A vehicle low-speed anti-collision auxiliary control method, device and equipment

Abstract

The application provides a vehicle low-speed anti-collision auxiliary control method, device and equipment, relates to the technical field of vehicle control, and comprises the following steps: when a vehicle is in a low-speed driving or waiting state, the current sound speed of the environment where the vehicle is located and a composite double-frequency reflected sound wave reflected by an obstacle in front of the vehicle after receiving a detection sound wave emitted by the vehicle are acquired; the composite double-frequency reflected sound wave is subjected to multi-domain interference feature analysis to obtain a multi-dimensional time-frequency feature; the current distance between the vehicle and the obstacle and the living body attribute of the obstacle are determined according to the current sound speed and the multi-dimensional time-frequency feature; and the anti-collision control strategy of the vehicle is determined based on the current distance and the living body attribute of the obstacle. The vehicle low-speed anti-collision auxiliary control method can greatly improve the active safety of the vehicle in the low-speed driving and waiting state.

Description

Technical Field

[0001] This application relates to the field of vehicle control technology, and more specifically, to a method, device, and equipment for low-speed collision avoidance auxiliary control of vehicles. Background Technology

[0002] With the increasing complexity of urban traffic scenarios and the widespread adoption of low-speed intelligent driving technology, the risk of collisions between vehicles and pedestrians, animals, and various obstacles continues to rise in scenarios such as parking, traffic jams, and factory operations. Acoustic detection, due to its low cost and stable short-range detection, has become the mainstream technology for low-speed collision avoidance assistance control of vehicles. Accurately distinguishing living beings from ordinary obstacles under complex reflection interference and adaptively executing braking strategies is a core requirement for improving active safety performance.

[0003] Most existing vehicle acoustic collision avoidance methods use single-frequency pulse echo ranging, which only obtains distance information through time-of-flight in the time domain, or supplements it with simple frequency domain features for conventional obstacle classification, and then performs fixed braking control according to a uniform threshold. This cannot effectively characterize the micro-motion-related living attributes of obstacles, resulting in the inability to accurately identify living targets, making it difficult to perform differentiated collision avoidance control, and having the defects of insufficient protection for active targets such as pedestrians or accidental braking. Summary of the Invention

[0004] The purpose of this application is to provide a vehicle low-speed collision avoidance auxiliary control method, device and equipment to solve the above-mentioned problems existing in the prior art, and to greatly improve the active safety of vehicles in low-speed driving and waiting states.

[0005] Firstly, a method for low-speed collision avoidance assist control of a vehicle is provided, the method including: When the vehicle is traveling at low speed or in a standby state, the current speed of sound in the environment where the vehicle is located and the composite dual-frequency reflected sound waves reflected by the obstacles in front of the vehicle after receiving the detection sound waves emitted by the vehicle are obtained. Multi-domain interference feature analysis was performed on the composite dual-frequency reflected sound wave to obtain multi-dimensional time-frequency features; Based on the current speed of sound and the multidimensional time-frequency characteristics, determine the current distance between the vehicle and the obstacle, as well as the liveness attribute of the obstacle; Based on the current distance and the liveness attribute of the obstacle, a collision avoidance control strategy for the vehicle is determined.

[0006] In an optional implementation, the method further includes: The ambient temperature, humidity, and noise levels of the vehicle's current environment are obtained. Calculate the current speed of sound based on the ambient temperature and humidity; Calculate the initial transmission parameters based on the current sound speed, the ambient noise, and the configured reference transmission parameters; The vehicle's acoustic transducer is controlled to emit detection acoustic waves around the vehicle according to the initial emission parameters and receive echo signals. The echo signal is evaluated for quality based on the configured quality scoring model to obtain a comprehensive quality score. A model-free optimization algorithm is used to iteratively optimize the initial launch parameters within a preset launch parameter range based on the comprehensive quality score to obtain the current launch parameters; The acoustic transducer is controlled to emit detection acoustic waves around the vehicle with the current emission parameters, and to receive composite dual-frequency reflected acoustic waves reflected by obstacles in front of the vehicle after receiving the detection acoustic waves.

[0007] In an optional implementation, the composite dual-frequency reflected sound wave includes a pulse echo component and a first echo component and a second echo component at each time point; the multidimensional time-frequency characteristics include: a continuous phase difference sequence and an amplitude ratio sequence; Multi-domain interference feature analysis was performed on the composite dual-frequency reflected sound wave to obtain multi-dimensional time-frequency features, including: The first echo component and the second echo component at each time point are quadrature demodulated to obtain the first complex baseband signal and the second complex baseband signal at each time point. Calculate the difference frequency complex signal at each time point based on the first complex baseband signal and the second complex baseband signal at each time point; Extract the phase sequence of the difference frequency complex signal at each time point to obtain a continuous phase difference sequence; The amplitudes of the first complex baseband signal and the second complex baseband signal at each time point are calculated respectively to obtain the first echo amplitude and the second echo amplitude at each time point; An amplitude ratio sequence is generated based on the ratio of the first echo amplitude to the second echo amplitude at each time point.

[0008] In an optional implementation, the multidimensional time-frequency features further include: time delay features and a spectrum centroid sequence energy distribution width sequence; The step of performing multi-domain interference feature analysis on the composite dual-frequency reflected sound wave to obtain multi-dimensional time-frequency features also includes: Short-time Fourier transform is performed on the first complex baseband signal at each time point to obtain the time-frequency energy spectrum at each time point; Calculate the spectral centroid sequence and energy distribution width sequence based on the time-frequency energy spectrum at each time point; For the pulse echo component, the time difference from transmission to reception of each pulse is obtained as a time delay feature.

[0009] In an optional implementation, the current transmission parameters include: a first frequency corresponding to the first echo component and a second frequency corresponding to the second echo component; Determining the current distance between the vehicle and the obstacle based on the current speed of sound and the multidimensional time-frequency characteristics includes: Calculate the pulse ranging value based on the time delay characteristics and the current sound speed; Calculate the interferometric ranging value based on the current sound speed, the continuous phase difference sequence, the first frequency, and the second frequency; Calculate the initial distance based on the pulse ranging value and the interferometric ranging value; The continuous phase difference sequence is differentiated to obtain the phase change rate; the product of the phase change rate and the preset conversion coefficient is taken as the radial velocity of the obstacle relative to the vehicle. Integrating the radial velocity over time yields the distance change. The current distance is obtained based on the initial distance and the change in distance.

[0010] In an optional implementation, the method for determining the living properties of the obstacle includes: Based on the continuous phase difference sequence and the amplitude ratio sequence, a dual-frequency reflection joint feature map is constructed; The statistical characteristics of the spectral centroid sequence and the energy distribution width sequence are calculated respectively to obtain the first statistical characteristic and the second statistical characteristic; The dual-frequency reflection joint feature map, the first statistical feature, and the second statistical feature are input into a pre-trained obstacle material type classification model to obtain the material type of the obstacle; The continuous phase difference sequence is subjected to micro-motion energy extraction and time-frequency ridge feature extraction to obtain micro-motion features and time-frequency ridge features; Based on the current distance, the time-frequency ridge feature, the material type of the obstacle, and the micro-motion feature, the living attribute of the obstacle is determined.

[0011] In an optional implementation, the liveness attribute includes: live and non-live; Based on the current distance and the liveness attribute of the obstacle, a collision avoidance control strategy for the vehicle is determined, including: Match the target distance threshold corresponding to the configured liveness attribute from the lookup table of different liveness attributes and different distance thresholds; If the current distance is greater than the target distance threshold, then a warning strategy is determined based on the current distance, and a warning is issued. If the current distance is not greater than the target distance threshold and the liveness attribute indicates that the subject is alive, then the configured first anti-collision control strategy is used for braking. If the current distance is not greater than the target distance threshold and the liveness attribute is non-liveness, then the configured second anti-collision control strategy is used for braking.

[0012] Secondly, a low-speed collision avoidance assist control device for vehicles is provided, the device comprising: The acquisition unit is used to acquire the current speed of sound in the environment where the vehicle is located and the composite dual-frequency reflected sound waves reflected by obstacles in front of the vehicle after receiving the detection sound waves emitted by the vehicle when the vehicle is in a low-speed driving or waiting state. The analysis unit is used to perform multi-domain interference feature analysis on the composite dual-frequency reflected sound wave to obtain multi-dimensional time-frequency features; The first determining unit is configured to determine the current distance between the vehicle and the obstacle, as well as the liveness attribute of the obstacle, based on the current sound speed and the multi-dimensional time-frequency characteristics. The second determining unit is used to determine the collision avoidance control strategy of the vehicle based on the current distance and the liveness attribute of the obstacle.

[0013] Thirdly, an electronic device is provided, which includes a processor, a communication interface, a memory, and a communication bus, wherein the processor, the communication interface, and the memory communicate with each other through the communication bus; Memory, used to store computer programs; When a processor executes a program stored in memory, it implements any of the steps described in the first aspect above.

[0014] Fourthly, a computer-readable storage medium is provided, wherein a computer program is stored therein, and when executed by a processor, the computer program implements the steps of any of the methods described in the first aspect above.

[0015] This application, by combining real-time ambient sound velocity with multi-domain interference feature analysis of composite dual-frequency reflected sound waves, can accurately determine the distance between the vehicle and obstacles and the liveness attribute of the obstacles, effectively improving the accuracy of obstacle detection and recognition. Then, based on the distance and liveness attribute, a collision avoidance control strategy can be formulated, which greatly improves the active safety of the vehicle in low-speed driving and waiting states, and avoids control failure or malfunction caused by detection errors. Attached Figure Description

[0016] To more clearly illustrate the technical solutions of the embodiments of this application, the accompanying drawings used in the embodiments of this application will be briefly introduced below. It should be understood that the following drawings only show some embodiments of this application and should not be regarded as a limitation of the scope. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0017] Figure 1 An architecture diagram of a vehicle low-speed collision avoidance assist control system provided in this application embodiment; Figure 2 A flowchart illustrating a low-speed collision avoidance assist control method for vehicles provided in this application embodiment; Figure 3 A schematic diagram of the structure of a vehicle low-speed collision avoidance auxiliary control device provided in this application embodiment; Figure 4 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. Detailed Implementation

[0018] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of the embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application. Unless otherwise defined, the technical or scientific terms used in this application should have the ordinary meaning understood by those skilled in the art. The words "first," "second," and similar terms used in this application do not indicate any order, quantity, or importance, but are only used to distinguish different components. The words "comprising" or "including," etc., mean that the element or object preceding the word covers the element or object listed after the word and its equivalents, but do not exclude other elements or objects. The words "connected," "coupled," or "connected," etc., are not limited to physical or mechanical connections, but can include electrical connections, whether direct or indirect. "Up," "down," "left," "right," etc., are only used to indicate relative positional relationships. When the absolute position of the described object changes, the relative positional relationship may also change accordingly.

[0019] The vehicle low-speed collision avoidance auxiliary control method provided in this application embodiment can be applied to... Figure 1 In the system architecture shown, such as Figure 1 As shown, the system may include: a vehicle controller, multiple acoustic transducers deployed around the vehicle, and multiple data acquisition components. The vehicle controller is used to execute the low-speed collision avoidance auxiliary control method for vehicles provided in the embodiments of this application; Multiple acoustic transducers are used to emit continuous detection acoustic waves of a first frequency, continuous detection acoustic waves of a second frequency, and a short-time pulse detection acoustic wave to the area around the vehicle. Multiple data acquisition components are used to obtain the current speed of sound in the environment in which the vehicle is located, as well as the composite dual-frequency reflected sound waves reflected by obstacles in front of the vehicle after receiving the detection sound waves emitted by the vehicle.

[0020] The preferred embodiments of this application are described below with reference to the accompanying drawings. It should be understood that the preferred embodiments described herein are for illustration and explanation only and are not intended to limit this application. Furthermore, the embodiments and features in the embodiments of this application can be combined with each other without conflict.

[0021] Figure 2 This is a flowchart illustrating a low-speed collision avoidance assist control method for vehicles provided in an embodiment of this application. Figure 2 As shown, the method may include: Step S210: When the vehicle is in a low-speed driving or waiting state, acquire the composite dual-frequency reflected sound wave reflected by the obstacle in front of the vehicle after receiving the detection sound wave emitted by the vehicle.

[0022] The composite dual-frequency reflected sound wave can include: a pulse echo component and a first echo component and a second echo component at each time point.

[0023] In practice, when the vehicle is traveling at low speed or stationary, multiple acoustic transducers deployed around the vehicle simultaneously emit three types of detection acoustic waves: a continuous detection acoustic wave at a first frequency, a continuous detection acoustic wave at a second frequency, and a short-time pulse detection acoustic wave. The difference between the first and second frequencies is a preset non-zero constant value, the duration of the short-time pulse is equal to a predetermined pulse duration, and the emission time of the short-time pulse is synchronized with the emission time of the continuous acoustic wave. The system receives composite echo signals reflected from obstacles in front of the vehicle within a preset time period and filters and separates these signals using three parallel digital bandpass filters to obtain a first echo component, a second echo component, and a pulse echo component. The first echo component corresponds to the first frequency, the second echo component corresponds to the second frequency, and the pulse echo component corresponds to the short-time pulse.

[0024] In one embodiment of this application, the method may further include: Read the current values ​​from the temperature and humidity sensor and the noise sensor to obtain the ambient temperature, ambient humidity and ambient noise; Calculate the current speed of sound based on the ambient temperature and humidity; where the ambient humidity refers to relative humidity; the formula for calculating the current speed of sound is as follows: ;in, Indicates the current speed of sound; Indicates ambient temperature; Indicates ambient temperature; Based on the current speed of sound, ambient noise, and configured reference transmission parameters, the initial transmission parameters are calculated. These reference transmission parameters may include: a first reference frequency, a second reference frequency, a reference transmission power, and a reference pulse duration. Specifically, the first and second reference frequencies are adjusted according to the current speed of sound to obtain the first and second initial frequencies, calculated using the following formulas: ; ;in, Indicates the first initial frequency; Indicates the first reference frequency; This represents the preset reference sound velocity, which is user-defined and can be 343 m / s, corresponding to 20℃ and 50% humidity. This represents the difference in reference frequencies, which can be 800Hz. The second initial frequency is indicated. The reference transmit power is adjusted based on the ambient noise to obtain the initial transmit power. The noise difference between the ambient noise and the reference noise is calculated, where the reference noise is the noise corresponding to the reference transmit power. The initial transmit power is calculated based on this noise difference and the configured linear compensation rule. The linear compensation rule may include: if the noise difference is greater than 0, the initial transmit power is calculated by multiplying the noise difference by a configuration coefficient (which can be 0.8). The sum of this product and the reference transmit power is determined as the initial transmit power. If the noise difference is not greater than 0, the initial transmit power is the sum of the calculated noise difference and the reference transmit power. The reference transmit power can be obtained by back-calculating the transducer sensitivity curve based on the target signal-to-noise ratio set according to the ambient noise. The formula for calculating the reference transmit power is as follows: , express; Indicates the minimum available power of the transducer; Indicates environmental noise; This represents the typical noise level in a quiet environment, either measured by the user in a quiet environment or preset. This indicates the preset target signal-to-noise ratio; if the reference transmit power exceeds the preset maximum transducer power, then the reference transmit power is set as the maximum transducer power; the reference pulse duration can be 0.5ms; the reference pulse duration is set as the initial pulse duration. The acoustic transducer is controlled to emit detection acoustic waves around the vehicle with initial emission parameters and receive echo signals. The quality of the echo signal is evaluated based on the configured quality scoring model to obtain a comprehensive quality score. The comprehensive quality score is equal to the weighted sum of the echo signal-to-noise ratio and the echo peak energy stability, minus the multipath interference intensity penalty term. A model-free optimization algorithm is used to iteratively adjust the initial transmission parameters within a preset range to maximize the output of the quality evaluation function, thus obtaining the current transmission parameters. The model-free optimization algorithm can be a Bayesian optimization algorithm or a random perturbation random approximation algorithm. In each iteration, a probe wave corresponding to the initial transmission parameters is transmitted, the echo is received, and the quality score is calculated. The internal model of the optimization algorithm is updated until the increment of the quality score is lower than the preset threshold three times consecutively or the maximum number of iterations is reached. The final set of transmission parameters obtained by optimization is used as the current transmission parameters under the current environment.

[0025] Step S220: Perform multi-domain interference feature analysis on the composite dual-frequency reflected sound wave to obtain multi-dimensional time-frequency features.

[0026] Among them, multidimensional time-frequency features can include: time delay features, continuous phase difference sequence, amplitude ratio sequence, spectral centroid sequence, and energy distribution width sequence.

[0027] In practice, the first echo component and the second echo component at each time point within a preset time period are quadrature demodulated to obtain the first complex baseband signal and the second complex baseband signal at each time point. Specifically, for any time point, i.e., the sampling point, the first echo component and the second echo component are simultaneously sent to two parallel quadrature demodulation channels: First channel: The local oscillator frequency is set to the first frequency, the input signal is multiplied and low-pass filtered, and the first in-phase component and the first quadrature component are output, which together constitute the first complex baseband signal; Second channel: The local oscillator frequency is set to the second frequency, the input signal is multiplied and low-pass filtered, and the second in-phase component and the second quadrature component are output, which together constitute the second complex baseband signal; The first complex baseband signal and the second complex baseband signal retain the amplitude and phase information of the original reflected echo, and the time sampling rate of each signal is synchronized with the transmitted pulse; Based on the first and second complex baseband signals at each time point, the difference frequency complex signal at each time point is calculated; the phase sequence of the difference frequency complex signal at each time point is extracted to obtain a continuous phase difference sequence; specifically, the first and second complex baseband signals are multiplied by their conjugates to obtain the difference frequency complex signal; for each sampling point of the difference frequency complex signal, its phase angle (angle value) is calculated to obtain a continuous phase difference sequence; the continuous phase difference sequence is used to reflect the relative phase change of two frequencies due to the change in propagation path length; The amplitudes of the first and second complex baseband signals at each time point are calculated to obtain the first and second echo amplitudes at each time point. The ratio of the first echo amplitude to the second echo amplitude at each time point is used as the amplitude ratio to obtain an amplitude ratio sequence. Specifically, for any time point, the square root of the sum of the squares of the first in-phase component and the first quadrature component of the first complex baseband signal is taken as the instantaneous amplitude of the first complex baseband signal, and this instantaneous amplitude is used as the first echo amplitude. The square root of the sum of the squares of the second in-phase component and the second quadrature component of the second complex baseband signal is taken as the instantaneous amplitude of the second complex baseband signal, and this instantaneous amplitude is used as the second echo amplitude. The ratio of the first echo amplitude to the second echo amplitude is used as the amplitude ratio for each sampling point to form an amplitude ratio sequence. This amplitude ratio sequence has characteristic responses to sound-absorbing materials, smooth surfaces, etc. A short-time Fourier transform is performed on the first complex baseband signal to obtain the time-frequency energy spectrum. Based on the time-frequency energy spectrum, the spectral centroid sequence and energy distribution width sequence are calculated. Specifically, for the first complex baseband signal (or the second complex baseband signal) at each time point, a short-time Fourier transform is performed with a fixed-length time window (e.g., 128 sampling points, with 50% overlap between windows) to obtain the spectral energy distribution corresponding to each time window. For any given time window, each frequency value is multiplied by its corresponding energy value, summed, and then divided by the total energy of all frequencies to obtain the spectral centroid of that window. The squared deviation of each frequency value from the spectral centroid is calculated, multiplied by the energy value of that frequency, summed, and then divided by the total energy to obtain the energy distribution width of that window. The spectral centroids and energy distribution widths of each time window are arranged in chronological order to obtain the spectral centroid sequence and energy distribution width sequence. The spectral centroid sequence and energy distribution width sequence are used to reflect the roughness and structural uniformity of the obstacle surface. For the pulse echo component, the time difference from transmission to reception of each pulse is measured as a time delay feature; and the peak amplitude of each pulse echo is extracted as a peak energy feature; specifically, for the pulse echo component, the transmission time of each short pulse is recorded, and the reflected echo of the pulse is searched in the received signal; the position with the largest peak amplitude in the echo is found, and the time difference between the peak position and the transmission time is measured, and this time difference is used as the time delay feature corresponding to each pulse.

[0028] Step S230: Determine the current distance between the vehicle and the obstacle, as well as the liveness attribute of the obstacle, based on the current sound speed and multi-dimensional time-frequency characteristics.

[0029] Among them, the living attribute can include living and non-living.

[0030] In one embodiment of this application, the method for determining the current distance may include: Calculate the pulse ranging value based on the time delay characteristics and the current speed of sound; specifically, ;in, Indicates the pulse ranging value; This represents the current speed of sound, which is calculated based on ambient temperature and humidity. This indicates the time delay characteristic, namely the time difference between the transmission and reception of a pulse; Based on the current sound speed, the continuous phase difference sequence, the first frequency, and the second frequency, the interferometric ranging value is calculated. Specifically, the initial cumulative phase difference is obtained by taking the dual-frequency phase difference of the continuous phase difference sequence at the initial time point (i.e., the first effective echo time) within a preset time period; the difference between the first frequency and the second frequency is taken as the dual-frequency difference value; the interferometric ranging value is calculated based on the initial cumulative phase difference and the dual-frequency difference value; the calculation formula for the interferometric ranging value is as follows: ;in, Indicates the interferometric ranging value; This represents the initial accumulated phase difference, based on the value of the continuous phase difference sequence at the initial time. Indicates the difference between the two frequencies; The initial distance is calculated based on the pulse ranging value and the interferometric ranging value. Specifically, the periodic ambiguity of the interferometric ranging value is eliminated using the pulse ranging value: an integer number of periods is calculated such that the sum of the integer period number multiplied by the unambiguous ranging range and the interferometric ranging value minimizes the absolute value of the difference between the pulse ranging value and the unambiguous ranging value. The initial absolute distance is obtained by multiplying the integer period number by the unambiguous ranging range and adding it to the interferometric ranging value. The initial distance calculation formula is as follows: ;in, Indicates the initial distance; Let n represent the candidate integers, and select the integer n that minimizes the absolute value of the difference; Represents a set of integers; The phase change rate is obtained by differentiating the continuous phase difference sequence; the phase change rate is multiplied by a preset conversion coefficient to obtain the radial velocity of the obstacle relative to the vehicle. Integrating the radial velocity over time yields the change in distance. The sum of the initial distance and the change in distance is used as the current distance.

[0031] In one embodiment of this application, the method for determining the living properties of an obstacle may include: Based on the continuous phase difference sequence and amplitude ratio sequence, a dual-frequency reflection joint feature map is constructed. Specifically, the continuous phase difference sequence and amplitude ratio sequence are normalized to the [0,1] interval respectively. The normalized continuous phase difference sequence and amplitude ratio sequence are stacked side by side along the time axis to obtain a dual-frequency two-dimensional feature map. The first row of the dual-frequency two-dimensional feature map is the normalized continuous phase difference sequence, and the second row is the normalized amplitude ratio sequence. Calculate the statistical characteristics of the spectral centroid sequence and the energy distribution width sequence respectively to obtain the first statistical characteristic and the second statistical characteristic; the first statistical characteristic may include: the mean and variance of the spectral centroid sequence; the first statistical characteristic may include: the mean and variance of the energy distribution width sequence; The dual-frequency reflection joint feature map, the first statistical feature, and the second statistical feature are input into a pre-trained obstacle material type classification model to obtain the material type of the obstacle; the material type may include: human skin, fabric, rigid metal, rigid plastic, sound-absorbing foam, rubber, branches / shrubs, or irregular rocks / concrete. Micro-motion energy extraction and time-frequency ridge feature extraction are performed on the continuous phase difference sequence to obtain micro-motion features and time-frequency ridge features. Specifically, the continuous phase difference sequence is input into a bandpass digital filter for filtering to obtain the filtered signal. The passband frequency range of this filter is 0.5 Hz to 5 Hz, corresponding to the micro-motion frequencies generated by the swaying of human or animal limbs and the rise and fall of the chest cavity. The Hilbert transform is performed on the filtered signal to obtain the analytic signal. The magnitude of the analytic signal is calculated by taking the square root of the sum of the squares of the real part and the squares of the imaginary part at each time point to obtain the envelope amplitude sequence. Each value in this sequence represents the instantaneous amplitude of the filtered signal at that time point. The arithmetic mean of the envelope amplitude sequence is calculated by summing the envelope amplitudes at all time points and then dividing by the total number of time points. This average value is the micro-motion characteristic. The larger the micro-motion characteristic, the more obvious the small periodic motion on the obstacle surface, and the more likely it is to be a living organism. A short-time Fourier transform is performed on the continuous phase difference sequence to obtain the time spectrum. The frequency value with the highest energy at each time point is extracted and used as the time-frequency ridge value at that time point. The time-frequency ridge values ​​at each time point are arranged in chronological order to obtain the time-frequency ridge sequence. The difference between adjacent time points in the time-frequency ridge sequence is calculated to obtain the ridge change rate sequence. The standard deviation of the ridge change rate sequence is calculated as the time-frequency ridge feature. Based on the current distance, time-frequency ridge features, obstacle material type, and micro-motion features, the liveness attribute of the obstacle is determined. Specifically, from a configured lookup table of different material types and different baseline thresholds, the baseline threshold corresponding to the obstacle's material type is matched. The baseline threshold for human skin can be 0.01, for fabric 0.02, for branches or shrubs 0.04, and for others 0.05. An adaptive threshold is then calculated based on the current distance and the baseline threshold. The formula for calculating the adaptive threshold is as follows: ;in, Indicates an adaptive threshold; Indicates the baseline threshold; This indicates the current distance. If the material type is human skin or fabric, and the micro-motion characteristic is greater than the adaptive threshold, then the obstacle's living attribute within this time period is determined to be a movable living being, i.e., a living being. If the material type is human skin or fabric, but the micro-motion energy index is not greater than the adaptive threshold, then it is determined to be a static non-living being, i.e., a non-living being. If the material type is tree branches or shrubs, and the time-frequency ridge feature is greater than 0.05, it is determined to be a static non-living entity. If the material type is tree branches or shrubs, and the time-frequency ridge feature is less than 0.2, and the micro-motion energy index is greater than 0.03, it is determined to be a mobile living entity. If the material type is tree branches or shrubs, but does not meet the above two conditions, it is determined to be a static non-living entity. If the material type is rigid metal, rigid plastic, or irregular rock concrete, it is directly determined to be a static non-living entity, without considering the micro-motion energy index. If the material type is sound-absorbing foam or rubber, or the confidence level of the material type is less than 0.6, the material probability vector, micro-motion feature, time-frequency ridge feature, and current distance for that time period are input into a pre-trained logistic regression classifier, which outputs the liveness probability. If the liveness probability is greater than 0.6, it is determined to be a mobile living entity; otherwise, it is determined to be a static non-living entity.

[0032] In one embodiment of this application, the obstacle material type classification model may include: The input layer is used to input the joint feature map of dual-frequency reflection, the first statistical feature, and the second statistical feature. The image processing layer is used to perform feature mapping on the joint feature map of dual-frequency reflections to obtain the image feature vector. Specifically, physical prior convolutional layers and deformable convolutional layers are used to extract spatiotemporal features from the joint feature map of dual-frequency reflections to obtain the first feature map. Attention weighting is applied to the first feature map to obtain the second feature map. Global average pooling is performed on the second feature map along the time axis to reduce the dimensionality and obtain the image feature vector. The statistical processing layer concatenates the first and second statistical features to obtain a statistical fusion vector; the statistical fusion vector is then subjected to batch normalization, fully connected transformation and Dropout operations to obtain a statistical feature vector. The fusion layer is used to concatenate the image feature vector with the statistical feature vector to obtain the fused vector. The classification layer is used to perform fully connected transformation and Softmax normalization on the fusion vector in sequence, and output the probability of eight material categories; The output layer is used to output the material category with the highest probability, which is the material type of the obstacle.

[0033] Step S240: Determine the vehicle's collision avoidance control strategy based on the current distance and the liveness attributes of the obstacles.

[0034] The collision avoidance control strategy may include: target braking force value, braking mode, and braking force application method.

[0035] In practice, the target distance threshold corresponding to the configured liveness attribute is matched from a lookup table of different liveness attributes and different distance thresholds; where, when the liveness attribute is live, the target distance threshold can be 0.8 meters; when the liveness attribute is not live, the target distance threshold can be 0.3 meters. If the current distance is greater than the target distance threshold, the braking conditions are not met, braking is not triggered, and a warning strategy is determined and a warning is issued based solely on the current distance. Specifically, the warning frequency is calculated based on the current distance; the warning frequency is inversely proportional to the distance: when the current distance is greater than 1.5 meters, the warning frequency is once per second; when the current distance is between 0.8 meters and 1.8 meters, the warning frequency increases linearly to five times per second as the distance decreases; when the current distance is less than 0.8 meters but braking has not yet been triggered, the warning frequency remains at five times per second; an intermittent warning tone is emitted through the vehicle's buzzer, with the pitch increasing as the frequency increases. If the current distance is not greater than the target distance threshold, the braking condition is determined, and the braking mode is determined based on the liveness attribute. Specifically, when the subject is identified as a living entity, the first collision avoidance control strategy is adopted: the braking mode is progressive braking mode; the target braking force value is determined based on the current distance; if the current distance is greater than 0.4 meters, the target braking force value is 30% of the configured maximum braking force; if the current distance is no greater than 0.4 meters but greater than 0.2 meters, the target braking force value is 60% of the maximum braking force; if the current distance is no greater than 0.2 meters, the target braking force value is the maximum braking force; the actual application of braking force adopts a linear ramp increase, with an increase rate of 50% of the maximum braking force per second, until the target braking force value is reached; at the same time, if the radial velocity is positive (moving away) and lasts for 0.1 seconds, the braking mode is exited in advance. When the liveness attribute is non-liveness, the second collision avoidance control strategy is adopted: the braking mode is step braking mode; the product of the maximum braking force and the configured braking force coefficient is used as the braking force target value; where the braking force coefficient can be 60%; the actual application of braking force reaches the target value within 50 milliseconds after triggering (step response); if the current distance is less than 0.1 meters, the braking force target value is forcibly increased to 100% of the maximum braking force, and the liveness attribute is not considered; A collision avoidance control strategy is generated based on the target braking force value, braking mode, and braking force application method.

[0036] In another embodiment of this application, the method further includes: The system monitors the rate of change of the current distance in real time. If, during braking, the distance begins to increase (i.e., the vehicle moves away from the obstacle) and the increase exceeds 0.05 meters, the braking force is gradually reduced, and the deceleration rate is proportional to the rate at which the current distance increases. If the current distance decreases to less than 0.05 meters, the maximum braking force will be forcibly maintained until the vehicle comes to a complete stop or the driver intervenes actively (such as pressing the accelerator pedal beyond the preset threshold).

[0037] Corresponding to the above method, embodiments of this application also provide a vehicle low-speed collision avoidance assist control device, such as... Figure 3 As shown, the device includes: The acquisition unit 310 is used to acquire the current speed of sound in the environment where the vehicle is located and the composite dual-frequency reflected sound waves reflected by obstacles in front of the vehicle after receiving the detection sound waves emitted by the vehicle when the vehicle is in a low-speed driving or waiting state. Analysis unit 320 is used to perform multi-domain interference feature analysis on composite dual-frequency reflected sound waves to obtain multi-dimensional time-frequency features; The first determining unit 330 is used to determine the current distance between the vehicle and the obstacle, as well as the liveness attribute of the obstacle, based on the current sound speed and multi-dimensional time-frequency characteristics. The second determining unit 340 is used to determine the vehicle's collision avoidance control strategy based on the current distance and the liveness attributes of the obstacle.

[0038] The functions of each functional unit of the vehicle low-speed collision avoidance assist control device provided in the above embodiments of this application can be implemented through the above methods and steps. Therefore, the specific working process and beneficial effects of each unit in the vehicle low-speed collision avoidance assist control device provided in the embodiments of this application will not be repeated here.

[0039] This application also provides an electronic device, such as... Figure 4 As shown, it includes a processor 410, a communication interface 420, a memory 430, and a communication bus 440, wherein the processor 410, the communication interface 420, and the memory 430 communicate with each other through the communication bus 440.

[0040] Memory 430 is used to store computer programs; When the processor 410 executes the program stored in the memory 430, it performs the following steps: When the vehicle is traveling at low speed or in a standby state, the current speed of sound in the environment where the vehicle is located and the composite dual-frequency reflected sound waves reflected by obstacles in front of the vehicle after receiving the detection sound waves emitted by the vehicle are obtained. Multi-domain interference feature analysis was performed on the composite dual-frequency reflected sound wave to obtain multi-dimensional time-frequency features; Based on the current speed of sound and multi-dimensional time-frequency characteristics, determine the current distance between the vehicle and the obstacle, as well as the liveness attribute of the obstacle; Based on the current distance and the liveness attributes of the obstacles, a collision avoidance control strategy for the vehicle is determined.

[0041] The communication bus mentioned above can be a Peripheral Component Interconnect (PCI) bus or an Extended Industry Standard Architecture (EISA) bus, etc. This communication bus can be divided into address bus, data bus, control bus, etc. For ease of illustration, only one thick line is used to represent it in the diagram, but this does not mean that there is only one bus or one type of bus.

[0042] The communication interface is used for communication between the aforementioned electronic devices and other devices.

[0043] The memory may include random access memory (RAM) or non-volatile memory (NVM), such as at least one disk storage device. Optionally, the memory may also be at least one storage device located remotely from the aforementioned processor.

[0044] The processors mentioned above can be general-purpose processors, including central processing units (CPUs), network processors (NPs), etc.; they can also be digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components.

[0045] The implementation methods and beneficial effects of the various components of the electronic device in the above embodiments for solving the problem can be found in [reference needed]. Figure 2 The steps in the illustrated embodiments are used to implement the electronic device. Therefore, the specific working process and beneficial effects of the electronic device provided in this application will not be repeated here.

[0046] In another embodiment provided in this application, a computer-readable storage medium is also provided, which stores instructions that, when executed on a computer, cause the computer to perform any of the vehicle low-speed collision avoidance assist control methods described in the above embodiments.

[0047] In another embodiment provided in this application, a computer program product containing instructions is also provided, which, when run on a computer, causes the computer to execute any of the vehicle low-speed collision avoidance auxiliary control methods described in the above embodiments.

[0048] Those skilled in the art will understand that the embodiments in this application can be provided as methods, systems, or computer program products. Therefore, the embodiments in this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the embodiments in this application can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0049] This application describes embodiments of methods, apparatus (systems), and computer program products according to embodiments of this application with reference to flowchart illustrations and / or block diagrams. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0050] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0051] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0052] Although preferred embodiments have been described in this application, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments as well as all changes and modifications falling within the scope of the embodiments of this application.

[0053] Obviously, those skilled in the art can make various modifications and variations to the embodiments of this application without departing from the spirit and scope of the embodiments of this application. Therefore, if these modifications and variations to the embodiments of this application fall within the scope of this application and its equivalents, then these modifications and variations are also intended to be included in the embodiments of this application.

Claims

1. A vehicle low-speed collision avoidance auxiliary control method, characterized in that, The method includes: When the vehicle is traveling at low speed or in a standby state, the current speed of sound in the environment where the vehicle is located and the composite dual-frequency reflected sound waves reflected by the obstacles in front of the vehicle after receiving the detection sound waves emitted by the vehicle are obtained. Multi-domain interference feature analysis was performed on the composite dual-frequency reflected sound wave to obtain multi-dimensional time-frequency features; Based on the current speed of sound and the multidimensional time-frequency characteristics, determine the current distance between the vehicle and the obstacle, as well as the liveness attribute of the obstacle; Based on the current distance and the liveness attribute of the obstacle, a collision avoidance control strategy for the vehicle is determined.

2. The method as described in claim 1, characterized in that, The method further includes: The ambient temperature, humidity, and noise levels of the vehicle's current environment are obtained. Calculate the current speed of sound based on the ambient temperature and humidity; Calculate the initial transmission parameters based on the current sound speed, the ambient noise, and the configured reference transmission parameters; The vehicle's acoustic transducer is controlled to emit detection acoustic waves around the vehicle according to the initial emission parameters and receive echo signals. The echo signal is evaluated for quality based on the configured quality scoring model to obtain a comprehensive quality score. A model-free optimization algorithm is used to iteratively optimize the initial launch parameters within a preset launch parameter range based on the comprehensive quality score to obtain the current launch parameters; The acoustic transducer is controlled to emit detection acoustic waves around the vehicle with the current emission parameters, and to receive composite dual-frequency reflected acoustic waves reflected by obstacles in front of the vehicle after receiving the detection acoustic waves.

3. The method as described in claim 2, characterized in that, The composite dual-frequency reflected sound wave includes a pulse echo component and a first echo component and a second echo component at each time point. The multidimensional time-frequency features include: a continuous phase difference sequence and an amplitude ratio sequence; Multi-domain interference feature analysis was performed on the composite dual-frequency reflected sound wave to obtain multi-dimensional time-frequency features, including: The first echo component and the second echo component at each time point are quadrature demodulated to obtain the first complex baseband signal and the second complex baseband signal at each time point. Calculate the difference frequency complex signal at each time point based on the first complex baseband signal and the second complex baseband signal at each time point; Extract the phase sequence of the difference frequency complex signal at each time point to obtain a continuous phase difference sequence; The amplitudes of the first complex baseband signal and the second complex baseband signal at each time point are calculated respectively to obtain the first echo amplitude and the second echo amplitude at each time point; An amplitude ratio sequence is generated based on the ratio of the first echo amplitude to the second echo amplitude at each time point.

4. The method as described in claim 3, characterized in that, The multidimensional time-frequency features also include: time delay features and the energy distribution width sequence of the spectral centroid sequence; The multi-domain interference feature analysis of the composite dual-frequency reflected sound wave yields multi-dimensional time-frequency features, including: Short-time Fourier transform is performed on the first complex baseband signal at each time point to obtain the time-frequency energy spectrum at each time point; Calculate the spectral centroid sequence and energy distribution width sequence based on the time-frequency energy spectrum at each time point; For the pulse echo component, the time difference from transmission to reception of each pulse is obtained as a time delay feature.

5. The method as described in claim 4, characterized in that, The current transmission parameters include: a first frequency corresponding to the first echo component and a second frequency corresponding to the second echo component; Determining the current distance between the vehicle and the obstacle based on the current speed of sound and the multidimensional time-frequency characteristics includes: Calculate the pulse ranging value based on the time delay characteristics and the current sound speed; Calculate the interferometric ranging value based on the current sound speed, the continuous phase difference sequence, the first frequency, and the second frequency; Calculate the initial distance based on the pulse ranging value and the interferometric ranging value; The continuous phase difference sequence is differentiated to obtain the phase change rate; the product of the phase change rate and the preset conversion coefficient is taken as the radial velocity of the obstacle relative to the vehicle. Integrating the radial velocity over time yields the distance change. The current distance is obtained based on the initial distance and the change in distance.

6. The method as described in claim 4, characterized in that, The method for determining the living properties of the obstacle includes: Based on the continuous phase difference sequence and the amplitude ratio sequence, a dual-frequency reflection joint feature map is constructed; The statistical characteristics of the spectral centroid sequence and the energy distribution width sequence are calculated respectively to obtain the first statistical characteristic and the second statistical characteristic; The dual-frequency reflection joint feature map, the first statistical feature, and the second statistical feature are input into a pre-trained obstacle material type classification model to obtain the material type of the obstacle; The continuous phase difference sequence is subjected to micro-motion energy extraction and time-frequency ridge feature extraction to obtain micro-motion features and time-frequency ridge features; Based on the current distance, the time-frequency ridge feature, the material type of the obstacle, and the micro-motion feature, the living attribute of the obstacle is determined.

7. The method as described in claim 1, characterized in that, The living organism attribute includes: living organism and non-living organism; Based on the current distance and the liveness attribute of the obstacle, a collision avoidance control strategy for the vehicle is determined, including: Match the target distance threshold corresponding to the configured liveness attribute from the lookup table of different liveness attributes and different distance thresholds; If the current distance is greater than the target distance threshold, then a warning strategy is determined based on the current distance, and a warning is issued. If the current distance is not greater than the target distance threshold and the liveness attribute indicates that the subject is alive, then the configured first anti-collision control strategy is used for braking. If the current distance is not greater than the target distance threshold and the liveness attribute is non-liveness, then the configured second anti-collision control strategy is used for braking.

8. A vehicle low-speed collision avoidance auxiliary control device, characterized in that, The device includes: The acquisition unit is used to acquire the current speed of sound in the environment where the vehicle is located and the composite dual-frequency reflected sound waves reflected by obstacles in front of the vehicle after receiving the detection sound waves emitted by the vehicle when the vehicle is in a low-speed driving or waiting state. The analysis unit is used to perform multi-domain interference feature analysis on the composite dual-frequency reflected sound wave to obtain multi-dimensional time-frequency features; The first determining unit is configured to determine the current distance between the vehicle and the obstacle, as well as the liveness attribute of the obstacle, based on the current sound speed and the multi-dimensional time-frequency characteristics. The second determining unit is used to determine the collision avoidance control strategy of the vehicle based on the current distance and the liveness attribute of the obstacle.

9. An electronic device, characterized in that, The electronic device includes a processor, a communication interface, a memory, and a communication bus, wherein the processor, the communication interface, and the memory communicate with each other through the communication bus; Memory, used to store computer programs; A processor, when executing a program stored in memory, implements the method of any one of claims 1-7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed by a processor, implements the method described in any one of claims 1-7.

Images