Vehicle door opening early warning method and vehicle

Abstract

The invention discloses a vehicle door opening early warning method and a vehicle, and the method comprises the steps: a plurality of UWB signal detectors arranged on a vehicle body detect UWB echo signals at the rear side of the vehicle body in real time, and the UWB echo signals are UWB signals reflected by a standard UWB signal emitted by a UWB signal emitter when the standard UWB signal encounters an object; analyzing the UWB echo signal to obtain a detection signal; the detection signal comprises the distance between a moving object corresponding to the current UWB echo signal and the vehicle body and the movement direction; when it is detected that the vehicle door state of the vehicle body is changed from the locked state to the opened state, if the detection signal represents that a moving object approaches the vehicle body at present and the distance between the moving object and the vehicle body is smaller than a preset threshold value, warning information is output according to the detection signal.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of vehicle safety warning, in particular to a vehicle door opening warning method and a vehicle. BACKGROUND

[0002] With the popularization of modern transportation tools, more and more attention is paid to the safety of vehicles. In daily life, when the driver or passenger opens the door to get off the vehicle after parking, they often ignore the situation behind the vehicle, which may collide with the moving objects (such as pedestrians, bicycles, motorcycles, etc.) close to the rear, causing accidents. Therefore, a vehicle door opening warning system has emerged as the times require.

[0003] The vehicle door opening warning system mainly refers to detecting the moving objects close to the sides and rear of the vehicle through sensors when the vehicle is stopped and ready to open the door, so as to prevent collision accidents when the door is opened. At present, the door opening warning systems on the market mainly use camera recognition or angular millimeter wave radar technology to perceive the targets behind the vehicle.

[0004] However, the existing camera and millimeter wave radar solutions have certain limitations. These solutions generally have high cost, high consumption of computing power, high power consumption and other problems, resulting in a low market penetration rate of the door opening warning function. This situation makes only part of high-end vehicles equipped with this safety function, and the general consumers are difficult to experience the safety convenience brought by technology.

[0005] CONTENT OF THE APPLICATION The technical scheme adopted by the present application to solve the above problems is: The purpose of the present application is to provide a vehicle door opening warning method and vehicle which effectively reduce power consumption and cost.

[0006] In order to achieve the above purpose, the present application provides a vehicle door opening warning method, which comprises: detecting the UWB echo signals behind the vehicle body in real time based on a plurality of UWB signal detectors arranged on the vehicle body, the UWB echo signals being UWB signals reflected back by objects after the standard UWB signals emitted by UWB signal transmitters hit the objects; analyzing the UWB echo signals to obtain a detection signal; When it is detected that the door state of the vehicle body changes from locking to opening, if the detection signal represents that there is a moving object currently approaching the vehicle body and the distance between the moving object and the vehicle body is less than a preset threshold, outputting warning information according to the detection signal.

[0007] Compared with the prior art, the vehicle door opening warning method provided by the above technical solution of the application adopts a UWB signal detector to detect a UWB echo signal behind the side of the vehicle body, and generates a detection signal accordingly. When the user opens the door, if it is judged that there is a dangerous situation according to the detection signal, a warning signal is sent. As can be seen, the above method can accurately judge the distance and movement direction of the moving object and the vehicle body, thereby significantly reducing accidents caused by improper opening of the door, improving the driving safety, and the UWB radar signal has higher measurement accuracy and lower requirements for hardware, without the need for complex image processing algorithms, so that the power requirement of the processor can be reduced, thereby effectively reducing the power consumption, and compared with the traditional camera and millimeter wave radar warning system, the cost is lower. BRIEF DESCRIPTION OF DRAWINGS

[0008] Figure 1 The vehicle door opening warning method flowchart in the application; Figure 2 The vehicle door opening warning method working state reference diagram in the application; Figure 3 The flowchart of the step of analyzing the UWB signal in the application. DETAILED DESCRIPTION

[0009] To explain the technical content, structural features, purposes and effects of the application in detail, the following will be described in detail in combination with the embodiments and the accompanying drawings.

[0010] In an embodiment of the application, a vehicle door opening warning method is disclosed for warning of dangerous situations when the vehicle door is opened. For example, when the vehicle door is opened, if there are pedestrians or bicycles approaching from the side and rear, according to the warning method, a warning signal can be given when the door is unlocked to remind the user to check the situation outside the vehicle before deciding whether to open the door, thereby improving the safety performance of the vehicle. For example, Figure 1 The warning method in the embodiment includes the following steps: S1: Real-time detection of UWB echo signals behind the side of the vehicle body based on a plurality of UWB (Ultra Wide Band) signal detectors arranged on the vehicle body, the UWB echo signals being UWB signals reflected back by objects after the standard UWB signals emitted by the UWB signal transmitters; S2: Analysis of the UWB echo signals to obtain a detection signal; The detection signal includes any one or several of the following: The distance between the moving object corresponding to the current UWB echo signal and the vehicle body; The speed of the moving object corresponding to the current UWB echo signal; The azimuth angle of the moving object corresponding to the current UWB echo signal relative to the vehicle body; The moving direction of the moving object corresponding to the current UWB echo signal.

[0011] The moving direction of the moving object can be calculated and obtained by the distance, speed, and azimuth information between the moving object and the vehicle body.

[0012] For example, the first spatial position of the moving object at the first time and the second spatial position of the moving object at the second time can be determined by the distance and azimuth information, and the moving direction of the moving object can be determined by combining the first spatial position and the second spatial position of the moving object.

[0013] S3: When it is detected that the door state of the vehicle body changes from locking to opening, if the detection signal represents that there is currently a moving object approaching the vehicle body and the distance between the moving object and the vehicle body is less than a preset threshold, an alarm information is output according to the detection signal.

[0014] It should be noted that the UWB signal detector and the UWB signal transmitter constitute a UWB radar system, and unlike the traditional UWB ranging system, the UWB radar system does not need to configure a UWB tag on the target object, so it can detect any object around the vehicle body.

[0015] Specifically, a small car, such as Figure 2 , is installed with several UWB signal detectors on the vehicle body for real-time monitoring of the environmental conditions behind the vehicle body. Then, the execution process of the above early warning method is as follows: 1. UWB signal detection: The UWB signal detector installed behind the vehicle body continuously receives the UWB echo signal.

[0016] For example, a bicycle approaches the parked small car from behind, and the UWB signal detector will receive the UWB echo signal reflected by the bicycle and the person on the bicycle.

[0017] 2. Signal analysis: The received UWB echo signal is analyzed to obtain a detection signal. The detection signal includes but is not limited to the distance and moving direction between the current reflecting object (bicycle) and the vehicle body.

[0018] Suppose the bicycle is detected to be 5 meters away from the vehicle body and is approaching the vehicle body.

[0019] 3. Door state detection: The door state is detected in real time. When the door state changes from locking to opening, further detection is triggered.

[0020] For example, the driver is preparing to open the door to get off, and at this time the door state changes to opening.

[0021] 4. Early warning judgment and output: The detection signal is checked, and if a moving object (a bicycle) is detected to approach the vehicle body and the distance is less than a preset safety threshold (for example, 2 meters), a warning information is sent out.

[0022] In this example, the bicycle is 5 meters away from the vehicle body and approaches, and it is assumed that the set early warning threshold is 2 meters. The current detection signal shows that the bicycle approaches but does not reach the early warning threshold, so no warning is sent out at present.

[0023] Continuing to monitor, if the bicycle continues to approach and enters the 2-meter range, the system will immediately send out a warning information, such as a warning sound in the vehicle or a prompt of “Attention: object approaching” on the rearview mirror display.

[0024] As can be seen, the above vehicle door opening warning method can accurately judge the distance and motion direction of the moving object and the vehicle body, thereby significantly reducing accidents caused by improper opening of the door and improving driving safety. Moreover, the UWB radar signal has higher measurement accuracy and lower requirements for hardware, without the need for complex image processing algorithms, so the power requirement of the processor can be reduced, thereby reducing power consumption. Moreover, compared with the traditional camera and millimeter wave radar warning system, the cost is lower.

[0025] Since a conventional vehicle door opening accident is generally a collision with a human body, a bicycle or an electric bicycle. For other vehicles that cannot see the human body, it is less likely to occur a collision with the vehicle door because it is more obvious on the rearview mirror. Therefore, the moving object corresponding to the UWB echo signal in the embodiment is a living body, which can not only warn pedestrians, bicycles and electric bicycles approaching the vehicle from the rear of the vehicle, but also further reduce the calculation amount of the UWB echo signal, and only the living body needs to be detected.

[0026] Further, by analyzing and identifying the signal components related to the physiological activities (such as breathing actions) of the living body in the micro motion changes in the UWB echo signal, a detection signal is generated.

[0027] Specifically, the “signal components related to the physiological activities (such as breathing actions) of the living body in the micro motion changes in the UWB echo signal” include but are not limited to: Distance, speed, azimuth angle and other information of the moving object and the calculated acceleration and its change mode (such as the gait cycle frequency of the pedestrian and the uniform motion characteristics of the vehicle).

[0028] The motion characteristics of the moving object are different, and the speed / acceleration and its change mode are different. For example, a pedestrian walks with a gait cycle → the speed fluctuates periodically with small amplitude (such as one step fast and one step slow); a car usually moves at a constant speed or accelerates uniformly → the speed is stable and has no periodic fluctuation.

[0029] By analyzing the above signal components, a corresponding target type detection signal can be generated to realize detection of a mobile object type as a living body.

[0030] Further, multi-dimensional feature data of mobile objects of known types (various types of living bodies) can be collected to form a sample database for model training, wherein the multi-dimensional feature data includes a high-dimensional feature vector composed of the above signal components and a type label of the corresponding mobile object.

[0031] In an optional solution, the type label can be divided into three types (pedestrian, animal or other living body), and three different target types correspond to different high-dimensional feature vectors.

[0032] The sample data of the sample database is input into a deep learning model (such as a 2D-CNN model) to enable the deep learning model to learn the mapping rule between the high-dimensional feature vector and the target type, and to realize type classification output of different mobile objects.

[0033] On the other hand, in order to improve the accuracy of UWB echo signal processing and avoid the influence of environmental noise data on it, the processing method of UWB echo signal is as follows: 1. Determine the target direction of the mobile object causing the change of the UWB signal by the change amount of a plurality of UWB echo signals; 2. Generate a core detection result based on the UWB echo signal consistent with the target direction; 3. Generate an auxiliary detection result based on the UWB echo signal deviating from the target direction; 4. Analyze the auxiliary detection result to distinguish and extract environmental noise data to obtain an estimate of background noise; 5. Filter the core detection result based on the background noise to obtain a detection signal.

[0034] In this embodiment, by estimating the background noise level and filtering the core detection data, noise interference can be effectively eliminated, thereby improving the accuracy of the detection signal, reducing the false positive rate, and improving the reliability of the early warning system. It provides more timely and accurate collision warning for the driver, enhances the safety performance of the vehicle, and reduces the risk of traffic accidents.

[0035] On the other hand, at least one UWB signal detector is arranged on each side of the rear, middle and front of the vehicle body, and preferably on each side of the rear of the vehicle body to detect the situation behind the left and right sides of the vehicle.

[0036] On the other hand, the UWB signal transmitter is arranged in a signal base station near the vehicle body or outside the vehicle body.

[0037] In one preferred embodiment, the analysis of the S2, UWB echo signal at least includes the following steps: S201, distance and speed information acquisition: structured processing of the received UWB echo signal, extracting and outputting the distance and speed information of the moving object.

[0038] Specifically, through the structured processing of the UWB echo signal, the basic quantitative data is provided for the subsequent steps, and the subsequent calculation / data processing is facilitated.

[0039] S202, azimuth information acquisition: based on the signal feature difference of the received same UWB echo signal, the azimuth information of the moving object is extracted and outputted.

[0040] Specifically, based on the feature difference of the same UWB echo signal (such as the phase difference received by multiple receiving modules in the same UWB signal detector), the azimuth information is extracted, which breaks through the limitation of single receiving module that cannot perceive direction. Through the analysis of the signal feature difference, the azimuth of the moving object can be stably outputted in complex scenes, improving the accuracy of azimuth measurement, supplementing the direction dimension information of the spatial positioning of the moving object, and enhancing the comprehensive perception ability of the relative position of the target.

[0041] S203, target information fusion: multi-source data fusion processing of distance and speed information and azimuth information to obtain the spatial position and motion state of the moving object. The motion state includes the speed, acceleration, and motion direction of the moving object.

[0042] Specifically, multi-source fusion of distance, speed, and azimuth information can eliminate the limitations and errors of single-source data. Multi-source data fusion processing can improve data robustness and integrate scattered parameters into unified spatial position (such as three-dimensional coordinates) and motion state (such as trajectory trend), which can more comprehensively reflect the target dynamics and provide complete and reliable feature input for subsequent higher-order type recognition functions.

[0043] In one preferred embodiment of the embodiment, S201, distance and speed information acquisition step includes: 1. When structurally processing the received UWB echo signal, first construct a multi-dimensional signal model, and the multi-dimensional signal model includes signal representation associated with distance and speed information (but not limited to distance and speed information, it can also be associated with azimuth, phase difference, etc.), and then extract distance and speed information based on the signal representation.

[0044] By constructing a multi-dimensional signal model containing distance and speed associated representation, both are extracted from the signal. Moreover, the multi-dimensional signal model can map the inherent association between signal features and target motion parameters, provide more associated basic data for subsequent fusion, and at the same time, can accelerate the calculation efficiency.

[0045] 2、In the extraction of distance and speed information, interference suppression processing is performed on the received UWB echo signal first, and the effective signal features directly related to the moving object are enhanced.

[0046] By performing interference suppression on the UWB echo signal in advance, irrelevant signals such as environmental noise and multipath reflection can be filtered out; the target directly related features are enhanced, and the effective components of the direct path signal can be highlighted. The above steps can reduce invalid signal interference, make the subsequently extracted distance and speed information purer, and directly improve the accuracy and stability of parameter calculation.

[0047] Further, the "constructing a multi-dimensional signal model" of the above preferred solution can be implemented by the following steps: A CIR matrix containing fast time dimension and slow time dimension is constructed; wherein the fast time dimension corresponds to the distance information of the moving object, and the slow time dimension corresponds to the speed information of the moving object.

[0048] In the above steps, the CIR matrix captures the energy distribution of the signal in the distance (such as the direct / multipath peak position) through the fast time dimension, records the trend of the signal change over time (such as the amplitude fluctuation caused by the Doppler shift) through the slow time dimension, and associates the distance and speed information in the matrix, so that the signal features of the two can be extracted cooperatively, avoiding the problems of parameter asynchronization or data fragmentation caused by separate processing or stage-by-stage processing, and improving the correlation and extraction efficiency of the distance and speed parameters.

[0049] Moreover, the structured storage mode of constructing a multi-dimensional signal model makes the distance and speed information no longer isolated, but stored uniformly in the constructed matrix, and the subsequent steps can directly extract the required parameters based on the row / column dimensions of the CIR matrix, avoiding data calling confusion. At the same time, the synchronous processing mode ensures that the distance (based on the fast time) and the speed (based on the slow time) are calculated based on the same set of signal data and the same time reference, eliminating the parameter error caused by the time difference from the source, and improving the consistency of the distance and speed.

[0050] More specifically, each row of the CIR matrix corresponds to all sampling points (TAPs) of a single pulse, and the values of each row correspond to the complex signal values (including amplitude and phase) of the pulse. In the complex signal values in each row of the CIR matrix, the direct path signal reflected by the moving object (the signal that best represents the true distance) will produce the maximum amplitude fluctuation at a certain TAP. Finding this maximum amplitude TAP can calculate the distance of the moving object through the time delay corresponding to it, thereby converting the fast time dimension (row) into distance information.

[0051] Meanwhile, each column of the CIR matrix corresponds to a set of corresponding TAP signal values of all pulses at the same sampling time, and based on the Doppler effect, the movement of a moving object causes a change in the phase of the signal. By analyzing the phase change rule in the column direction of the matrix, the Doppler frequency shift can be calculated, and then the radial velocity of the moving object can be calculated, so as to convert the slow time dimension (column) into velocity information.

[0052] Further, the "performing interference suppression processing on the received UWB echo signal" of the above preferred solution can be implemented by the following steps: The time domain signal of the CIR matrix is subjected to Fourier transform (FFT) to obtain the frequency domain signal of the range Doppler spectrum (RD spectrum), and then the signal selective fading caused by multipath interference is compensated by frequency domain equalization.

[0053] Further, the "performing interference suppression processing on the received UWB echo signal" of the above preferred solution can be implemented by the following steps: Then, the frequency domain signal of the range Doppler spectrum is converted back to the time domain signal of the CIR matrix by inverse Fourier transform (inverse FFT) to enhance the prominence of the pulse peak of the direct path of the UWB echo signal.

[0054] The essence of multipath interference is that signals of different frequencies are attenuated to different degrees after passing through the channel, and this difference is more intuitive in the frequency domain (such as some frequency bands are severely attenuated, and some bands are relatively normal). Multipath interference is difficult to separate in the time domain signal, but the frequency selective fading caused by multipath interference in the frequency domain signal has a regular pattern.

[0055] Therefore, the above steps convert the time domain signal of the CIR matrix into the frequency domain signal of the range Doppler spectrum by FFT for each pulse, which can facilitate the compensation of the selective fading caused by multipath interference by frequency domain equalization in the frequency domain, and accurately suppress the attenuation imbalance of each frequency component.

[0056] Optionally, after the frequency domain signal is converted back to the time domain signal by inverse FFT, the amplitude and recognition of the pulse peak of the direct path can be significantly enhanced, and the interference superposition of the multipath reflected signal can be weakened, so that the proportion of effective signals is higher when extracting the distance and velocity information from the CIR matrix subsequently, and the parameter extraction error caused by interference is reduced.

[0057] More specifically, the "frequency domain equalization" can be implemented by designing a minimum mean square error (MMSE) equalizer, specifically: By analyzing the attenuation law of the frequency domain signal, the compensation coefficient of the equalizer is calculated. For the frequency with serious attenuation, the compensation coefficient is greater than 1 (gain); for the frequency with normal attenuation, the compensation coefficient is close to 1 (no additional adjustment). Multiply each frequency in the CFR matrix by the corresponding compensation coefficient to obtain the equalized CFR matrix. At this time, the signal strengths of each frequency have tended to be consistent, the frequency imbalance caused by multipath interference is corrected, and the signals of all frequencies are restored to the ideal state without multipath interference.

[0058] Further, the "performing interference suppression processing on the received UWB echo signal" of the above preferred scheme can be implemented by the following steps: The constant false alarm rate (CFAR) detection algorithm is used to process the time domain signal of the CIR matrix or the frequency domain signal of the range Doppler spectrum to identify and lock the pulse peak position of the direct path of the UWB echo signal. The pulse peak position of the direct path is used as the reference for extracting the range and velocity values.

[0059] The constant false alarm rate detection algorithm is a dynamic threshold detection algorithm that dynamically adjusts the threshold (i.e. non-fixed threshold) according to the noise intensity around the detection point. If the amplitude of the detection point exceeds the dynamic threshold, it is determined to be the direct path peak; if it does not exceed, it is determined to be noise or multipath interference and is directly filtered out. Finally, false positives or false negatives caused by noise are avoided, and the true direct path peak (i.e. the pulse peak of the direct path) is locked.

[0060] Therefore, using the constant false alarm rate detection algorithm to process the time domain signal of the CIR matrix or the frequency domain signal of the range Doppler spectrum can stabilize the false alarm rate when the noise fluctuates, avoid mistaking noise or weak multipath peaks as direct path peaks, accurately identify and lock the pulse peak position of the direct path, provide a reliable reference for range (fast time dimension) and velocity (slow time dimension) value extraction, and significantly reduce the error of subsequent parameter calculation.

[0061] In one preferred embodiment of the embodiment, S201, the distance and velocity information acquisition step further comprises: Optimizing the extracted range and velocity values: based on the pulse peak position of the direct path, improving the parameter accuracy of the range and velocity values through resolution enhancement means; fusing multi-source observation data (i.e. range and velocity information) to reduce noise interference and output optimized range and velocity values.

[0062] Specifically, the above steps take the pulse peak position of the direct path as a reliable reference, refine the quantization accuracy of the distance and speed parameters through resolution enhancement means (for example, reducing the minimum scale of distance measurement, increasing the speed sampling interval, etc.), and fuse multi-source observation data (such as simultaneous signals of different receiving modules, sampling values of continuous time slices, etc.) to offset the noise fluctuations of single data, reduce the influence of random interference, and finally output distance and speed optimization values with higher precision and stronger stability, providing a better dynamic parameter basis for subsequent azimuth information fusion and target type identification.

[0063] Further, the "improving the parameter accuracy of distance and speed values through resolution enhancement means" of the above preferred solution can be implemented by the following steps: 1. Defining a local window with the pulse peak position of the direct path as the center, performing smoothing interpolation algorithm on the real and imaginary parts of the signal in the local window respectively, and generating subdivided sampling points by fitting the signal change trend.

[0064] In the above steps, the local window (for example, a 6x6 grid) is defined with the pulse peak position of the direct path, which can focus on the effective signal area, exclude irrelevant data interference outside the local window, and reduce the subsequent calculation amount. The real and imaginary parts of the signal are respectively subjected to smoothing interpolation, which can completely retain the phase and amplitude characteristics of the signal and avoid information loss caused by single processing. The generated subdivided sampling points fill the gaps of the original sampling interval, significantly improve the signal resolution, and provide high-density and high-fidelity data support for subsequent accurate positioning of the pulse peak optimization position.

[0065] More specifically, the "smoothing interpolation" can use Spline interpolation method to perform Spline interpolation on the real and imaginary parts in the local window, that is, to draw a continuous and smooth curve between the discrete points sampled by hardware, and then to insert more virtual sampling points on the curve. The density of the sampling points is greatly improved after interpolation, so as to realize sub-pixel level resolution enhancement through algorithmic point filling.

[0066] 2. Re-positioning the pulse peak optimization position of the direct path within the constraint range of the subdivided sampling points, and taking the pulse peak optimization position of the direct path as the reference for extracting the distance and speed optimization values.

[0067] Among them, a large number of virtual sampling points will appear in the local window after performing interpolation, but the real direct path peak value cannot be too far away from the initial detection point found in the previous step. Therefore, a constraint range is set, which can ensure that the real peak value is found and avoid the interference of pseudo-peak outside the constraint range.

[0068] In the above steps, relying on the high-density data of the subdivided sampling points, the real extreme point of the direct path pulse peak can be captured more clearly, the positioning deviation of the direct path pulse peak position caused by the large sampling interval can be corrected effectively, the optimized position of the direct path pulse peak is taken as the reference for extracting the distance and speed optimization values, the parameter error caused by inaccurate peak positioning can be directly eliminated, the scale of distance measurement is more accurate, the interval of speed calculation is more accurate, and the quantization accuracy and reliability of the two parameters are further improved. More specifically, the distance and speed optimization values can be obtained through the following preferred schemes.

[0069] In one preferred embodiment of this embodiment, S201, the distance and speed information acquisition step further comprises: 1. Distance calculation: the distance information of the moving object is calculated by the formula d = c * Δt / 2; wherein d is the distance between the moving object and the vehicle body, c is the speed of light, and Δt is the round-trip time of the UWB signal (theoretically twice the return time of the UWB echo signal).

[0070] In the above steps, relying on the constant characteristics of the speed of light c and the measured value of the UWB signal round-trip time Δt, a linear correspondence between time and distance is established, the above calculation method avoids complex conversion errors, and combining the advantages of high time resolution of the UWB echo signal, the subtle changes of the signal round-trip time Δt can be accurately converted into distance values, ensuring the centimeter-level accuracy of distance measurement, and providing direct and reliable quantitative basis for the spatial positioning of the moving object.

[0071] 2. Speed calculation: the speed information of the moving object is calculated by the formula v = Δf * λ / 2; wherein v is the radial speed of the moving object, Δf is the frequency offset of the UWB echo signal, and λ is the wavelength of the UWB echo signal.

[0072] In the above steps, the frequency offset Δf of the UWB echo signal and the physical correlation of the radial speed v are used, and the fixed parameter of the signal wavelength λ is combined to directly convert the frequency domain signal features into speed values. The above method does not need complex motion model derivation, can respond to the frequency change caused by the movement of the moving object in real time, improve the dynamic tracking ability of speed measurement, and accurately reflect the speed of the moving object approaching or moving away.

[0073] More specifically, the frequency offset Δf is calculated by the formula Δf = f_d / f_c, wherein f_d is the Doppler shift obtained by FFT transformation of the slow time dimension of the CIR matrix, and f_c is the carrier frequency.

[0074] It should be noted that the distance and speed information calculated and output in the above steps can correspond to the distance and speed values before optimization, or the distance and speed optimization values after optimization.

[0075] Further, the "fusing multi-source observation data to reduce noise interference" of the above preferred solution can be implemented by the following steps: The observation values of the plurality of receiving modules are fused by a weighted least squares method, and the final distance information is calculated by the formula d_final=(w1*d1+w2*d2) / (w1+w2). Wherein, the UWB signal detector includes a plurality of receiving modules arranged at a preset interval, w1 and w2 are inversely proportional to the signal-to-noise ratio of different receiving modules of the same UWB signal detector, d1 and d2 are the distance observation values (i.e. distance calculation results) of different receiving modules of the same UWB signal detector, and d_final is the final distance information.

[0076] Specifically, the observation value (i.e. measurement data) of a single receiving module will be affected by noise, resulting in errors in distance calculation. Therefore, the above steps rely on the plurality of receiving modules (such as antenna arrays) arranged at a preset interval in the UWB signal detector, and take advantage of the observation complementarity of the spatial distribution of the receiving modules to avoid local interference and field of view limitations of a single module. By making the weights w1 and w2 inversely proportional to the signal-to-noise ratio of each receiving module, the observation values d1 and d2 with higher accuracy (smaller variance) have a larger proportion in the fusion, and the interference of lower precision data is weakened. After weighted averaging by the formula d_final=(w1*d1+w2*d2) / (w1+w2), the random noise of each receiving module is effectively offset, and the final output d_final not only significantly improves the distance measurement accuracy, but also reduces the error caused by the fluctuation of a single module, providing a more stable quantitative basis for subsequent target positioning.

[0077] It should be noted that the formula described in the above steps corresponds to the fusion calculation method of the distance observation values (d1 and d2) of only two receiving modules of the same UWB signal detector. When the same UWB signal detector has three or more receiving modules, there will also be three or more distance observation values. At this time, a third or more parameters (i.e. d3, w3; d4, w4; d5, w5...) are introduced, and the final distance information is obtained after fusion.

[0078] In addition, the above steps can also include a fusion processing step for speed information, which calculates the final speed information by fusing multi-source speed observation values.

[0079] In one preferred solution of this embodiment, S202, the azimuth angle information acquisition step includes: Based on the signal feature difference between the same UWB echo signals received by the plurality of receiving modules, the azimuth angle value of the moving object is output by stability optimization and precision improvement processing. Wherein, the UWB signal detector includes a plurality of receiving modules arranged at a preset interval, and each receiving module will produce signal feature difference when receiving the same UWB echo signal.

[0080] The above steps use a plurality of receiving modules (for example, an antenna array) arranged at a plurality of preset intervals to provide multi-dimensional reference for the azimuth value calculation by using the characteristic differences (such as phase difference, amplitude ratio, etc.) of the same UWB echo signal when received by the receiving modules, thereby breaking through the limitation that a single receiving module cannot perceive the direction. Meanwhile, by filtering out transient interference through stability optimization and refining angle quantization through precision improvement processing, the output azimuth value is not only more accurate, but also can resist signal fluctuation influence, thereby supplementing reliable directional dimension data for target space positioning and enhancing the all-around perception ability of the relative position of the target, thereby laying a precise angle information foundation for subsequent information fusion.

[0081] Further, the same UWB echo signal received by each receiving module forms a phase difference, and the signal characteristic difference described in the above step is the phase difference. Based on this, the "calculating the azimuth initial value of the moving object" of the above preferred solution can be implemented through the following steps: The Capon beam forming algorithm is used to search for the peak value of the spatial spectrum function, and the phase difference corresponding to the peak value is taken as the phase difference value. The spatial spectrum function is constructed based on the steering vector and the signal covariance matrix, the steering vector is the phase distribution template of the UWB echo signal incident on different receiving modules from a certain angle, and the covariance matrix is calculated from the distance between the receiving modules and the wavelength of the UWB echo signal.

[0082] In the above steps, when constructing the spatial spectrum function, the steering vector is taken as the phase distribution template of the UWB echo signal incident on different receiving modules from a certain angle when the UWB echo signal reaches the UWB signal detector, and the covariance matrix calculated from the distance between the receiving modules and the wavelength of the signal is combined, so that the spatial spectrum function can truly map the signal energy distribution at different angles, and the physical correlation between the azimuth value and the signal characteristic is strengthened. At the same time, the Capon beam forming algorithm is used to search for the spectral peak, which can effectively suppress sidelobe interference and accurately lock the angle corresponding to the strongest energy direction as the azimuth value, thereby laying a high-quality parameter foundation for subsequent processing.

[0083] More specifically, when a plurality of receiving modules receive the UWB signal of the same target, the signal waveforms are similar, but there is a phase shift, and the phase shift amount is the phase difference value Δ The above steps are used to record the phase difference value Δ .

[0084] Further, the "stability optimization and precision improvement processing" of the above preferred solution can be implemented through the following steps: The signal covariance matrix R is subjected to diagonal loading processing, and a diagonal perturbation term with a set strength is added to the signal covariance matrix R to improve the numerical stability of the azimuth initial value.

[0085] The step adds a diagonal perturbation term to the signal covariance matrix R, which can effectively improve the singularity problem of the matrix caused by insufficient samples or excessive noise, enhance the numerical stability of the matrix, avoid the spectrum peak splitting and pseudo-peak phenomenon of the Capon beamforming algorithm in the case of weak signal or complex interference, keep the initial value of the azimuth angle stable in the noise fluctuation, reduce the extreme value jump, and provide a more reliable benchmark for subsequent optimization.

[0086] More specifically, the signal covariance matrix R needs to be calculated by sampling data of multiple groups of receiving modules, but in a dynamic scene of vehicle monitoring (such as fast movement of the target and short sampling time), the amount of sampling data may be insufficient, resulting in the signal covariance matrix R becoming a singular matrix (mathematically irreversible), and the Capon beamforming needs to calculate the inverse matrix of the signal covariance matrix. The singular matrix will cause the algorithm to crash and the spatial spectrum function P(θ) cannot be calculated. Therefore, the essence of the above step is to make the signal covariance matrix R always reversible, and the diagonal loading is equivalent to adding a stability term to ensure that the Capon beamforming algorithm can run normally.

[0087] In an optional solution, the initial value of the azimuth angle can also be directly calculated by the spatial spectrum function, as follows: At this time, the spatial spectrum function formula is where a(θ) is a steering vector, R is a signal covariance matrix, is an inverse matrix of the signal covariance matrix, and θ is an azimuth angle to be calculated. The spatial spectrum function P(θ) is used to measure the matching degree of the actual received signal and the steering vector a(θ), and the higher the value of P(θ), the greater the possibility of signal incidence from the θ direction. Therefore, the θ value corresponding to the peak value of the spatial spectrum is the initial value of the azimuth angle of the target object; the signal covariance matrix R is used to describe the correlation between the signals received by multiple receiving modules; and aH(θ) is the conjugate transpose of a(θ).

[0088] Optionally, after the initial value of the azimuth angle is calculated by the spatial spectrum function, more reliable and accurate angle information, i.e., the optimized value of the azimuth angle, can be obtained by combining the historical angle and the current angle information and through a Kalman filtering algorithm.

[0089] The above step takes the current azimuth angle as the initial value, avoids redundant calculation and irrelevant interference of global search, and through fine iteration in the local range by the Kalman filtering algorithm, can use a probability model to fit the mapping relationship between the signal characteristics and the angle, accurately locate the angle value with the minimum error, significantly weaken the quantization error of the initial value, make the azimuth angle value closer to the real angle, and improve the fineness of the angle measurement.

[0090] Wherein, the initial azimuth value is a relative concept, and the initial azimuth value obtained for the first time is obtained from the CIR matrix, and the value before each iteration in the iterative calculation of the spatial spectrum function can be called the initial azimuth value.

[0091] In one preferred embodiment of the embodiment, the azimuth information obtaining step S202 further comprises: Based on the phase difference physical model formula Δ =2πd*sinθ / λ to solve the azimuth value. Wherein, θ is the azimuth value, Δ is the phase difference value, d is the distance between the receiving modules, and λ is the wavelength of the UWB echo signal.

[0092] The above step solves the azimuth value by the phase difference physical model formula, and directly associates the phase difference value Δ , the distance between the receiving modules d, and the signal wavelength λ with the azimuth θ of the moving object.

[0093] In one preferred embodiment of the embodiment, the target information fusion step S203 comprises: 1. Time synchronization: synchronize all observation values to the same time reference; wherein the observation values include distance and speed information observed by different UWB signal detectors and azimuth information.

[0094] The essence of the above time synchronization step is to eliminate the target position deviation caused by the time difference, ensure that all subsequent data comparison and fusion are based on the target state at the same time, and avoid misjudgment caused by different time synchronization (for example, it may be mistaken for two different targets).

[0095] The above step synchronizes the distance, speed, and azimuth observation values of different UWB signal detectors to the same time reference, eliminates the target position deviation caused by the time difference between the data, ensures that all subsequent data comparison and fusion are based on the target state at the same time, and avoids misjudgment caused by different time synchronization (for example, it may be mistaken for two different moving objects).

[0096] Specifically, the same time reference can be the system clock of the vehicle (such as the vehicle-mounted ECU clock), and the sampling time of all UWB signal detectors is converted to this time reference. If the sampling time of a UWB signal detector does not coincide with the target time of the unified reference (non-synchronous sampling data), linear interpolation is used to complete the data, and finally all module observation data are unified to the same time.

[0097] 2. Coordinate system: convert the observation values in the local coordinate system with the UWB signal detector as the origin to the global coordinate system with the center of the rear axle of the vehicle as the origin.

[0098] The above step converts the observation value of the local coordinate system (the UWB signal detector as the origin) to the global coordinate system (the center of the rear axle of the vehicle as the origin), unifies the spatial reference, eliminates the coordinate deviation caused by the position difference of different UWB signal detectors, makes the multi-source data be fused in the same spatial framework, and improves the integrity of the target spatial positioning.

[0099] Specifically, when the global coordinate system is defined, a Cartesian coordinate system (x, y) is established with the center of the rear axle of the vehicle as the origin, the observation value (d, θ) of each module is calculated, the coordinates of the UWB signal detector in the global coordinate system are calculated, the local Cartesian coordinates of the moving object relative to the UWB signal detector are calculated, the local coordinates are converted into global coordinates, and finally the observation values of all UWB signal detectors are converted into (x, y) coordinates in the global coordinate system, and then the positions can be directly compared.

[0100] 3. Uncertainty initialization: Assign a covariance matrix P to the observation value; wherein the covariance matrix P is set by the reliability of each observation value.

[0101] Each UWB signal detector has an error (for example, the vehicle head module has less shielding and smaller distance error; the vehicle tail module has more multipath interference and larger distance error), if all observation data are treated equally, the data with large error will reduce the fusion precision, therefore, the initial uncertainty is needed to quantify the reliability of each data (the higher the reliability, the smaller the uncertainty), to provide a basis for subsequent weighted fusion, and to avoid the influence of data with large error on the final result. The above step assigns a covariance matrix P according to the reliability of the observation value, quantifies the error range of each observation data, so that the fusion algorithm can differentially process the data according to the reliability, reduce the interference weight of the low-quality observation value, provide an error quantification basis for the accurate fusion of multi-source data, and improve the result robustness.

[0102] Specifically, the defined covariance matrix P is the initial uncertainty matrix of each observation value, the observation value (x, y) of each UWB signal detector is represented by an N*N covariance matrix P, the variance value is calibrated according to the hardware precision of the UWB signal detector and the actual scene, and each global coordinate (x, y) is assigned a corresponding covariance matrix P, and the data weight is adjusted according to the size of the covariance matrix P during subsequent fusion (the smaller P is, the greater the weight is).

[0103] In one preferred embodiment of the embodiment, S203, the target information fusion step further comprises: 1. Redundancy detection: Set a spatial distance threshold; for any two observation values from different UWB signal detectors, calculate the Euclidean distance of the two in the global coordinate system; and determine whether the observation values have redundant association by comparing the Euclidean distance with the spatial distance threshold.

[0104] Specifically, if the observation values of two UWB signal detectors correspond to the same moving object, their spatial distance (i.e., Euclidean distance) in the global coordinate system will be small, and if they correspond to different moving objects or false targets, the distance will be large. Therefore, the above step determines whether it is a redundant observation (i.e., multiple observations of the same moving object) by using a Euclidean distance threshold. This step is used to preliminarily screen potential same target observation groups and prepares for subsequent clustering grouping, while excluding obvious false targets (e.g., only a single UWB signal detector observes and the observation distance is far beyond the threshold compared with other UWB signal detectors).

[0105] The above step sets a spatial distance threshold, calculates the Euclidean distance of observation values of different UWB signal detectors in the global coordinate system, and compares it with the threshold, which can quickly identify the redundant correlation of observation values among multiple UWB signal detectors, avoid repeated data interference caused by overlapping detection regions of moving objects, and ensure the simplicity and relevance of the data before fusion from the data source.

[0106] More specifically, for any two different observation values (x1, y1) and (x2, y2) of UWB signal detectors, the calculation method of Euclidean distance is as follows: first, calculate the coordinate difference of the two points in the x-axis direction and square it, then calculate the coordinate difference of the two points in the y-axis direction and square it, add the two square results, and finally take the square root of the sum, i.e., the calculation formula is dist=sqrt[(x2-x1)²+(y2-y1)²].

[0107] 2. Clustering grouping: a distance-based clustering algorithm is used to group observation values with similar positions in the global coordinate system into the same redundant cluster; and only when a redundant cluster contains at least two observation values from different UWB signal detectors, the redundant cluster is determined to be an effective redundant cluster.

[0108] The above step uses a distance-based clustering algorithm, which can aggregate redundant observation values of multiple UWB signal detectors into a cluster according to the similarity of spatial positions in the global coordinate system, i.e., multiple observation values of the same moving object are grouped into a redundant cluster, and only one target is calculated, which can avoid repeated counting. At the same time, the determination condition that a redundant cluster needs to contain at least two observation values from different detectors ensures that the data in the cluster is a multi-source effective observation of a real target (rather than a single-detector noise or false point), provides a reliable multi-source redundant set for subsequent fusion, enhances the information density by taking advantage of the observation complementarity of multiple detectors, filters meaningless single-point clusters (if a cluster has only one observation value of a UWB signal detector, it may be a false target and needs to be directly excluded), and enhances the effectiveness of the input data for fusion.

[0109] More specifically, the above-mentioned "distance-based clustering algorithm" can automatically cluster the observation values with close distances into a redundant cluster without knowing the number of moving objects in advance, and perform clustering with the coordinates in the global coordinate system as the clustering features, group all observation values according to the rule that the Euclidean distance is less than a set threshold, each redundant cluster represents a potential target, set the validity condition of the redundant cluster, and finally obtain a plurality of valid redundant clusters, each cluster corresponds to all redundant observations of a real moving object.

[0110] In one preferred embodiment of the embodiment, S203, the target information fusion step further comprises: The fusion output: based on the target motion model, the state prediction information of the moving object at the next time is obtained; all observation values in the valid redundant cluster and their corresponding observation noise covariance matrices are obtained; the state prediction information is updated by using the observation values and their corresponding observation noise covariance matrices through the Kalman filtering algorithm to obtain the fused state optimization information; and the optimized state optimization information is output. The state prediction information and the state optimization information include the position and velocity information of the moving object.

[0111] Specifically, the target motion model in the above-mentioned step preferably adopts a constant velocity model (assuming that the speed of the target object does not change in a short time) for describing the motion state change rule of the moving object at the predicted time; the observation noise covariance matrix (R1, R2,...) is the covariance matrix corresponding to each observation value, which is used to quantify the noise level of each observation value (Z1, Z2,...) in the valid redundant cluster; and the updating process of the Kalman filtering algorithm specifically includes: calculating the state deviation based on the state prediction information and the observation value, correcting the deviation to optimize the target state in combination with the observation noise covariance matrix, and obtaining the state optimization information.

[0112] Since the target is dynamically moving, the observation value at a certain time cannot accurately reflect its position and velocity information, and different observation values in the redundant cluster also have error differences in essence. Therefore, the above-mentioned step combines the target motion model and the observation reliability to fuse and optimize the observation value, the Kalman filtering algorithm can simultaneously realize state prediction and observation update, so that the prediction stage can conform to the actual motion law of the target object (such as uniform speed, acceleration and deceleration), avoiding the deviation of state estimation without basis, and then using the multi-source reliable observation values of the valid redundant cluster to accurately correct the deviation between the predicted value and the actual observation, and simultaneously updating the covariance corresponding to the observation value to quantify the error range of the current state, and finally obtaining the state optimization information of the target object, providing more quantitative basis for subsequent decision-making of the vehicle.

[0113] In another preferred embodiment of the application, a vehicle is also disclosed, which comprises a vehicle body and a plurality of UWB signal detectors and a controller arranged on the vehicle body, and the controller works based on the vehicle door opening warning method in the above-mentioned embodiment.

[0114] In another aspect, the vehicle is further provided with a sound alarm electrically connected with the controller, and the sound alarm is used to output the warning information.

[0115] The application further discloses a vehicle door opening warning system, which comprises one or more processors, a memory and one or more programs, wherein the one or more programs are stored in the memory and configured to be executed by the one or more processors, and the programs comprise instructions for executing the vehicle door opening warning method as described above. The processor can be a general-purpose central processing unit (CPU), a microprocessor, an application specific integrated circuit (ASIC), or one or more integrated circuits, which are used to execute the related programs to realize the functions required by the modules in the vehicle door opening warning system of the embodiments of the application or execute the vehicle door opening warning method of the method embodiments of the application.

[0116] The application further discloses a computer readable storage medium comprising a computer program, wherein the computer program can be executed by a processor to complete the vehicle door opening warning method as described above. The computer readable storage medium can be any available medium accessible by a computer or a data storage device such as a server, a data center and the like integrated with one or more available media. The available medium can be a read-only memory (ROM), a random access memory (RAM), a magnetic medium such as a floppy disk, a hard disk, a magnetic tape, a magnetic disc or an optical medium such as a digital versatile disc (DVD), or a semiconductor medium such as a solid state disk (SSD) and the like.

[0117] The embodiments of the application further disclose a computer program product or a computer program, which comprises computer instructions stored in a computer readable storage medium. A processor of an electronic device reads the computer instructions from the computer readable storage medium, and the processor executes the computer instructions to enable the electronic device to execute the vehicle door opening warning method as described above.

Claims

1. A vehicle door opening warning method characterized by comprising: The application comprises: Real-time detection of UWB echo signals behind the vehicle body by a plurality of UWB signal detectors arranged on the vehicle body, wherein the UWB echo signals are UWB signals reflected by objects after being emitted by a UWB signal emitter; Analysis of the UWB echo signals to obtain a detection signal; When the state of the vehicle door is detected to change from being locked to being opened, if the detection signal represents that there is currently a moving object approaching the vehicle body and the distance between the moving object and the vehicle body is less than a preset threshold, an alarm information is output according to the detection signal.

2. The vehicle door opening warning method according to claim 1, characterized by, The detection signal includes any one or several of the following: The distance between the moving object corresponding to the current UWB echo signal and the vehicle body; The speed of the moving object corresponding to the current UWB echo signal; The azimuth angle of the moving object corresponding to the current UWB echo signal relative to the vehicle body; The moving direction of the moving object corresponding to the current UWB echo signal.

3. The vehicle door opening warning method according to claim 1, characterized by, The analysis of the UWB echo signals includes at least the following steps: Distance and speed information acquisition: structured processing of the received UWB echo signals to extract and output the distance and speed information of the moving object; Azimuth angle information acquisition: based on the signal feature difference of the received UWB echo signals, the azimuth angle information of the moving object is extracted and output; Target information fusion: multi-source data fusion processing of the distance and speed information and the azimuth angle information to obtain the spatial position and motion state of the moving object.

4. The vehicle door opening warning method according to claim 3, characterized by The distance and speed information acquisition step includes: When the received UWB echo signals are structured, a multi-dimensional signal model is first constructed, which includes signal representation related to distance and speed information, and then the distance and speed information is extracted based on the signal representation; Before extracting the distance and speed information, interference suppression processing is performed on the received UWB echo signals, and the effective signal features directly related to the moving object are enhanced.

5. The vehicle door opening warning method according to claim 4, characterized by, The construction of the multi-dimensional signal model includes: Construction of a CIR matrix containing fast time dimension and slow time dimension; characterized in that the fast time dimension corresponds to the distance information of the moving object, and the slow time dimension corresponds to the speed information of the moving object.

6. The vehicle door opening warning method according to claim 5, characterized by The interference suppression processing of the received UWB echo signals includes: Fourier transform of the time domain signal of the CIR matrix to obtain a distance Doppler spectrum; Compensation of signal selective fading caused by multipath interference through frequency domain equalization.

7. The vehicle door opening warning method according to claim 6, characterized by, The interference suppression processing of the received UWB echo signals further includes: Inverse Fourier transform of the frequency domain signal of the distance Doppler spectrum back to the time domain signal of the CIR matrix to enhance the prominence of the direct path pulse peak of the UWB echo signal.

8. The vehicle door opening warning method according to claim 6, characterized by, The interference suppression processing of the received UWB echo signals further includes: The constant false alarm rate detection algorithm is used to process the frequency domain signal of the distance Doppler spectrum, to identify and lock the pulse peak position of the direct path of the UWB echo signal, and to use the pulse peak position of the direct path as a reference for extracting the distance and velocity values.

9. The vehicle door opening warning method according to claim 7, characterized by, The interference suppression processing performed on the received UWB echo signal further includes: The constant false alarm rate detection algorithm is used to process the time domain signal of the CIR matrix, to identify and lock the pulse peak position of the direct path of the UWB echo signal, and to use the pulse peak position of the direct path as a reference for extracting the distance and velocity values.

10. The vehicle door opening warning method according to any one of claims 8 or 9, characterized by, The distance and velocity information acquisition step further includes an optimization process, which includes: Based on the pulse peak position of the direct path, the parameter accuracy of the distance and velocity values is improved through a resolution enhancement means; Fusion of multi-source observation data to reduce noise interference, output distance and velocity optimization value.

11. The vehicle door opening warning method according to claim 10, characterized by, The parameter accuracy of the distance and velocity values is improved through a resolution enhancement means, which includes: A local window is defined around the pulse peak position of the direct path, and a smoothing interpolation algorithm is performed on the real and imaginary parts of the signal in the local window to generate a subdivided sampling point by fitting the signal trend; Based on the constraint range of the subdivided sampling point, the pulse peak optimization position of the direct path is relocated, and the pulse peak optimization position of the direct path is used as a reference for extracting the distance and velocity optimization value.

12. The vehicle door opening warning method according to claim 11, characterized by, The distance and velocity information acquisition step further includes distance calculation and velocity calculation, which includes: The distance information of the moving object is calculated by the formula d=c*Δt / 2; wherein d is the distance between the moving object and the vehicle body, c is the speed of light, and Δt is the round trip time of the UWB signal; The velocity information of the moving object is calculated by the formula v=Δf*λ / 2; wherein v is the radial velocity of the moving object, Δf is the frequency offset of the UWB echo signal, and λ is the wavelength of the UWB echo signal.

13. The vehicle door opening warning method according to claim 12, characterized by, The fusion of multi-source observation data to reduce noise interference includes: The observation values of multiple receiving modules are fused by a weighted least squares method, and the final distance information is calculated by the formula d_final=(w1*d1+w2*d2) / (w1+w2); wherein the UWB signal detector includes multiple receiving modules arranged at a predetermined interval, w1 and w2 are inversely proportional to the signal-to-noise ratio of different receiving modules of the same UWB signal detector, d1 and d2 are distance observation values of different receiving modules of the same UWB signal detector, and d_final is the final distance information.

14. The vehicle door opening warning method according to claim 3, characterized by, The azimuth angle information acquisition step includes: Based on the signal characteristic difference between the same UWB echo signals received by multiple receiving modules, the azimuth angle value of the target object is output through stability optimization and precision improvement processing; wherein the UWB signal detector includes multiple receiving modules arranged at a predetermined interval.

15. The vehicle door opening warning method according to claim 14, characterized by, The azimuth angle value of the target object is output, which includes: The Capon beamforming algorithm is used to search for the peak value of the spatial spectrum function, and the phase difference corresponding to the peak value is used as the phase difference value. The spatial spectrum function is characterized by being constructed based on a steering vector and a signal covariance matrix. The steering vector is a phase distribution template of the UWB echo signal in different receiving modules, and the signal covariance matrix is ​​calculated from the spacing between multiple receiving modules and the wavelength of the UWB echo signal.

16. The vehicle door opening warning method according to claim 15, characterized by, The stability optimization and accuracy improvement processes include: The signal covariance matrix is ​​subjected to diagonal loading processing, and a diagonal perturbation term of a set strength is added to the signal covariance matrix to improve the numerical stability of the azimuth angle value.

17. The vehicle door opening warning method of claim 16, wherein The acquisition of azimuth information also includes: Based on a phase difference physical model formula Δ =2πd*sinθ / λ to solve the azimuth value; characterized in that θ is the azimuth value, Δ is the phase difference value, d is the interval of the receiving module, and λ is the wavelength of the UWB echo signal.

18. The vehicle door opening warning method according to claim 3, characterized by, The target information fusion step includes time synchronization, coordinate system 1, and uncertainty initialization. The time synchronization, coordinate system 1, and uncertainty initialization include: All observations are synchronized to the same time reference; the observations include distance and velocity information and azimuth information observed by different UWB signal detectors. The observed values ​​in the local coordinate system with the UWB signal detector as the origin are transformed to the global coordinate system with the rear axle center of the vehicle as the origin; A covariance matrix is ​​assigned to the observations; characterized in that the covariance matrix is ​​set by the reliability of each observation.

19. The vehicle door opening warning method of claim 18, wherein The target information fusion step further includes redundancy detection and clustering grouping, wherein the redundancy detection and the clustering grouping include: Set a spatial distance threshold; For any two observations from different UWB signal detectors, calculate the Euclidean distance between them in the global coordinate system; Determine the magnitude of the Euclidean distance and the spatial distance threshold, and determine whether there is redundant correlation between the observed values ​​based on the comparison result; A distance-based clustering algorithm is used to group the observations that are close in location in the global coordinate system into the same redundant cluster; Determine whether the redundant cluster contains at least two observations from different UWB signal detectors; if so, the redundant cluster is determined to be a valid redundant cluster.

20. The vehicle door opening warning method according to claim 19, characterized by, The target information fusion step further includes a fusion output, which includes: Based on the target motion model, the state prediction information of the moving object at the next moment is obtained; Obtain all observations within the effective redundant cluster and their corresponding observation noise covariance matrix; Using the observed values ​​and the observed noise covariance matrix, the state prediction information is updated using the Kalman filter algorithm to obtain the fused state optimization information; The optimized state information is output; characterized in that the state prediction information and the state optimization information include the spatial position and motion state of the moving object.

21. A vehicle, comprising a vehicle body and a plurality of UWB signal detectors and a controller disposed on the vehicle body, the controller operating based on the vehicle door opening warning method according to any one of claims 1 to 20.

Images