A deceleration strip feature calculation method and system based on millimeter-wave radar

Abstract

This application relates to the field of vehicle assisted driving technology, and in particular, to a method and system for calculating deceleration belt features based on millimeter wave radar. The calculation method includes: receiving an intermediate frequency signal; analyzing the spectrum of the intermediate frequency signal to determine the shortest reflection path corresponding to the peak spectral line in the spectrum, and determining the frequency calculation value and spatial spectrum of the peak spectral line in the spectrum; determining a high-precision distance calculation value of the core position point according to the frequency calculation value, and determining a high-precision angle calculation value of the core position point according to the spatial spectrum; repeating the above steps to determine the high-precision distance calculation values and high-precision angle calculation values of a number of consecutive core position points; establishing a deceleration belt feature calculation model, and calculating the features of the deceleration belt according to the deceleration belt feature calculation model. This application improves the accuracy and stability of deceleration belt feature calculation through a deceleration belt feature calculation method based on the shortest reflection path.

Description

Technical Field

[0001] This application relates to the technical field of vehicle assisted driving, and particularly to a method and system for calculating deceleration strip features based on millimeter wave radar. Background Art

[0002] Deceleration strips are essential traffic facilities on urban roads, which can make drivers slow down. Before the vehicle reaches the deceleration strip, even if the driver makes a deceleration operation, the vehicle will still experience varying degrees of bumps when passing over the deceleration strip, which will bring an unpleasant driving and riding experience to the driver and passengers.

[0003] With the development of technology, the increasing demand for vehicle ride comfort has promoted the development of controllable suspension technology; the suspension system of a vehicle can alleviate the bumps caused by the road surface. However, traditional passive suspensions can only achieve the best performance under specific road conditions and cannot effectively cope with different road conditions. With the development of controllable suspension technology, active suspension systems and semi-active suspension systems have emerged, which can adjust the stiffness or damping coefficient of the suspension according to the road surface information captured by sensors according to a certain control strategy, enabling the vehicle to have good ride comfort under different road conditions. Among them,

[0004] The semi-active suspension has the advantages of small volume, low cost, and short response time. The semi-active suspension system based on magnetorheological dampers can not only achieve continuously adjustable damping, but the response time can even reach a few milliseconds. Non-contact perception of the deceleration strip that the vehicle is about to encounter can enable the semi-active suspension system to make effective adjustments for the deceleration strip, allowing the vehicle to have good ride comfort when passing over the deceleration strip.

[0005] However, since the semi-active suspension system generally uses an accelerometer built into the suspension to collect acceleration signals, this method is mainly used for perceiving different grades of road surfaces. Since the road surface grade generally remains consistent on a section of the road surface, the semi-active suspension system can make adjustments for different road surface grades. However, in actual road conditions, there are also discrete impact road surfaces containing deceleration strips. This method can only identify them after the wheels pass over the deceleration strip, and it is too late for the semi-active suspension system to make a response at this time. Currently, road surface recognition technology can pre-scan the road surface in front of the vehicle through methods such as cameras or lidar. This method does not rely on existing road surface information, and the vehicle can independently judge whether there is a deceleration strip in the front road and determine the spatial position of the deceleration strip. However, to enable the semi-active suspension system to make corresponding adjustments for the deceleration strip, it is necessary to rely on the important features of the deceleration strip, including the height and width of the deceleration strip. After obtaining the features of the deceleration strip, according to the corresponding semi-active suspension control strategy, the vehicle can still ensure the optimal ride comfort when passing over the deceleration strip.

[0006] Thus, cameras, lidar, millimeter-wave radars, and ultrasonic radars installed on the vehicle can provide the vehicle with non-contact perception capabilities of the surrounding environment. Cameras can achieve good quantitative analysis capabilities, but are affected by visual conditions and cannot work properly in environments with low visibility such as at night and in foggy weather. Although lidar can provide reliable and accurate results, its high cost limits its widespread application under cost constraints. Ultrasonic radars are restricted by the slow propagation speed of sound signals and cannot guarantee performance during the high-speed movement of the vehicle.

[0007] In summary, in the existing technologies for non-contact perception of speed bumps, there are problems of unstable performance, poor adaptability, and high cost. Summary of the Invention

[0008] This application provides a speed bump feature calculation method and system based on millimeter-wave radar, which can solve the problems of unstable performance, poor adaptability, and high cost in the existing technologies for non-contact perception of speed bumps.

[0009] The first technical solution of this application is a speed bump feature calculation method based on millimeter-wave radar. The millimeter-wave radar is installed on the front chassis of the vehicle and emits an initial signal to the road surface, and the method includes:

[0010] S1: Receive the intermediate-frequency signal obtained after the initial signal is reflected by several position points on the road surface during the vehicle's driving process;

[0011] S2: Analyze the spectrum of the intermediate-frequency signal, determine the shortest reflection path corresponding to the peak spectral line in the spectrum, and determine the frequency calculation value and spatial spectrum of the peak spectral line in the spectrum;

[0012] S3: Determine the high-precision distance calculation value of the core position point based on the shortest reflection path according to the frequency calculation value, and determine the high-precision angle calculation value of the core position point according to the spatial spectrum;

[0013] S4: Continuously and repeatedly execute the steps S1 to S3 during the vehicle's driving process, and respectively determine the high-precision distance calculation values and high-precision angle calculation values of several consecutive core position points during the vehicle's driving process;

[0014] S5: Establish a speed bump feature calculation model according to the high-precision distance calculation values and high-precision angle calculation values of several consecutive core position points, and calculate the features of the speed bump according to the speed bump feature calculation model.

[0015] Optionally, the antenna array of the millimeter-wave radar is set in a two-transmit and four-receive MIMO form and can virtualize eight receiving antennas;

[0016] And, step S2 includes:

[0017] S21: Sampling the intermediate frequency signal to obtain a discrete time series. The expression of the discrete time series is as follows:

[0018]

[0019] where, x IF (n) represents the discrete time series; N represents the number of sampling points; n represents the sampling sequence number; A represents the amplitude of the intermediate frequency signal; f s represents the sampling rate; f IF represents the frequency of the intermediate frequency signal, represents the initial phase of the intermediate frequency signal;

[0020] S22: Performing spectrum analysis on the discrete time series through FFT to obtain an amplitude spectrum, and determining the amplitude value, spectral line number, and the shortest reflection path corresponding to the peak spectral line in the amplitude spectrum, and determining the calculated frequency value of the peak spectral line according to the spectral line number. The expression of the calculated frequency value is as follows:

[0021]

[0022] where, f coarse represents the calculated frequency value; m represents the spectral line number of the peak spectral line;

[0023] S23: Performing zero-padding processing on the amplitude value of the peak spectral line to obtain a 256-point discrete sequence, and determining the spatial spectrum of the peak spectral line according to the 256-point discrete sequence. The expression of the spatial signal frequency resolution is as follows:

[0024]

[0025] where, Q represents the number of points of the angular FFT; l represents the distance between receiving antennas.

[0026] Optionally, step S3 includes:

[0027] S31: Performing error elimination processing on the calculated frequency value through an iterative method based on DFT interpolation to obtain a high-precision calculated frequency value, and determining a high-precision distance calculated value of the core position point based on the shortest reflection path according to the high-precision calculated frequency value;

[0028] S32: Determining the spatial frequency corresponding to the maximum amplitude value in the spatial spectrum, and determining a high-precision angle calculated value of the core position point based on the shortest reflection path according to the spatial frequency corresponding to the maximum amplitude value.

[0029] Optionally, step S31 includes:

[0030] S311: Calculate the DFT of single points located on both sides of the peak spectral line. The expressions of the discrete-time sequences of the two single points are X IF (m + 1 / 2) and X IF (m - 1 / 2);

[0031] S312: Substitute the expressions X IF (m + 1 / 2) and X IF (m - 1 / 2) into the following formula: Obtain the calculation formula for the frequency calculation error, and the formula is as follows:

[0032]

[0033] In the formula, represents the frequency calculation error;

[0034] And, the calculation formula for the high-precision frequency calculation value is as follows:

[0035]

[0036] In the formula, f fine represents the high-precision frequency calculation value;

[0037] S313: Determine the number of iterations α, assign 1 to the iteration control variable i and assign 0 to , and calculate and

[0038]

[0039] S314: Calculate the frequency calculation error according to the following formula:

[0040]

[0041] S315: Judge the magnitude relationship between the number of iterations α and the control variable i. If i ≤ α, increase the iteration control variable from i to i + 1, and iteratively execute the step S314;

[0042] S316: If i > α, calculate the high-precision frequency calculation value through the following formula:

[0043]

[0044] S317: Determine the high-precision distance calculation value of the core position point based on the shortest reflection path according to the high-precision frequency calculation value and the following formula;

[0045]

[0046] In the formula, c represents the speed of light; T c represents the duration of the chirp; B represents the bandwidth of the chirp.

[0047] Optionally, the step S32 includes:

[0048] S321: Determine the spatial frequency corresponding to the maximum amplitude value in the spatial spectrum and determine the high-precision angle calculation value of the core position point based on the shortest reflection path according to the spatial frequency corresponding to the maximum amplitude value and the following formula:

[0049] θ = sin -1 (λf spatial );

[0050] In the formula, f spatial represents the spatial frequency corresponding to the maximum amplitude value in the spatial spectrum; θ represents the high-precision angle calculation value; λ represents the wavelength.

[0051] Optionally, the characteristics of the speed bump include: the height and width of the speed bump;

[0052] And, the step S5 includes:

[0053] S51: Establish a speed bump characteristic calculation model according to the high-precision distance calculation values and high-precision angle calculation values of several consecutive core position points;

[0054] S52: Determine the high-precision distance trajectory and high-precision angle trajectory respectively according to the speed bump characteristic calculation model;

[0055] S53: Calculate the characteristics of the speed bump according to the high-precision distance trajectory and high-precision angle trajectory, and according to the following formula:

[0056] H = d - d hT ;

[0057] W = 2hl;

[0058] In the formula, H represents the height of the speed bump; W represents the width of the speed bump; d represents the high-precision distance calculation value between the millimeter-wave radar and the road surface when the millimeter-wave radar has not passed over the speed bump; d hT represents the high-precision distance calculation value when the millimeter-wave radar reaches the vertex of the speed bump.

[0059] The second technical solution of the present application is a speed bump characteristic calculation system based on a millimeter-wave radar, including:

[0060] A radar module for transmitting an initial signal and receiving an intermediate-frequency signal at several consecutive core position points;

[0061] A calculation module, configured to respectively determine a frequency calculation value and a spatial frequency spectrum of a peak spectral line in the intermediate frequency signals of a plurality of consecutive core position points according to the intermediate frequency signals of the plurality of consecutive core position points;

[0062] It is also configured to respectively determine high-precision distance calculation values of a plurality of consecutive core position points according to the frequency calculation values of the plurality of consecutive core position points, and respectively determine high-precision angle calculation values of the plurality of consecutive core position points according to the spatial frequency spectra of the plurality of consecutive core position points;

[0063] Moreover, it is also configured to establish a speed bump feature calculation model according to the high-precision distance calculation values and high-precision angle calculation values of a plurality of consecutive core position points, and calculate the features of the speed bump according to the speed bump feature calculation model.

[0064] Beneficial effects:

[0065] (1) A method for calculating speed bump features based on a millimeter-wave radar:

[0066] (1) By proposing the shortest reflection path of millimeter-wave signals, the scientificity of speed bump feature calculation is improved. When the millimeter-wave radar moves on the trajectory, the distance-angle information is accurately obtained according to the shortest reflection path of the millimeter-wave signals.

[0067] (2) By using a high-precision distance-angle calculation method, the accuracy of speed bump feature calculation is improved. Through a high-precision distance calculation algorithm based on DFT and a high-precision angle calculation algorithm based on MIMO, the distance-angle measurement accuracy of the millimeter-wave radar is improved, and high-precision distance-angle calculation values corresponding to the shortest reflection path are obtained.

[0068] (3) By determining a speed bump feature calculation method based on the shortest reflection path, the convenience and stability of speed bump feature calculation are improved. After obtaining high-precision distance-angle calculation values at different positions on the surface of the speed bump, Hampel filtering is used for outlier detection and removal. The target is determined as a speed bump through the Euclidean distance, and then the height and width of the speed bump are calculated respectively, improving the calculation efficiency, so it has strong stability.

[0069] (4) Strong adaptability. Compared with existing technologies such as lidar, the millimeter-wave radar in this application can still work normally even when blocked. In addition, if the chassis of the vehicle is contaminated with a certain degree of road debris, including mud and snow, in practice, this situation will affect the use of lidar, etc., but it will not affect the use of the millimeter-wave radar. Therefore, it has the problem of strong adaptability.

[0070] (5) High cost performance. The present application calculates the characteristics of the speed bump through a millimeter-wave radar. Compared with existing technologies such as lidar, the millimeter-wave radar is affordable and can accurately calculate the characteristics of the speed bump through this method. Therefore, it has the characteristic of high cost performance.

[0071] (2) A speed bump feature calculation system based on a millimeter-wave radar

[0072] (6) Through a speed bump feature calculation system based on a millimeter-wave radar, the height and width of the speed bump can be effectively obtained before the front wheels of the vehicle reach the speed bump, and corresponding speed bump data can be provided to the semi-active suspension system in a timely manner, facilitating the semi-active suspension system to make corresponding adjustments, so that the vehicle can still ensure the optimal riding comfort when passing through the speed bump, bringing a good driving experience to the vehicle user.

[0073] In summary, the present application can solve the problems of unstable performance, poor adaptability, and high cost existing in the prior art of sensing speed bumps through non-contact methods. BRIEF DESCRIPTION OF THE DRAWINGS

[0074] In order to more clearly illustrate the technical solutions of the present application, the drawings required for use in the embodiments will be briefly introduced below. Obviously, for those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0075] Figure 1 It is a schematic diagram of the relative position change between the vehicle and the speed bump during the driving process in the embodiment of the present application;

[0076] Figure 2 It is a schematic flow diagram of a speed bump feature calculation method based on a millimeter-wave radar in the embodiment of the present application;

[0077] Figure 3 It is a schematic diagram of the change of the shortest reflection path when the vehicle passes through the speed bump during the driving process in the embodiment of the present application;

[0078] Figure 4 It is a schematic diagram of the speed bump features in the embodiment of the present application;

[0079] Figure 5 It is a schematic diagram of the speed bump feature calculation model in the embodiment of the present application;

[0080] Figure 6 It is a schematic structural diagram of a speed bump feature calculation system based on a millimeter-wave radar in the embodiment of the present application;

[0081] Figure 7 It is a schematic diagram of the error comparison between using the speed bump feature calculation system and the laser rangefinder to calculate the speed bump features in the embodiment of the present application;

[0082] Figure 8 This is a high-precision distance comparison diagram of the speed bump and the cardboard box in the embodiment of the present application;

[0083] Figure 9 This is a high-precision angle comparison diagram of the speed bump and the cardboard box in the embodiment of the present application;

[0084] Figure 10 This is an error schematic diagram of calculating the speed bump characteristics at different measurement intervals in the embodiment of the present application;

[0085] Figure 11 This is an error schematic diagram of calculating the speed bump characteristics of speed bumps of different models in the embodiment of the present application;

[0086] Figure 12 This is an error schematic diagram of calculating the speed bump characteristics at different vehicle speeds in the embodiment of the present application;

[0087] Among them, radar module - 1; calculation module - 2. Specific embodiments

[0088] The embodiments will be described in detail below, and the examples are shown in the drawings. When the following description refers to the drawings, unless otherwise indicated, the same numbers in different drawings represent the same or similar elements. The embodiments described in the following examples do not represent all embodiments consistent with the present application. They are merely examples of systems and methods consistent with some aspects of the present application detailed in the claims. (1)

[0090] The first technical solution of the present application is a method for calculating speed bump characteristics based on a millimeter-wave radar. The millimeter-wave radar is installed on the front chassis of the vehicle and emits an initial signal to the road surface.

[0091] Specifically, compared with optical sensors such as cameras and lidar, the millimeter-wave radar is not affected by visual conditions, has a strong ability to penetrate fog and dust, can work normally at night, and has the characteristics of all-day and all-weather operation. In addition, it has the advantage of low cost. Compared with ultrasonic radars, the millimeter-wave radar has the characteristics of small size, light weight, and high spatial resolution.

[0092] Therefore, the millimeter-wave radar can perform non-contact sensing of speed bumps at a sufficiently low cost in a wide range of scenarios. Millimeter-wave radars have been widely arranged on vehicles, and the functions that can be achieved include adaptive cruise control, forward collision warning, and automatic emergency braking. Researchers have also proposed a large number of applications based on commercial millimeter-wave radars, including such applications on vehicles and such applications in indoor environments. The embodiment of the present application proposes a method for calculating speed bump characteristics based on a millimeter-wave radar.

[0093] In the embodiments of the present application, the millimeter-wave radar is used to calculate the characteristics of the speed bump in a non-contact manner, and the height and width of the speed bump are obtained. The millimeter-wave radar is placed at the very front end of the central axis of the vehicle chassis, and the antenna is facing the road surface directly. During the driving of the vehicle, when the road surface is in a flat state, the distance-angle information obtained by the millimeter-wave radar remains unchanged. When the millimeter-wave radar passes over the speed bump, the appearance of the speed bump will cause changes in the distance-angle information obtained by the millimeter-wave radar. Based on this, the embodiments of the present application can determine whether the vehicle is about to pass over the speed bump, calculate the characteristics of the speed bump, and finally obtain the height and width of the speed bump.

[0094] In practical applications, due to the high-frequency characteristics of millimeter waves, the incident signal will form a specular reflection on the surface of the object, that is, the angle between the incident signal and the normal is equal to the angle between the reflected signal and the normal. The transmitting antenna and the receiving antenna of the millimeter-wave radar are in a juxtaposed state. To receive a reflected signal with sufficient intensity at the receiving end, the antenna of the millimeter-wave radar needs to face the detection target directly. In many ADAS application scenarios, the millimeter-wave radar is placed at the front of the vehicle, with the transmitting antenna facing the front of the vehicle. This arrangement is beneficial for the millimeter-wave radar to detect the targets in front of the vehicle, but the perception ability of the road surface is very weak. When the transmitting antenna of the millimeter-wave radar is perpendicular to the detection target, the reflected signal cannot return to the receiving antenna due to specular reflection. Correspondingly, when the millimeter-wave radar is in the front position of the vehicle, the reflected signal from the road surface cannot return to the receiving antenna. Considering that the road surface is rough, part of the signal will return to the receiving antenna due to diffuse reflection, but the intensity of this part of the echo signal is often very weak and is not sufficient to realize the recognition and perception of the road surface. Since the detection target in the embodiments of the present application is the speed bump, and the speed bump is part of the road surface, the millimeter-wave radar is placed on the chassis of the vehicle, with the antenna of the millimeter-wave radar facing the road surface directly. In order to allow the front wheels of the vehicle to have sufficient time to complete the calculation of the speed bump characteristics before reaching the speed bump, the embodiments of the present application finally place the millimeter-wave radar at the very front of the central axis of the vehicle chassis, as Figure 1 shown Figure 1 is a schematic diagram of the relative position change between the vehicle and the speed bump during the driving process of the vehicle in the embodiments of the present application. During the process of the vehicle driving from the Figure 1 solid line position to the dotted line position in, by continuously collecting data through the millimeter-wave radar, the speed bump characteristic perception system can obtain the height and width of the speed bump.

[0095] As Figure 2 shown Figure 2 is a schematic flowchart of a method for calculating the characteristics of a speed bump based on a millimeter-wave radar in the embodiments of the present application. The calculation method includes:

[0096] S1: Receive the intermediate-frequency signal obtained after the initial signal is reflected by several position points on the road surface during the driving of the vehicle.

[0097] Specifically, as Figure 3 shown Figure 3 is a schematic diagram of the change in the shortest reflection path when the vehicle passes over a speed bump during driving in the embodiment of the present application. When the millimeter-wave radar is above the road surface, the road surface is simplified to a plane. As long as the shortest reflection path among all specular reflection paths is determined, the distance between the radar and the plane can be obtained. By performing spectral analysis on the initial intermediate-frequency signal obtained after the ground reflection of the initial signal, the frequency corresponding to the maximum amplitude corresponds to the shortest reflection path. At this time, the shortest reflection path is perpendicular to the transmitting antenna of the millimeter-wave radar and also perpendicular to the plane, as Figure 3 shown. When the millimeter-wave radar passes over the speed bump, the specular reflection formed by the initial signal is completely different from that when it is above the road surface. After simplifying the surface of the speed bump to a curved surface, the shortest reflection path of the initial signal on the curved surface is determined.

[0098] Through the idea of calculus, the curved surface can be divided into countless tiny planes, and each plane forms a different angle with the transmitting antenna of the millimeter-wave radar. When the transmitting antenna of the millimeter-wave radar forms a certain angle with the target, the shortest reflection path is still perpendicular to the target surface, but forms a certain angle with the transmitting antenna of the millimeter-wave radar. When the millimeter-wave radar is at different positions on the trajectory above the speed bump, the shortest reflection path formed by the initial signal on the surface of the speed bump has a direction that is the normal direction at the corresponding point on the surface of the speed bump.

[0099] S2: Analyze the spectrum of the intermediate-frequency signal, determine the shortest reflection path corresponding to the peak spectral line in the spectrum, and determine the frequency calculation value and spatial spectrum of the peak spectral line in the spectrum.

[0100] Specifically, the antenna array of the millimeter-wave radar is set in a two-transmit and four-receive MIMO form and can virtualize eight receiving antennas.

[0101] And, step S2 includes:

[0102] S21: Perform sampling processing on the intermediate-frequency signal to obtain a discrete-time sequence. The expression of the discrete-time sequence is as follows:

[0103]

[0104] In the formula, x IF (n) represents the discrete-time sequence; N represents the number of sampling points; n represents the sampling sequence number; A represents the amplitude of the intermediate-frequency signal; f s represents the sampling rate; f IF represents the frequency of the intermediate-frequency signal, represents the initial phase of the intermediate-frequency signal;

[0105] S22: Perform spectral analysis on the discrete time series through FFT to obtain the amplitude spectrum, determine the amplitude value, spectral line number, and the shortest reflection path corresponding to the peak spectral line in the amplitude spectrum, and determine the calculated frequency value of the peak spectral line according to the spectral line number. The expression of the calculated frequency value is as follows:

[0106]

[0107] In the formula, f coarse represents the calculated frequency value; m represents the spectral line number of the peak spectral line;

[0108] S23: Perform zero-padding processing on the amplitude value of the peak spectral line to obtain a 256-point discrete sequence and determine the spatial spectrum of the peak spectral line according to the 256-point discrete sequence. The expression of the spatial signal frequency resolution is as follows:

[0109]

[0110] In the formula, Q represents the number of points of the angular FFT; l represents the distance between receiving antennas.

[0111] Specifically, based on the hardware of the existing millimeter-wave radar, the embodiments of the present application meet the requirements of high angular resolution in the form of a MIMO radar. Assume that the number of transmitting antennas and receiving antennas of the millimeter-wave radar are N TX and N RX , compared with the traditional single-transmitter multi-receiver radar, the MIMO radar can expand the antenna aperture by N TX times through a virtual array without changing the number of receiving antennas, thereby improving the angular resolution of the millimeter-wave radar. Setting the antenna array of the millimeter-wave radar in the form of a two-transmitter four-receiver MIMO can double the antenna aperture, and the corresponding angular resolution also doubles. The radar signal modulation method of the TDM-MIMO form is simpler, requires less hardware resources, and can achieve very ideal orthogonality in theory. Therefore, the embodiments of the present application adopt the orthogonal form of TDM-MIMO to transmit chirp signals.

[0112] If the signal is transmitted in the two-transmitter four-receiver TDM-MIMO mode, the millimeter-wave radar can virtualize 8 receiving antennas. At this time, the angular resolution of the millimeter-wave radar at a 0-degree viewing angle is about 14.5 degrees.

[0113] The amplitude values of the 8 virtual antennas of the millimeter-wave radar at the peak spectral line m after range FFT, after zero-padding, obtain a 256-point discrete sequence. The spatial spectrum can be obtained through a 256-point FFT, and the angular resolution is improved to within 1 degree, meeting the requirements of the deceleration belt feature calculation for the angle measurement accuracy. Corresponding to the frequency resolution of the time-domain signal, the expression of the spatial signal frequency resolution is as follows:

[0114]

[0115] Wherein, Q represents the number of points of the angular FFT; l represents the distance between receiving antennas.

[0116] S3: Determine the high-precision distance calculation value of the core position point based on the shortest reflection path according to the calculated frequency value, and determine the high-precision angle calculation value of the core position point according to the spatial spectrum.

[0117] Wherein, step S3 includes:

[0118] S31: Perform error elimination processing on the calculated frequency value through an iterative method based on DFT interpolation to obtain a high-precision frequency calculation value, and determine the high-precision distance calculation value of the core position point based on the shortest reflection path according to the high-precision frequency calculation value.

[0119] S32: Determine the spatial frequency corresponding to the maximum amplitude value in the spatial spectrum and determine the high-precision angle calculation value of the core position point based on the spatial frequency corresponding to the maximum amplitude value.

[0120] Wherein, step S31 includes:

[0121] S311: Calculate the DFT of single points located on both sides of the peak spectral line, and the expressions of the discrete time series of the two single points are X IF (m + 1 / 2) and X IF (m - 1 / 2);

[0122] S312: Substitute the expressions X IF (m + 1 / 2) and X IF (m - 1 / 2) into the following formula: Obtain the calculation formula of the frequency calculation error, and the formula is as follows:

[0123]

[0124] In the formula, represents the frequency calculation error. Usually, is caused by the fence effect.

[0125] And, the calculation formula of the high-precision frequency calculation value is as follows:

[0126]

[0127] In the formula, f fine represents the high-precision frequency calculation value;

[0128] S313: Determine the number of iterations α, assign the iteration control variable i to 1, and assign to 0, and calculate according to the following formula and

[0129]

[0130] S314: Calculate the frequency calculation error according to the following formula:

[0131]

[0132] S315: Judge the magnitude relationship between the number of iterations α and the control variable i. If i ≤ α, increase the iteration control variable from i to i + 1, and iteratively execute the step S314;

[0133] S316: If i > α, calculate the high-precision frequency calculation value through the following formula:

[0134]

[0135] S317: Determine the high-precision distance calculation value of the core position point based on the shortest reflection path according to the high-precision frequency calculation value and the following formula;

[0136]

[0137] In the formula, c represents the speed of light; T c represents the duration of the chirp; B represents the bandwidth of the chirp.

[0138] Among them, step S32 includes:

[0139] S321: Determine the spatial frequency corresponding to the maximum amplitude value in the spatial spectrum, and determine the high-precision angle calculation value of the core position point based on the shortest reflection path according to the spatial frequency corresponding to the maximum amplitude value and the following formula:

[0140] θ = sin -1 (λf spatial ) ;

[0141] In the formula, f spatial represents the spatial frequency corresponding to the maximum amplitude value in the spatial spectrum; θ represents the high-precision angle calculation value; λ represents the wavelength.

[0142] S4: Continuously and repeatedly execute the steps S1 to S3 during the vehicle driving process, and respectively determine the high-precision distance calculation values and high-precision angle calculation values of several consecutive core position points during the vehicle driving process.

[0143] Specifically, during vehicle driving, the millimeter-wave radar continuously emits initial signals and continuously obtains high-precision distance calculation values and high-precision angle calculation values of several consecutive core position points.

[0144] S5: Establish a speed bump feature calculation model based on the high-precision distance calculation values and high-precision angle calculation values of several consecutive core position points, and calculate the features of the speed bump according to the speed bump feature calculation model.

[0145] Among them, step S5 includes:

[0146] S51: Establish a speed bump feature calculation model based on the high-precision distance calculation values and high-precision angle calculation values of several consecutive core position points.

[0147] Specifically, generally, as Figure 4 shown, Figure 4 is a schematic diagram of the speed bump feature in the embodiment of the present application. The color of the road surface speed bump is yellow and black, the height is 30 mm to 60 mm, and the width is 300 mm to 400 mm. In the embodiment of the present application, the height and width of the speed bump are respectively defined as H and W.

[0148] Since the vehicle is continuously approaching the speed bump during driving, the height and width of the part of the speed bump in contact are continuously increasing, and the vertical vibration acceleration of the vehicle when passing through the speed bump will also gradually become larger, that is, the bump of the vehicle caused by the speed bump will gradually intensify. Therefore, the height and width of the speed bump can be regarded as the effective features of the speed bump to characterize different speed bumps. To make the semi-active suspension control strategy make corresponding adjustments to the suspension system according to these two features to adapt to the impact caused by speed bumps of different heights and widths on the vehicle, it is necessary to obtain the height and width of the speed bump before the vehicle reaches the speed bump.

[0149] The speed bump is very small in size relative to the vehicle. During the driving of the vehicle, the millimeter-wave radar stays above the speed bump for a very short time. Therefore, it can be considered that the moving speed of the millimeter-wave radar remains constant at v during this process. When the millimeter-wave radar emits chirps at the same time interval T, the calculation formula for the trajectory interval between two adjacent measurements is as follows:

[0150] l = vT;

[0151] Combined with the high-precision distance calculation values and high-precision angle calculation values corresponding to the shortest reflection paths obtained by the millimeter-wave radar at different positions on its movement trajectory, the embodiment of the present application can establish a speed bump feature calculation model suitable for it, as Figure 5 shown, Figure 5 is a schematic diagram of the speed bump feature calculation model in the embodiment of the present application.

[0152] S52: Determine the high-precision distance trajectory and the high-precision angle trajectory respectively according to the speed bump feature calculation model.

[0153] When the millimeter-wave radar is at each position on the trajectory, the high-precision angle calculation value θ is obtained through the high-precision angle calculation method based on MIMO. When the millimeter-wave radar is above the ground, θ is close to 0 degrees. At this time, the high-precision distance calculation value d obtained through the high-precision ranging algorithm based on DFT is the ground distance. Since the mechanical components of the vehicle will generate a certain degree of vibration during operation, this vibration will cause additional errors in d. The average value of several groups of ground distances obtained can be taken to weaken the additional errors caused by vehicle body vibration. At the same time, it is also necessary to firmly install the millimeter-wave radar on the vehicle body to limit the unnecessary additional mechanical vibration caused by loose installation.

[0154] When the millimeter-wave radar encounters an obstacle, d or θ will show a sudden change. Mark the high-precision distance calculation value at this time as d 0 , and mark the high-precision angle calculation value as θ 0 . During the process of the millimeter-wave radar continuing to move forward, mark the obtained high-precision distance calculation value and high-precision angle calculation value as θ kT and d kT , where k = 1, 2, 3, …. Mark the d kT and θ kT obtained by the millimeter-wave radar on its trajectory to form a high-precision distance sequence and a high-precision angle sequence, where k = 0, 1, 2, 3, ….

[0155] The surface of the speed bump has specific textures, and the textures will cause multipath interference to the reflected signals. The high-precision distance-angle calculation values obtained at different positions on the surface of the speed bump will contain some outliers. In the embodiments of the present application, the Hampel filter is first used to detect and remove these outliers. Secondly, the embodiments of the present application need to determine whether the obstacle is a speed bump, and calculate the similarity between the currently obtained high-precision distance trajectory and high-precision angle trajectory and the high-precision distance trajectory and high-precision angle trajectory corresponding to the speed bump respectively.

[0156] S53: According to the high-precision distance trajectory and the high-precision angle trajectory, and calculate the features of the speed bump according to the following formula:

[0157] H = d - d hT ;

[0158] W = 2hl;

[0159] In the formula, H represents the height of the speed bump; W represents the width of the speed bump; d represents the high-precision distance calculation value between the millimeter-wave radar and the road surface when the millimeter-wave radar has not passed through the speed bump; d hT represents the high-precision distance calculation value when the millimeter-wave radar reaches the vertex of the speed bump.

[0160] Specifically, to meet the real-time requirements of the speed bump feature perception system, the embodiments of the present application calculate the similarity by using the Euclidean distance with extremely low computational complexity. Calculating the Euclidean distance (referring to the actual distance between two points in an m-dimensional space) requires keeping the same number of points for two trajectories. Therefore, the embodiments of the present application select the same number of trajectory points to calculate the Euclidean distances of two high-precision distance trajectories and high-precision angle trajectories respectively. When the Euclidean distances of the two curves are both less than the threshold, it is considered that the target passed by the millimeter-wave radar is a speed bump. Finally, the height and width of the speed bump are calculated. When the obstacle encountered by the millimeter-wave radar is a speed bump, during the forward movement of the millimeter-wave radar, θ kT gradually increases from the minimum value at the mutation to 0 degrees, and d kT gradually decreases. When the millimeter-wave radar reaches the vertex of the speed bump, d kT takes the minimum value d hT .

[0161] Therefore, the calculation formula for the height H of the speed bump is as follows:

[0162] H = d - d hT ;

[0163] The calculation formula for the width W of the speed bump is as follows:

[0164] W = 2hl;

[0165] Finally, the height and width of the speed bump are used as the output of feature calculation. (2)

[0167] The second technical solution of the present application is a speed bump feature calculation system based on a millimeter-wave radar, as Figure 6 shown, Figure 6 which is a schematic structural diagram of a speed bump feature calculation system based on a millimeter-wave radar in the embodiments of the present application. The calculation system includes a radar module and a calculation module.

[0168] The radar module 1 is used to transmit an initial signal and receive an intermediate frequency signal at a number of consecutive core position points;

[0169] The calculation module 2 is used to determine the frequency calculation value of the peak spectral line and the spatial frequency spectrum in the intermediate frequency signals of a number of consecutive core position points respectively according to the intermediate frequency signals of a number of consecutive core position points;

[0170] It is also used to determine the high-precision distance calculation values of a number of consecutive core position points respectively according to the frequency calculation values of a number of consecutive core position points and determine the high-precision angle calculation values of a number of consecutive core position points respectively according to the spatial frequency spectra of a number of consecutive core position points;

[0171] It is also used to establish a speed bump feature calculation model based on the high-precision distance calculation values and high-precision angle calculation values of several consecutive core position points, and calculate the features of the speed bump according to the speed bump feature calculation model.

[0172] Specifically, the radar module 1 is a millimeter-wave radar installed on the front chassis of the vehicle.

[0173] In the static scenario, the average absolute error of the calculated speed bump height by the calculation system of the embodiment of the present application is 0.5 cm, and the average absolute error of the speed bump width is 2.4 cm. In the dynamic experimental scenario, the average absolute error of the calculated speed bump height by the calculation system of the embodiment of the present application is 0.5 cm, and the average absolute error of the speed bump width is 4.1 cm. Considering the influence of vehicle speed error, the experimental results demonstrate the effectiveness and reliability of the system. The system has good real-time performance, and the highest vehicle speed at which the speed bump feature calculation can be effectively realized is 25.7 km / h. (Three)

[0175] Embodiment 1

[0176] (1) Implementation equipment: The speed bump feature calculation system is implemented on AWR1642 BOOST, and a data acquisition board DCA1000EVM and a laptop computer are equipped for this millimeter-wave radar evaluation board.

[0177] (2) Implementation parameters: The starting frequency of the millimeter-wave radar is set to 77 GHz;

[0178] The frequency modulation rate of the millimeter-wave radar is set to 36.017 MHz / us;

[0179] The period for the millimeter-wave radar to transmit one chirp is 150 us, the bandwidth is 3961.87 MHz, its transmission time is 110 us, and the idle time is 40 us; the starting sampling time is set to 7 us.

[0180] The sampling rate of the receiving channel of the millimeter-wave radar is 2560 ksps, and each received chirp contains 256 sampling data.

[0181] The millimeter-wave radar is set to the TDM-MIMO signal transmission form of 2 transmit and 4 receive. Each frame only contains a pair of chirp groups that implement TDM-MIMO.

[0182] (3) Implementation conditions: The embodiment of the present application evaluates the speed bump feature calculation system in static experimental scenarios and dynamic experimental scenarios respectively. The dynamic experimental scenario is a real vehicle driving scenario, and the static experimental scenario replaces the movement of the vehicle by moving the speed bump.

[0183] Four speed bumps of different models are required during the evaluation process, and their differences are reflected in height, width, and surface texture. The 1st, 2nd, and 3rd speed bumps have different heights and widths, among which the 1st speed bump has the largest height and width, and the 3rd speed bump has the smallest height and width. The surface textures of the three speed bumps are all wavy. The 3rd and 4th speed bumps have the same height and width, but the surface texture of the 4th speed bump is diagonal.

[0184] (4) Inspection of implementation results:

[0185] 1) System effectiveness:

[0186] The embodiment of this application is compared with the basic in-vehicle millimeter-wave radar intermediate frequency signal processing method, that is, the results obtained without using the high-precision algorithm are compared. When not using the high-precision algorithm, the distance measurement values at any position of the trajectory are integer multiples of the distance resolution. Even if the appearance of the speed bump causes changes in the distance measurement values, the finally obtained height of the speed bump is also an integer multiple of the distance resolution, which cannot reflect any effective information of the speed bump. This also confirms the necessity of realizing high-precision distance-angle measurement through the high-precision algorithm for the calculation of speed bump characteristics.

[0187] At the same time, the results obtained by using the laser rangefinder are also compared with the results of speed bump characteristic calculation. The 1st speed bump is used in the experiment, as Figure 7 shown, Figure 7 is a schematic diagram of the error comparison between the speed bump characteristic calculation system and the laser rangefinder for calculating the speed bump characteristics in the embodiment of this application. In each group of comparison diagrams, the left side represents the height and the right side represents the width. Figure 7 It shows the measurement errors when using the speed bump characteristic calculation method and the laser rangefinder. The measurement error results are all calculated from 8 groups of data. The average absolute error of the speed bump height obtained through the speed bump characteristic calculation system is only 0.2 cm, and the average absolute error of the width is 1.5 cm. Therefore, it can be considered that the speed bump characteristic calculation method proposed in the embodiment of this application has good performance.

[0188] Although there is a slight gap compared with the method of the laser rangefinder, the laser rangefinder, as an optical-based sensor, cannot work properly when blocked. And when the vehicle is driving, the chassis will always be contaminated with a certain degree of road debris, including mud and snow. In this case, the advantages of the millimeter-wave radar are reflected.

[0189] 2) Speed bump feature recognition:

[0190] This group of experiments evaluates the ability of the speed bump feature calculation system to determine whether the obstacle encountered by the vehicle is a speed bump. Data of speed bumps and non-speed bump objects are collected at a measurement interval of 0.5 cm (the influence of different measurement intervals will be evaluated later). The speed bump used in the experiment is Speed Bump No. 1, and the non-speed bump object is a cardboard box with a height of 10 cm. Since the speed bump feature calculation system will not further calculate the features of non-speed bump objects when it detects that the target object is not a speed bump, this group of experiments shows the high-precision distance map and high-precision angle map corresponding to different positions on the trajectory when the millimeter-wave radar passes above the speed bump and the cardboard box, and evaluates the effectiveness of this system in identifying speed bumps through the variation characteristics of high-precision distance and high-precision angle. Since the texture of the object surface will cause certain fluctuations in the high-precision distance and high-precision angle obtained at adjacent positions, the high-precision distance map and high-precision angle map after sliding mean filtering are shown here. Through sliding mean filtering, the data can be made smoother, making it easier to observe the variation characteristics of high-precision distance and high-precision angle.

[0191] As Figure 8 and Figure 9 shown, Figure 8 is the high-precision distance comparison chart of the speed bump and the cardboard box in the embodiment of this application, Figure 9 is the high-precision angle comparison chart of the speed bump and the cardboard box in the embodiment of this application. Whether the target is a speed bump can be judged by the shapes of the high-precision distance curve and high-precision angle curve of the speed bump and the cardboard box. When the millimeter-wave radar passes above the speed bump, the distance from the speed bump shows a gradually decreasing trend, and after reaching the apex of the speed bump, the distance gradually increases. The change trend of the angle is that when the millimeter-wave radar is above the ground, the angle is about 0 degrees, and when it encounters the speed bump, the angle suddenly changes. After the angle reaches the minimum value, it gradually increases. When the millimeter-wave radar reaches the apex of the speed bump, the angle returns to about 0 degrees again. When the millimeter-wave radar passes over the surface of the cardboard box, the change trends of the distance and angle show completely different characteristics. The high-precision distance curve shows a sudden change at the edge of the cardboard box and then remains stable. At this time, the distance value is the distance from the millimeter-wave radar to the surface of the cardboard box. Except for a slight change at the boundary, the angle is generally about 0 degrees.

[0192] 3) Real-time performance and measurement interval:

[0193] In Figure 1In the scenario shown, when the vehicle is at the dotted line position, assuming the distance between the tire and the speed bump is 30 cm, if the vehicle speed is 36 km / h, the wheel will encounter the speed bump in only 30 ms. Considering that sufficient time needs to be left for the semi-active suspension to make corresponding adjustments, the system should have an extremely fast response time and be able to achieve quantitative prediction of the speed bump in an extremely short time. Real-time analysis of the speed bump feature calculation system is carried out on MATLAB. A total of 8 experiments are conducted to obtain 5000 frames of data to calculate the processing delay of the high-precision distance calculation value and the high-precision angle calculation value. The average value of the processing delay of the 8 experiments is 3.479 s, that is, it can be obtained that the speed bump feature calculation system needs about 0.7 ms to obtain the high-precision distance calculation value and the high-precision angle calculation value from 1 frame of data. After obtaining the last frame of data above the vertex of the speed bump, the speed bump feature calculation system can calculate the height and width of the speed bump through feature calculation by combining several previous groups of data. The time for feature calculation can be ignored. If the high-precision distance-angle calculation can be completed for each frame of data before obtaining the next frame of data, the time for the vehicle to pass through a measurement interval must be greater than the processing delay of 1 frame of data, which is 0.7 ms. At this time, the system has good real-time performance.

[0194] Under different measurement intervals, evaluate the accuracy of calculating the height and width of the speed bump through the speed bump feature calculation system. This part of the experiment is carried out in a static scenario to finely control the measurement interval when the millimeter-wave radar passes above the speed bump. The measurement intervals are respectively set to 0.5 cm, 1 cm, 1.5 cm, and 2 cm to measure the No. 1 speed bump.

[0195] As Figure 10 shown, Figure 10 is the error schematic diagram of calculating the speed bump features under different measurement intervals in the embodiment of the present application. In each group of comparison diagrams, the left side represents the height and the right side represents the width. As the measurement interval increases, the errors of the height and width also show an increasing trend. Therefore, in order to obtain higher accuracy, the embodiment of the present application selects a measurement interval of 0.5 cm to implement the speed bump feature calculation system. At a measurement interval of 0.5 cm, to make the time for the vehicle to pass through a measurement interval greater than 0.7 ms, the vehicle speed needs to be less than 25.7 km / h.

[0196] 4) Influence of different speed bumps

[0197] In this part, experiments are carried out on four different types of speed bumps to evaluate the performance of the speed bump feature calculation system under different types of speed bumps. The experimental scenario is a static experimental scenario. As Figure 11 shown, Figure 11This is a schematic diagram of the error in calculating the characteristics of speed bumps of different models in the embodiments of this application. In each set of comparison diagrams, the left side represents the height and the right side represents the width. For speed bumps with the same texture, as the size decreases, the errors in their height and width gradually increase. However, the average error in height remains within 0.7 cm, and the error in width remains within 3.1 cm. The reduction in the size of the speed bump has a more significant impact on the error in the calculated width value than on the error in the calculated height value. This is because the reduction in the size of the speed bump makes the slope of the speed bump gentler, thereby causing a larger error in determining the starting point of the speed bump. However, this has less impact on determining the highest point of the speed bump. The final calculation results of speed bumps with the same size but different surface textures also show certain differences. This is because the texture on the surface of the speed bump causes multipath interference to the reflected signal. Compared with the No. 3 speed bump, the dense raised points on the surface of the No. 4 speed bump deepen the impact of multipath interference.

[0198] 5) Influence of different vehicle speeds

[0199] In the embodiments of this application, experiments are carried out at different vehicle speeds to verify the influence of vehicle speed on the calculation accuracy of the height and width of speed bumps. The vehicle speeds are respectively set to 0 km / h, 10 km / h, 20 km / h, and 30 km / h. Among them, 0 km / h means the experiment is carried out in a static experimental scenario, and the latter three groups of experiments are carried out in a dynamic experimental scenario. The speed bump used is the No. 1 speed bump. Since the embodiments of this application process the data after collecting the complete data, there is no need to consider the maximum vehicle speed at which the system can work effectively. In order to keep the measurement intervals consistent for the three vehicle speeds in the dynamic experimental scenario, the frame intervals are adjusted to 1.8 ms, 0.9 ms, and 0.6 ms respectively. At this time, the measurement intervals are all 0.5 cm.

[0200] As Figure 12 shown, Figure 12 This is a schematic diagram of the error in calculating the characteristics of speed bumps at different vehicle speeds in the embodiments of this application. In each set of comparison diagrams, the left side represents the height and the right side represents the width. The error in the dynamic experimental scenario is greater than that in the static experimental scenario. The reason for the increase in the error of the speed bump height is that the vibration of the vehicle itself in the dynamic experimental scenario introduces additional variables to the calculation of the speed bump height, making the height calculation unable to be as stable as in the static experimental scenario. However, the average error in height is still within 0.6 cm. Therefore, it can be considered that the speed bump characteristic calculation system has good speed bump height calculation ability in the real scenario. In the embodiments of this application, the calculated value of the vehicle speed is obtained through the vehicle's dashboard. The error in the calculated value of the vehicle speed further causes an error in the calculation of the speed bump width. Therefore, if a high-precision vehicle speed can be obtained, the speed bump characteristic calculation system can also have good speed bump width calculation ability.

[0201] The above has described the embodiments of the present application in detail, but the content is only the preferred embodiments of the present application and cannot be considered as limiting the scope of implementation of the present application. All equivalent changes and improvements made within the scope of the present application shall still fall within the scope covered by the patent of the present application.

Claims

1. A deceleration strip feature calculation method based on a millimeter-wave radar, characterized in that, the millimeter-wave radar is installed on the front chassis of the vehicle and emits an initial signal to the road surface, and the method includes: S1: Receive the intermediate-frequency signal obtained after the initial signal is reflected by several position points on the road surface during the driving of the vehicle; S2: Analyze the spectrum of the intermediate-frequency signal, determine the shortest reflection path corresponding to the peak spectral line in the spectrum, and determine the frequency calculation value and spatial spectrum of the peak spectral line in the spectrum; S3: Determine the high-precision distance calculation value of the core position point based on the shortest reflection path according to the frequency calculation value, and determine the high-precision angle calculation value of the core position point according to the spatial spectrum; S4: Continuously repeat the execution of steps S1 to S3 during the driving of the vehicle, and respectively determine the high-precision distance calculation value and high-precision angle calculation value of several consecutive core position points during the driving of the vehicle; S5: Establish a deceleration strip feature calculation model according to the high-precision distance calculation value and high-precision angle calculation value of several consecutive core position points, and calculate the features of the deceleration strip according to the deceleration strip feature calculation model; The step S3 includes: S31: Perform error elimination processing on the frequency calculation value through an iterative method based on DFT interpolation to obtain a high-precision frequency calculation value, and determine the high-precision distance calculation value of the core position point based on the shortest reflection path according to the high-precision frequency calculation value; S32: Determine the spatial frequency corresponding to the maximum amplitude value in the spatial spectrum and determine the high-precision angle calculation value of the core position point based on the shortest reflection path according to the spatial frequency corresponding to the maximum amplitude value; The step S31 includes: S311: Calculate the DFT of single points respectively located on both sides of the peak spectral line. The expressions of the discrete-time sequences of the two single points are X IF (m + 1 / 2) and X IF (m - 1 / 2); S312: Expressions X of the discrete time series of the two single points IF (m + 1 / 2) and X IF (m - 1 / 2) are substituted into the following formula: The calculation formula for the frequency calculation error is obtained, and the formula is as follows: In the formula, represents the frequency calculation error; and, the calculation formula of the high-precision frequency calculation value is as follows: where f fine represents the calculated value of high-precision frequency; S313: Determine the number of iterations α, assign the iteration control variable i to 1, and assign to 0, and calculate according to the following formula and S314: Calculate the frequency calculation error according to the following formula: S315: Judge the size of the iteration times α and the control variable i. If i ≤ α, the iteration control variable increases from i to i + 1, and iteratively execute the step S314; S316: If i > α, calculate the high-precision frequency calculation value through the following formula: S317: Determine the high-precision distance calculation value of the core position point based on the shortest reflection path according to the high-precision frequency calculation value and the following formula; where c represents the speed of light; T c represents the duration of the chirp; B represents the bandwidth of the chirp.

2. The deceleration strip feature calculation method based on a millimeter-wave radar according to claim 1, characterized in that, the antenna array of the millimeter-wave radar is set in the MIMO form of two transmitters and four receivers and can virtualize eight receiving antennas; and, the step S2 includes: S21: Perform sampling processing on the intermediate-frequency signal to obtain a discrete time series. The expression of the discrete time series is as follows: where x IF (n) represents a discrete-time sequence; N represents the number of sampling points; n represents the sampling sequence number; A represents the amplitude of the intermediate-frequency signal; f s represents the sampling rate; f IF represents the frequency of the intermediate-frequency signal, represents the initial phase of the intermediate-frequency signal; S22: Perform spectrum analysis on the discrete time series through FFT to obtain an amplitude spectrum, and determine the amplitude value, spectral line number and shortest reflection path corresponding to the peak spectral line in the amplitude spectrum, and determine the frequency calculation value of the peak spectral line according to the spectral line number. The expression of the frequency calculation value is as follows: where f coarse represents the calculated frequency value; m represents the spectral line number of the peak spectral line; S23: Zero-fill the amplitude value of the peak spectral line to obtain a 256-point discrete sequence, and determine the spatial spectrum of the peak spectral line according to the 256-point discrete sequence. The expression of the spatial signal frequency resolution is as follows: In the formula, Q represents the number of points of the angular FFT; l represents the distance between receiving antennas.

3. A deceleration strip feature calculation method based on a millimeter-wave radar according to claim 1, characterized in that the step S32 includes: S321: Determine the spatial frequency corresponding to the maximum amplitude value in the spatial spectrum, and determine the high-precision angle calculation value of the core position point based on the shortest reflection path according to the spatial frequency corresponding to the maximum amplitude value and the following formula: θ = sin -1 (λf spatial ); where f spatial represents the spatial frequency corresponding to the maximum amplitude value in the spatial frequency spectrum; θ represents the high-precision angle calculation value; λ represents the wavelength.

4. A deceleration strip feature calculation method based on a millimeter-wave radar according to claim 3, characterized in that the features of the deceleration strip include: the height and width of the deceleration strip; and, the step S5 includes: S51: Establish a deceleration strip feature calculation model according to the high-precision distance calculation values and high-precision angle calculation values of several consecutive core position points; S52: Respectively determine the high-precision distance trajectory and high-precision angle trajectory according to the deceleration strip feature calculation model; S53: Calculate the features of the deceleration strip according to the high-precision distance trajectory and high-precision angle trajectory, and according to the following formula: H = d - d hT ; W = 2hl; Wherein, H represents the height of the speed bump; W represents the width of the speed bump; d represents the calculated value of the high-precision distance between the millimeter-wave radar and the road surface when the millimeter-wave radar has not passed over the speed bump; d hT represents the calculated value of the high-precision distance when the millimeter-wave radar reaches the apex of the speed bump.

5. A calculation system for implementing the deceleration strip feature calculation method based on a millimeter-wave radar according to claim 1, characterized in that it includes: a radar module, configured to transmit an initial signal and receive an intermediate-frequency signal at several consecutive core position points; a calculation module, configured to respectively determine the frequency calculation value and spatial spectrum of the peak spectral line in the intermediate-frequency signals of several consecutive core position points according to the intermediate-frequency signals of several consecutive core position points; is further configured to respectively determine the high-precision distance calculation values of several consecutive core position points according to the frequency calculation values of several consecutive core position points, and respectively determine the high-precision angle calculation values of several consecutive core position points according to the spatial spectra of several consecutive core position points; and is further configured to establish a deceleration strip feature calculation model according to the high-precision distance calculation values and high-precision angle calculation values of several consecutive core position points, and calculate the features of the deceleration strip according to the deceleration strip feature calculation model.

Images