensation in vehicle-mounted radar systems according to the present invention. Detailed Implementation

[0022] To facilitate a clearer understanding of the objectives, technical solutions, and advantages of this invention, the invention will be further described below in conjunction with the accompanying drawings and specific embodiments. Those skilled in the art can easily understand other advantages and effects of this invention from the content disclosed in this specification.

[0023] This invention can also be implemented or applied through other different specific examples, and various details in this specification can also be modified and changed based on different viewpoints and applications without departing from the spirit of this invention.

[0024] It should be noted that if the embodiments of the present invention involve directional indicators (such as up, down, left, right, front, back, etc.), the directional indicators are only used to explain the relative positional relationship and movement of the components in a certain specific posture (as shown in the figure). If the specific posture changes, the directional indicators will also change accordingly.

[0025] Furthermore, if the embodiments of this invention involve descriptions such as "first" or "second," these descriptions are for descriptive purposes only and should not be construed as indicating or implying their relative importance or implicitly specifying the number of technical features indicated. Therefore, a feature defined with "first" or "second" may explicitly or implicitly include at least one of those features. Secondly, the technical solutions of the various embodiments can be combined with each other, but this must be based on the ability of those skilled in the art to implement them. When the combination of technical solutions is contradictory or impossible to implement, it should be considered that such a combination of technical solutions does not exist and is not within the scope of protection claimed by this invention.

[0026] The present invention will be further described below with reference to the accompanying drawings.

[0027] like Figures 1-7 As shown, the present invention provides an environmental transfer function identification method for error compensation of vehicle-mounted radar systems, the method comprising the following steps: Create, generate, and acquire radar detection spatial spectrum data and vehicle motion state data respectively; Based on the preset environmental transfer function degradation criterion, the radar detection spatial spectrum data and vehicle motion state data are fused and processed, and it is identified whether the transfer function of the current driving environment is equivalent to the ideal far-field free space transfer function. When the identification results are equivalent, a corresponding error compensation enable command is generated; wherein, the error compensation enable command is used to trigger the acquisition of radar echo data at the current moment in order to extract the system error compensation coefficient.

[0028] The degradation criterion for the environment transfer function includes a spatial geometric consistency criterion; the verification of the spatial geometric consistency criterion includes: The roadside point cloud is segmented into linear fits, and the direction vector dispersion of each fitted line segment is calculated. When the dispersion is lower than a preset straight road segment threshold, the criterion is satisfied.

[0029] The degradation criterion for the environmental transfer function includes an electromagnetic boundary sparsity criterion; the verification of the electromagnetic boundary sparsity criterion includes: Using the target track frame as a reference, detect whether there are other track frames blocking the line of sight from the radar phase center to the track frame, and detect whether there are discrete reflectors with radar cross sections exceeding a preset clutter threshold in the target's lateral neighborhood; when there are no obstructions and no high cross section reflectors in the lateral neighborhood, the criterion is satisfied.

[0030] The degradation criterion for the environmental transfer function includes a terrain levelness criterion; the verification of the terrain levelness criterion includes: The vehicle pitch angle time series data reported by the vehicle body inertial measurement unit is obtained. When all sample values ​​of the time series data are within the preset horizontal threshold range, the criterion is determined to be met.

[0031] The degradation criterion for the environmental transfer function includes a target motion saliency criterion; the verification of the target motion saliency criterion includes: Based on the vehicle's speed and the target's radial speed, the target's absolute speed is calculated. When the absolute speed exceeds a preset dynamic target threshold, the criterion is satisfied.

[0032] The degradation criterion for the environment transfer function also includes a far-field distance criterion; the verification of the far-field distance criterion includes: Extract the radial distance of the target. If the radial distance is greater than or equal to a preset far-field boundary threshold, the criterion is satisfied.

[0033] The preset environment transfer function degradation criterion is a multi-criteria logical AND operation; when any criterion is not satisfied, the identification result is not equivalent, and the error compensation enable instruction is reset.

[0034] Specifically, in the embodiments of the present invention, the provided method steps are as follows: First, determine if the following 6 conditions are met simultaneously. If they are met simultaneously, start extracting points on the actual vehicle: 1) 150 meters away; 2) straight road; 3) no obstructions; 4) not a slope; 5) no strong reflective objects such as iron guardrails on both sides; 6) the target vehicle is moving.

[0035] Condition: 150 meters away. Since the radar itself is for ranging, it is very easy to determine whether the target is 150 meters away.

[0036] Condition 2) Straight road, road edge recognition and road edge fitting module. Road edge fitting uses several short straight line segments to fit the road edge. The slope of these short fitted straight line segments is taken as the standard deviation. If the standard deviation is less than a set threshold, it can be determined as a straight road. Condition 3) No obstruction: Track frame generation module. The track frame estimates the length and width of the target based on the target point cloud distribution. No obstruction is determined by connecting the target's track frame to the radar position; if the line does not intersect with the track frames of other targets, it is considered unobstructed.

[0037] Condition 4) If it is not a slope, continuously collect data reported by the slope sensor built into the vehicle for a period of time. If the slope value of the slope sensor is lower than the set threshold during this period, it is determined that the road is not a slope during this period.

[0038] Condition 5) If there are no strong reflective objects such as iron railings on both sides, and there are no objects with an RCS exceeding a certain threshold on both sides of the target, then the condition is met.

[0039] 6) The target vehicle is moving; there is a vehicle speed extraction module. If the absolute value of the difference between the target vehicle's speed and the vehicle's speed exceeds a set threshold, the target vehicle is considered to be moving. The entire process is as follows: Figures 2-3 As shown.

[0040] like Figure 4 As shown, this illustrates the principle of angle measurement. The distance between the two antennas is d. An electromagnetic wave returning from a target in the direction θ reaches both antennas. There is a path difference between the two antennas, dsinθ. The corresponding phase difference is Δφ = dsinθ / λ * 2π. Therefore, the angle of the target can be obtained from the phase difference between the antennas. This is the principle of angle measurement.

[0041] like Figure 5 As shown, in vehicle-mounted millimeter-wave radar, besides this simplest path difference, there are the following components. Specifically, the transmitting circuit, transmitting feed line, receiving circuit, receiving feed line, radome, bumper, and ground plane all contribute to changes in phase difference.

[0042] To eliminate unnecessary influences, the current method used is darkroom calibration on the production line. The flowchart for darkroom calibration is as follows: Figure 6 As shown.

[0043] From above Figure 5 and Figure 6 It can be seen that the transmitting circuit, transmitting feed line, receiving circuit, receiving feed line, and radome were taken into account in the anechoic chamber calibration. For production feasibility, the target simulation horn distance was shortened. Although the target simulation horn can simulate a long distance, the distance causing the path difference remains unchanged.

[0044] This reveals the following discrepancies between darkroom calibration and real-world scenarios: The amplitude and phase errors introduced by the bumper are not considered. Since radar production takes place at the radar manufacturer, and radars are used in various vehicle models, it's impossible to calibrate with different bumpers during radar production. Therefore, the bumper's influence is inherently unacceptable. The impact of the ground plane on the antenna's amplitude and phase characteristics is also not considered, especially its significant effect on the antenna's elevation and phase characteristics. Furthermore, the reduced distance introduces errors.

[0045] To eliminate the impact of the above three points, this solution proposes on-vehicle calibration, such as... Figure 7 As shown, the effects of the bumper and the ground plane have been calibrated. During the actual vehicle calibration, an actual vehicle 150 meters away was selected, and the target distance reached 150 meters. The target simulated horn distance of the dark box was 1.5 to 6 meters. Therefore, the effects of too close a distance were also eliminated. It can be seen that the three effects mentioned above have been eliminated.

[0046] There are specific requirements for the site during actual vehicle calibration: 1) At least 150 meters away. 2) Straight road. 3) Unobstructed. 4) Not a slope. 5) No strong reflective objects such as iron guardrails on both sides. 6) The target vehicle is in motion.

[0047] like Figure 3 As shown, the reason for the 150-meter requirement has already been explained. A straight road ensures that electromagnetic waves return to the radar primarily from the target, rather than being reflected back to the radar via the roadside after a curve, thus reducing errors. Unobstructed views prevent target obstruction from affecting calibration quality, and a non-sloping road ensures that the pitch angle of vehicles traveling on the ground is 0 degrees. The absence of strong reflective objects such as guardrails on both sides ensures that electromagnetic waves return primarily from the target, rather than being reflected back to the radar via strong reflective objects.

[0048] The advantages of this invention are as follows: it eliminates the effects of bumper interference, ground plane interference, and interference from non-far-field factors. The pitch angle measurement error rate is reduced by an order of magnitude.

[0049] Example 1: Environmental Transfer Function Identification Method This embodiment describes an environmental transfer function identification method for error compensation in vehicle-mounted millimeter-wave radar systems. The method includes the following steps: Step S1: Acquire radar detection spatial spectrum data and vehicle motion state data.

[0050] Radar detection spatial spectrum data includes at least the target point cloud's range, radial velocity, azimuth, elevation, signal-to-noise ratio (SNR), and estimated radar cross-section (RCS). This data is generated by the vehicle-mounted millimeter-wave radar's transmit / receive array and signal processing flow.

[0051] Vehicle motion status data includes at least: vehicle speed (usually obtained from the CAN bus) and real-time pitch angle reported by the vehicle inertial measurement unit (IMU).

[0052] Step S2: Based on the preset environmental transfer function degradation criterion, integrate the above data to identify whether the transfer function of the current driving environment is equivalent to the ideal far-field free space transfer function.

[0053] In this embodiment, the preset environment transfer function degradation criterion includes five sub-criteria, which are logically ANDed, meaning that all sub-criteria must be satisfied simultaneously for the function to be considered "equivalent".

[0054] Sub-criterion 1 (spatial geometric consistency criterion): Determines whether it is a straight path.

[0055] The specific method is as follows: A curb recognition module is used to extract the curb point cloud from both sides of the road. Random Sample Consensus (RANSAC) or least squares method is used to perform piecewise straight-line fitting on the curb point cloud, resulting in several short line segments (e.g., each segment is 5 meters long). The direction vector (or slope) of all these short line segments is calculated. Then, the dispersion of these direction vectors is calculated, preferably expressed in terms of standard deviation. If the standard deviation is less than a pre-defined threshold K... slope If the standard deviation is 0.05 (e.g., 0.05), the current road segment is determined to be a straight road, satisfying the spatial geometric consistency criterion. Conversely, if the standard deviation is too large, it indicates that the road is curved and does not meet the criterion. The physical meaning is that on curved roads, the target echo received by the radar may be mixed with reflections from both sides of the curve, violating the assumption of an ideal plane wavefront.

[0056] Sub-criterion 2 (electromagnetic boundary sparsity criterion): Determine whether there is no obstruction and no strong reflective objects on both sides.

[0057] The specific method includes two sub-steps: Unobstructedness determination: For the target bounding box generated by the tracking module, which estimates the target's length and width, a line is drawn connecting the radar's phase center (antenna array reference point) to the center of the target bounding box or the corner closest to the radar. This line is then checked to see if it intersects with the bounding boxes of any other targets. If they do not intersect, the system is determined to be unobstructed.

[0058] No strong reflective objects determination: Search for the RCS values ​​of all point clouds on both sides of the target's bounding box (e.g., within 5 meters to the left and right). If the RCS of any point cloud exceeds the preset threshold K... rcs (For example, a typical RCS value of 10 dBsm corresponds to iron guardrails and large road signs), so it is determined that there are no strong reflective objects on both sides.

[0059] When both of the above sub-steps pass, the electromagnetic boundary sparsity criterion is satisfied. This criterion ensures that the echo energy received by the radar mainly originates from the direct wave along the target's main path, rather than from multipath reflections or interference sources.

[0060] Sub-criterion 3 (terrain levelness criterion): Determine whether it is not a slope.

[0061] The specific method is as follows: continuously collect pitch angle data reported by the vehicle's IMU within a time window T (e.g., 0.5 seconds). If the absolute value of all pitch angle samples within this time window is less than a preset threshold K... pitch (For example, 1 degree) indicates that the current road segment is a level surface, satisfying the terrain levelness criterion. Its physical meaning is that on a slope, the target's true pitch angle is not 0, and directly using it for calibration will lead to an error in the direction reference of the system error compensation coefficient.

[0062] Sub-criterion 4 (Target Motion Significance Criterion): Determine whether the target vehicle is in motion.

[0063] The specific method is as follows: Extract the vehicle speed V. ego (Set forward as positive). Track a stationary target and obtain its radial velocity V relative to the radar. r Considering that the target is traveling in the same direction as the vehicle and has an azimuth angle θ, the target's absolute velocity V target It can be approximated as Calculate V target The absolute value of the value, if it is greater than the preset threshold K v If the velocity is 1 m / s, the target is considered to be in significant motion and meets the criterion. The echo characteristics of stationary targets (such as road signs and parked vehicles) are different from those of moving vehicles (e.g., they lack micro-Doppler features) and are not suitable as calibration targets.

[0064] Sub-criterion 5 (far-field distance criterion): Determine whether the target is more than 150 meters away.

[0065] The specific method is as follows: directly read the target distance R output by the radar; if R is greater than or equal to the preset far-field distance threshold R... farfield (For example, 150 meters) then the criterion is met. This threshold is determined based on the radar aperture and wavelength using the far-field condition formula R. farfield It is calculated as ≥ 2D² / λ, where D is the antenna aperture and λ is the wavelength.

[0066] Step S3: When the identification result is equivalent, generate an error compensation enable command.

[0067] Only when all five sub-criteria are met does the identification engine output an "equivalent" result and generate a high-level active pulse signal as an error compensation enable command. This command is sent to the radar's data acquisition module, triggering it to "latch" or "capture" the raw echo data (i.e., ADC raw data or range-Doppler spectrum) at the current moment. Subsequently, the back-end error compensation coefficient extraction module uses this high-quality data to calculate the comprehensive amplitude and phase error matrix, including bumper, ground, and near-field effects, and updates it to the radar's signal processing link to correct the elevation angle measurements of all subsequent targets.

[0068] If any sub-criterion is not met, the identification result is "inequivalent", the enable instruction is reset or prohibited from being generated, the system continues to monitor, and no calibration data is collected.

[0069] Variant Implementation: In some variations, the specific implementation algorithms of each sub-criterion can be replaced. For example, the determination of a straight road can be limited to fitting the road edge, and can also be combined with visual lane line recognition or high-precision map information. The determination of no occlusion can also utilize deep learning models to directly perform semantic segmentation on radar point clouds. These variations all fall within the scope of the inventive concept of this invention, as they all aim to verify whether the environment transfer function degenerates to an ideal state.

[0070] To achieve the above-mentioned objectives, an environmental transfer function identification system for error compensation in vehicle-mounted radar systems is also provided, such as... Figure 8 As shown, the system includes a data acquisition module, a transfer function identification engine, and an enable gating module; The data acquisition module is configured to acquire radar detection spatial spectrum data and vehicle motion state data; The transfer function identification engine is configured to identify whether the transfer function of the current driving environment is equivalent to the ideal far-field free space transfer function by fusing the radar detection spatial spectrum data and vehicle motion state data based on the environmental transfer function degradation criterion. The enable gating module is configured to generate an error compensation enable command when the identification result is equivalent, so as to trigger the acquisition of radar echo data at the current moment.

[0071] The transfer function identification engine includes a spatial geometry analysis unit, which determines spatial geometric consistency by quantitatively evaluating the dispersion of the direction vector of the roadside point cloud.

[0072] Example 2: Environmental Transfer Function Identification System This embodiment describes an environmental transfer function identification system corresponding to the above method, which can be integrated into the digital signal processor (DSP) or microcontroller (MCU) of an on-board millimeter-wave radar.

[0073] The system includes: a data acquisition module with a data interface to the radar signal processing chain and the vehicle CAN bus, used to acquire radar detection spatial spectrum data and vehicle motion status data in real time, and to perform necessary data format unification and timestamp alignment.

[0074] Transfer Function Identification Engine: This engine internally contains processing units corresponding to the five sub-criteria mentioned above: spatial geometry analysis unit, electromagnetic boundary analysis unit, terrain level analysis unit, target motion analysis unit, and far-field distance analysis unit. Each unit independently calculates its own criterion result. The core of the engine is a multi-input AND gate logic circuit or equivalent software logic that performs an AND operation on the Boolean outputs of all units to output the final identification result.

[0075] Enable gating module: Receives the output of the identification engine and generates a hardware or software-level enable signal. This signal can be connected to the interrupt control register of the radar data acquisition driver layer to selectively store the raw data stream. To achieve the above-mentioned objectives, a vehicle-mounted radar is also provided, comprising an antenna array, a transceiver channel, a digital signal processor, and a non-volatile storage medium. The digital signal processor, when executing a computer program stored in the non-volatile storage medium, implements the aforementioned method for identifying the environmental transfer function for error compensation in a vehicle-mounted radar system.

[0076] Example 3: Vehicle-mounted radar This embodiment provides a vehicle-mounted millimeter-wave radar, which includes: an antenna array (usually a multiple-transmit multiple-receive MIMO array); a transceiver front-end (including a transmitter, receiver, local oscillator, mixer, etc.); a digital signal processor (DSP) or a field-programmable gate array (FPGA); and a non-volatile memory (such as Flash or EEPROM) on which computer program code is stored.

[0077] When the digital signal processor or FPGA executes the computer program code, all the steps of any of the methods described in Embodiment 1 will be implemented.

[0078] This invention transforms the stringent scenario conditions required for real-vehicle calibration into a multi-dimensional "environmental transfer function degradation criterion" that can be calculated in real time and performs logical AND judgment. This invention achieves automated and highly reliable online identification of the ideal calibration environment, thereby enabling the precise triggering of system error compensation data capture in real driving scenarios. It calibrates the three major systematic deviations of bumper, ground multipath and non-far-field effects in one go, ultimately reducing the pitch angle measurement error rate of the vehicle radar by an order of magnitude compared to the traditional dark box calibration scheme, and improving the angle measurement accuracy and robustness of the radar in complex road environments.

[0079] The embodiments described above are merely illustrative of several implementations of the present invention, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of the present invention. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the present invention, and these modifications and improvements all fall within the scope of protection of the present invention. Therefore, the scope of protection of this patent should be determined by the appended claims.

Claims

1. A method for identifying the environmental transfer function for error compensation in vehicle-mounted radar systems, characterized in that, The method includes the following steps: Create, generate, and acquire radar detection spatial spectrum data and vehicle motion state data respectively; Based on the preset environmental transfer function degradation criterion, the radar detection spatial spectrum data and vehicle motion state data are fused and processed, and it is identified whether the transfer function of the current driving environment is equivalent to the ideal far-field free space transfer function. When the identification results are equivalent, a corresponding error compensation enable command is generated; wherein, the error compensation enable command is used to trigger the acquisition of radar echo data at the current moment in order to extract the system error compensation coefficient; The environmental transfer function degradation criteria include spatial geometric consistency criteria, electromagnetic boundary sparsity criteria, terrain levelness criteria, target motion saliency criteria, and far-field distance criteria.

2. The environmental transfer function identification method for error compensation in vehicle-mounted radar systems according to claim 1, characterized in that, The environmental transfer function degradation criterion includes the spatial geometric consistency criterion; The verification of the spatial geometric consistency criterion includes: The roadside point cloud is segmented into linear fits, and the direction vector dispersion of each fitted line segment is calculated. When the dispersion is lower than a preset straight road segment threshold, the criterion is satisfied.

3. The environmental transfer function identification method for error compensation in vehicle-mounted radar systems according to claim 1 or 2, characterized in that, The environmental transfer function degradation criterion includes the electromagnetic boundary sparsity criterion; The verification of the electromagnetic boundary sparsity criterion includes: Based on the target track frame, detect whether there are other track frames blocking the line of sight from the radar phase center to the track frame, and detect whether there are discrete reflectors with radar cross-sections exceeding the preset clutter threshold in the target's lateral neighborhood. The criterion is satisfied when there is no obstruction and no high-scattering-section reflector in the lateral neighborhood.

4. The environmental transfer function identification method for error compensation in vehicle-mounted radar systems according to claim 3, characterized in that, The degradation criterion for the environmental transfer function includes a terrain levelness criterion; the verification of the terrain levelness criterion includes: The vehicle pitch angle time series data reported by the vehicle body inertial measurement unit is obtained. When all sample values ​​of the time series data are within the preset horizontal threshold range, the criterion is determined to be met.

5. The environmental transfer function identification method for error compensation in vehicle-mounted radar systems according to claim 3, characterized in that, The degradation criterion for the environmental transfer function includes the salience criterion for target motion; The verification of the target motion saliency criterion includes: Based on the vehicle's speed and the target's radial speed, the target's absolute speed is calculated. When the absolute speed exceeds a preset dynamic target threshold, the criterion is satisfied.

6. The environmental transfer function identification method for error compensation in vehicle-mounted radar systems according to claim 3, characterized in that, The degradation criterion for the environment transfer function also includes a far-field distance criterion; the verification of the far-field distance criterion includes: Extract the radial distance of the target. If the radial distance is greater than or equal to a preset far-field boundary threshold, the criterion is satisfied.

7. The environmental transfer function identification method for error compensation in vehicle-mounted radar systems according to claim 1, characterized in that, The preset environment transfer function degradation criterion is a multi-criteria logical AND operation; when any criterion is not satisfied, the identification result is not equivalent, and the error compensation enable instruction is reset.

8. An environmental transfer function identification system for error compensation in vehicle-mounted radar systems, characterized in that, The system includes a data acquisition module, a transfer function identification engine, and an enable gating module; The data acquisition module is configured to acquire radar detection spatial spectrum data and vehicle motion state data; The transfer function identification engine is configured to identify whether the transfer function of the current driving environment is equivalent to the ideal far-field free space transfer function by fusing the radar detection spatial spectrum data and vehicle motion state data based on the environmental transfer function degradation criterion. The enable gating module is configured to generate an error compensation enable command when the identification result is equivalent, so as to trigger the acquisition of radar echo data at the current moment.

9. The environmental transfer function identification system for error compensation in a vehicle-mounted radar system according to claim 8, characterized in that, The transfer function identification engine includes a spatial geometry analysis unit, which determines spatial geometric consistency by quantitatively evaluating the dispersion of the direction vector of the roadside point cloud.

10. A vehicle-mounted radar, comprising an antenna array, a transceiver channel, a digital signal processor, and a non-volatile storage medium, characterized in that, When the digital signal processor executes the computer program stored in the non-volatile storage medium, it implements the environmental transfer function identification method for error compensation of vehicle radar system as described in any one of claims 1 to 7.

Images