A self-driving vehicle speed estimation method, system and platform suitable for multiple scenarios

By calculating the longitudinal speed and static point velocity distribution of bicycles, combined with histogram statistics and linear filtering processing, the accuracy problem of bicycle speed estimation in the existing technology is solved in scenarios with fewer acceleration and deceleration and static targets, and achieved the Accurate bicycle speed estimation in multiple scenarios.

CN117783572BActive Publication Date: 2024-11-26SUZHOU CHENGTAI TECH CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202311755974.6
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2023-12-20
Publication Date
2024-11-26
Estimated Expiration
2043-12-20

AI Technical Summary

Technical Problem

The existing bicycle speed estimation methods have errors in scenarios such as acute acceleration and deceleration, which affects the determination of target motion state, and cannot obtain effective static target clusters in scenarios with fewer static targets, resulting in inaccurate calculations.

Method used

By obtaining the distance, speed and angle data of the detection point, the longitudinal speed of the bicycle is calculated, and combining the speed, deflection angle, acceleration and wheel speed data of the bicycle, the speed distribution of the stationary point is generated, and the speed of the bicycle is calculated in real time, and statistics are used for histograms. and linear filtering to improve estimation accuracy.

Benefits of technology

It achieves a more accurate vehicle speed when the bicycle is accelerated by a large bicycle. It is suitable for multiple scenarios and improves the accuracy and real-timeness of vehicle speed estimation.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN117783572B_ABST
    Figure CN117783572B_ABST
Patent Text Reader

Abstract

The present invention discloses a self-vehicle speed estimation method, system and platform applicable to multiple scenarios. The method obtains first data of a detection point corresponding to the self-vehicle in the scene, and generates the self-vehicle longitudinal speed corresponding to the first data according to the first data; wherein the first data includes detection point distance data, detection point speed data and detection point angle data; according to the self-vehicle longitudinal speed, in combination with second data corresponding to the self-vehicle body, generates the static point speed distribution corresponding to the self-vehicle body; wherein the second data includes self-vehicle speed data, self-vehicle deflection angle data, self-vehicle longitudinal acceleration data and self-vehicle wheel speed data; according to the static point speed distribution, the self-vehicle speed data corresponding to the self-vehicle body is generated in real time, as well as the system and platform corresponding to the method, so as to achieve the effect of obtaining the accurate speed of the self-vehicle in real time, that is, obtaining a more accurate vehicle speed when the acceleration of the self-vehicle is large (starting, emergency braking, etc.).
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention belongs to the technical field of vehicle speed estimation and processing, and in particular relates to a vehicle speed estimation method, system and platform applicable to multiple scenarios. Background Art

[0002] Millimeter-wave radar is one of the key sensors in ADAS. Its basic functions include distance measurement, speed measurement, and angle measurement. Due to the speed measurement principle of millimeter-wave radar, the measured speed is a relative speed. In automotive-level application scenarios, it is necessary to judge the motion properties of the target based on the speed measurement information and the vehicle speed information.

[0003] The existing vehicle speed information is obtained through vehicle body information, which has a certain delay. The delay has little impact during the normal driving of the vehicle, but in some scenarios such as sudden acceleration and deceleration, the obtained vehicle speed information is significantly different from the actual vehicle speed, which will lead to errors in the judgment of the target motion state, and thus affect the subsequent fusion or function realization.

[0004] In addition, patent CN111308458A discloses a method for estimating the speed of a self-vehicle based on a vehicle-mounted millimeter-wave radar. The method obtains the distance, speed and angle of all targets, and calculates the relative speed between the target and the self-vehicle in the normal direction; the speeds of all stationary targets are counted, and confidence matching is performed to obtain the target cluster results with the top three target numbers; the speed of the self-vehicle is estimated based on the speed estimation confidence in the target cluster. This scheme discloses speed statistics based on stationary targets and uses smoothing filtering to reduce speed estimation errors, but it does not consider the body posture of the target vehicle, and the calculated relative speed in the normal direction is inaccurate. In addition, it is impossible to obtain an effective stationary target cluster in scenes with fewer stationary targets, such as congested slow traffic or open areas.

[0005] Therefore, in view of the above technical problems and defects, it is urgent to design and develop a self-vehicle speed estimation method, system and platform suitable for multiple scenarios. Summary of the invention

[0006] In order to overcome the shortcomings and difficulties of the above-mentioned prior art, the present invention provides a vehicle speed estimation method, system and platform suitable for multiple scenarios, so as to achieve the effect of obtaining the accurate speed of the vehicle in real time, that is, a more accurate vehicle speed can be obtained when the vehicle has a large acceleration (starting, sudden braking, etc.).

[0007] The first purpose of the present invention is to provide a self-vehicle speed estimation method applicable to multiple scenarios; the second purpose of the present invention is to provide a self-vehicle speed estimation system applicable to multiple scenarios; the third purpose of the present invention is to provide a self-vehicle speed estimation platform applicable to multiple scenarios.

[0008] The first object of the present invention is achieved by:

[0009] Acquire first data of a detection point corresponding to the vehicle in the scene, and generate a longitudinal speed of the vehicle corresponding to the first data according to the first data; wherein the first data includes detection point distance data, detection point speed data, and detection point angle data;

[0010] According to the longitudinal speed of the vehicle, combined with the second data corresponding to the vehicle body, a static point speed distribution corresponding to the vehicle body is generated; wherein the second data includes the vehicle speed data, the vehicle deflection angle data, the longitudinal acceleration data of the vehicle and the wheel speed data of the vehicle;

[0011] According to the static point speed distribution, the vehicle speed data corresponding to the vehicle body is generated in real time.

[0012] Furthermore, the calculation formula of the longitudinal speed of the vehicle is as follows:

[0013] V otg =V / cosθ (1)

[0014] Among them, V otg Z is the longitudinal velocity; V is the radial velocity of the detection point, and θ is the azimuth angle of the detection point.

[0015] Furthermore, the generating of the static point velocity distribution corresponding to the vehicle body according to the longitudinal velocity of the vehicle body in combination with the second data corresponding to the vehicle body also includes:

[0016] Creating a first threshold value corresponding to the dynamic and static attributes of the detection point; wherein the first threshold value is a threshold for determining the dynamic and static attributes of the detection point;

[0017] It is determined that first detection point data corresponding to the longitudinal speed of the vehicle and smaller than the first threshold value is generated.

[0018] Furthermore, the calculation formula of the first threshold is as follows:

[0019] Threshold1=V err +cosθ·lon+sinθ·lat (2)

[0020] Among them, V err is the minimum baseline error, lon is the longitudinal velocity correction, and lat is the lateral velocity correction.

[0021] Furthermore, the generating of the static point velocity distribution corresponding to the vehicle body according to the longitudinal velocity of the vehicle body in combination with the second data corresponding to the vehicle body also includes:

[0022] According to the vehicle scene, a second threshold corresponding to the dynamic and static attributes of the detection point is created; wherein the second threshold is a threshold of the number of valid points that meets the first threshold;

[0023] It is determined to generate second detection point data corresponding to the longitudinal speed of the vehicle and smaller than the second threshold value.

[0024] Furthermore, the generating of the static point velocity distribution corresponding to the vehicle body according to the longitudinal velocity of the vehicle body in combination with the second data corresponding to the vehicle body also includes:

[0025] In combination with histogram statistics, histogram parameters corresponding to the histogram are created; wherein the histogram parameters include range, group interval and number of groups;

[0026] According to the histogram parameters, group data corresponding to the detection points are generated; wherein the group data includes the group number of each detection point;

[0027] According to the group data, the number of detection points in each group is compared in turn, and group number data corresponding to the number of detection points is generated; wherein the group number data is the group number to which the detection point array is located.

[0028] Furthermore, the real-time generation of the vehicle speed data corresponding to the vehicle body according to the static point speed distribution also includes:

[0029] Combined with linear filtering, the deflection angle data corresponding to the vehicle body is smoothed;

[0030] A third threshold value corresponding to the longitudinal acceleration of the vehicle is created, and whether the vehicle speed data is abnormal is determined according to the third threshold value; wherein the third threshold value is 30% of the maximum longitudinal acceleration of the vehicle.

[0031] The second object of the present invention is achieved by: the system is applied to the vehicle speed estimation method; the system comprises:

[0032] An acquisition generating unit is used to acquire first data of a detection point corresponding to the vehicle in the scene, and generate a longitudinal speed of the vehicle corresponding to the first data according to the first data; wherein the first data includes detection point distance data, detection point speed data and detection point angle data;

[0033] A first generating unit is used to generate a static point velocity distribution corresponding to the vehicle body according to the longitudinal velocity of the vehicle and second data corresponding to the vehicle body; wherein the second data includes vehicle speed data, vehicle deflection angle data, vehicle longitudinal acceleration data and vehicle wheel speed data;

[0034] The second generating unit is used to generate the vehicle speed data corresponding to the vehicle body in real time according to the static point speed distribution.

[0035] Furthermore, the calculation formula of the longitudinal speed of the vehicle is as follows:

[0036] V otg =V / cosθ (1)

[0037] Among them, V otg Z is the longitudinal velocity; V is the radial velocity of the detection point, and θ is the azimuth angle of the detection point;

[0038] The first generating unit further includes:

[0039] A first creation module is used to create a first threshold value corresponding to the dynamic and static attributes of the detection point; wherein the first threshold value is used to determine the dynamic and static attributes of the detection point;

[0040] A first determination module, used for determining and generating first detection point data corresponding to the longitudinal speed of the vehicle and less than the first threshold value;

[0041] The calculation formula of the first threshold is as follows:

[0042] Threshold1=V err +cosθ·lon+sinθ·lat (2)

[0043] Among them, V err is the minimum baseline error, lon is the longitudinal velocity correction, and lat is the lateral velocity correction;

[0044] And / or, the first generating unit further includes:

[0045] A second creation module is used to create a second threshold value corresponding to the dynamic and static attributes of the detection point according to the vehicle scene; wherein the second threshold value is a threshold value of the number of valid points that meets the first threshold value;

[0046] A second determination module is used to determine and generate second detection point data corresponding to the longitudinal speed of the vehicle and less than the second threshold value;

[0047] And / or, the first generating unit further includes:

[0048] A third creation module is used to create histogram parameters corresponding to the histogram in combination with histogram statistics; wherein the histogram parameters include range, group interval and number of groups;

[0049] A first generating module, used to generate group data corresponding to the detection points according to the histogram parameters; wherein the group data includes the group number of each detection point;

[0050] The second generating module is used to compare the number of detection points in each group in turn according to the group data, and generate group number data corresponding to the number of detection points; wherein the group number data is the group number of the detection point array;

[0051] And / or, the second generating unit further includes:

[0052] The first processing module is used to combine linear filtering to smooth the deflection angle data corresponding to the vehicle body;

[0053] The third determination module is used to create a third threshold value corresponding to the longitudinal acceleration of the vehicle, and determine whether the vehicle speed data is abnormal according to the third threshold value; wherein the third threshold value is 30% of the maximum longitudinal acceleration of the vehicle.

[0054] The third object of the present invention is achieved as follows: comprising a processor, a memory and a self-vehicle speed estimation platform control program applicable to multiple scenarios; wherein the self-vehicle speed estimation platform control program applicable to multiple scenarios is executed in the processor, the self-vehicle speed estimation platform control program applicable to multiple scenarios is stored in the memory, and the self-vehicle speed estimation platform control program applicable to multiple scenarios implements the self-vehicle speed estimation method applicable to multiple scenarios.

[0055] The present invention obtains first data of a detection point corresponding to the self-vehicle in a scene, and generates the longitudinal speed of the self-vehicle corresponding to the first data according to the first data; wherein the first data includes detection point distance data, detection point speed data and detection point angle data; according to the longitudinal speed of the self-vehicle, in combination with second data corresponding to the self-vehicle body, generates the static point speed distribution corresponding to the self-vehicle body; wherein the second data includes self-vehicle speed data, self-vehicle deflection angle data, self-vehicle longitudinal acceleration data and self-vehicle wheel speed data; according to the static point speed distribution, the self-vehicle speed data corresponding to the self-vehicle body is generated in real time, as well as the system and platform corresponding to the method, so as to achieve the effect of obtaining the accurate speed of the self-vehicle in real time, that is, a more accurate vehicle speed can be obtained when the acceleration of the self-vehicle is large (starting, emergency braking, etc.). BRIEF DESCRIPTION OF THE DRAWINGS

[0056] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0057] Figure 1 A schematic diagram of a flow chart of a vehicle speed estimation method applicable to multiple scenarios of the present invention;

[0058] Figure 2 A schematic diagram of a self-vehicle speed estimation processing flow of a self-vehicle speed estimation method embodiment of the present invention applicable to multiple scenarios;

[0059] Figure 3 A schematic diagram of a measured expressway scene in an embodiment of a vehicle speed estimation method applicable to multiple scenes of the present invention;

[0060] Figure 4 A schematic diagram of the architecture of a vehicle speed estimation system applicable to multiple scenarios of the present invention;

[0061] Figure 5 A schematic diagram of the architecture of a vehicle speed estimation platform applicable to multiple scenarios of the present invention;

[0062] The purpose, features and advantages of the present invention will be further described with reference to the accompanying drawings in conjunction with the embodiments. DETAILED DESCRIPTION

[0063] In order to better understand the purpose, technical solutions and advantages of the present invention, the present invention is further described below in conjunction with the accompanying drawings and specific implementation methods. Those skilled in the art can easily understand other advantages and effects of the present invention from the contents disclosed in this specification.

[0064] The present invention may also be implemented or applied through other different specific examples, and the details in this specification may also be modified and changed in various ways based on different viewpoints and applications without departing from the spirit of the present invention.

[0065] It should be noted that if the embodiments of the present invention involve directional indications (such as up, down, left, right, front, back, etc.), the directional indications are only used to explain the relative position relationship, movement status, etc. between the components under a certain specific posture (as shown in the accompanying drawings). If the specific posture changes, the directional indication will also change accordingly.

[0066] In addition, if there are descriptions involving "first", "second", etc. in the embodiments of the present invention, the descriptions of "first", "second", etc. are only used for descriptive purposes and cannot be understood as indicating or suggesting their relative importance or implicitly indicating the number of the indicated technical features. Therefore, the features defined as "first" and "second" may explicitly or implicitly include at least one of the features. Secondly, the technical solutions between the various embodiments can be combined with each other, but they must be based on the ability of ordinary technicians in the field to implement them. When the combination of technical solutions is contradictory or cannot be implemented, it should be deemed that such a combination of technical solutions does not exist and is not within the scope of protection required by the present invention.

[0067] Preferably, the vehicle speed estimation method applicable to multiple scenarios of the present invention is applied in one or more terminals or servers. The terminal is a device that can automatically perform numerical calculations and / or information processing according to pre-set or stored instructions, and its hardware includes but is not limited to microprocessors, application specific integrated circuits (ASIC), programmable gate arrays (FPGA), digital signal processors (DSP), embedded devices, etc.

[0068] The terminal can be a computing device such as a desktop computer, a notebook, a PDA, a cloud server, etc. The terminal can interact with the client through a keyboard, a mouse, a remote control, a touch pad, or a voice control device.

[0069] The present invention is to realize a vehicle speed estimation method, system and platform applicable to multiple scenarios.

[0070] like Figure 1 , which is a flow chart of a vehicle speed estimation method applicable to multiple scenarios provided by an embodiment of the present invention.

[0071] In this embodiment, the vehicle speed estimation method applicable to multiple scenarios can be applied to a terminal or a fixed terminal with a display function, and the terminal is not limited to a personal computer, a smart phone, a tablet computer, a desktop computer or an all-in-one computer equipped with a camera, etc.

[0072] The vehicle speed estimation method applicable to multiple scenarios can also be applied to a hardware environment consisting of a terminal and a server connected to the terminal via a network. The network includes but is not limited to: a wide area network, a metropolitan area network or a local area network. The vehicle speed estimation method applicable to multiple scenarios of the embodiment of the present invention can be executed by a server, can be executed by a terminal, or can be executed by both a server and a terminal.

[0073] For example, for a terminal that needs to perform self-vehicle speed estimation applicable to multiple scenarios, the self-vehicle speed estimation function applicable to multiple scenarios provided by the method of the present invention can be directly integrated on the terminal, or a client for implementing the method of the present invention can be installed. For another example, the method provided by the present invention can also be run on a server or other device in the form of a software development kit (SDK), and an interface for the self-vehicle speed estimation function applicable to multiple scenarios is provided in the form of SDK. The terminal or other device can implement the self-vehicle speed estimation function applicable to multiple scenarios through the provided interface. The present invention is further described below in conjunction with the accompanying drawings.

[0074] like Figure 1 As shown, the present invention provides a vehicle speed estimation method applicable to multiple scenarios, the method comprising the following steps:

[0075] S1, obtaining first data of a detection point corresponding to the vehicle in the scene, and generating a longitudinal speed of the vehicle corresponding to the first data according to the first data; wherein the first data includes detection point distance data, detection point speed data and detection point angle data;

[0076] S2, generating a static point velocity distribution corresponding to the vehicle body according to the longitudinal velocity of the vehicle and combining the second data corresponding to the vehicle body; wherein the second data includes the vehicle velocity data, the vehicle deflection angle data, the longitudinal acceleration data of the vehicle and the wheel speed data of the vehicle;

[0077] S3. Generate vehicle speed data corresponding to the vehicle body in real time according to the stationary point speed distribution.

[0078] The calculation formula of the longitudinal velocity of the vehicle is as follows:

[0079] V otg =V / cosθ (1)

[0080] Among them, V otg Z is the longitudinal velocity; V is the radial velocity of the detection point, and θ is the azimuth angle of the detection point.

[0081] The step of generating a static point velocity distribution corresponding to the vehicle body according to the longitudinal velocity of the vehicle body in combination with the second data corresponding to the vehicle body further includes:

[0082] S21, creating a first threshold value corresponding to the dynamic and static attributes of the detection point; wherein the first threshold value is used to determine the dynamic and static attributes of the detection point;

[0083] S22: Determine and generate first detection point data corresponding to the longitudinal speed of the vehicle and smaller than the first threshold value.

[0084] The calculation formula of the first threshold is as follows:

[0085] Threshold1=V err +cosθ·lon+sinθ·lat (2)

[0086] Among them, V err is the minimum baseline error, lon is the longitudinal velocity correction, and lat is the lateral velocity correction.

[0087] The step of generating a static point velocity distribution corresponding to the vehicle body according to the longitudinal velocity of the vehicle body in combination with the second data corresponding to the vehicle body further includes:

[0088] S23, according to the vehicle scene, create a second threshold corresponding to the dynamic and static attributes of the detection point; wherein the second threshold is a threshold of the number of valid points that meets the first threshold;

[0089] S24: Determine and generate second detection point data corresponding to the longitudinal speed of the vehicle and smaller than the second threshold value.

[0090] The step of generating a static point velocity distribution corresponding to the vehicle body according to the longitudinal velocity of the vehicle body in combination with the second data corresponding to the vehicle body further includes:

[0091] S25, combining histogram statistics, creating histogram parameters corresponding to the histogram; wherein the histogram parameters include range, group distance and number of groups;

[0092] S26, generating group data corresponding to the detection points according to the histogram parameters; wherein the group data includes the group number of each detection point;

[0093] S27. According to the group data, the number of detection points in each group is compared in turn, and group number data corresponding to the number of detection points is generated; wherein the group number data is the group number to which the detection point array is located.

[0094] The method of generating the vehicle speed data corresponding to the vehicle body in real time according to the stationary point speed distribution also includes:

[0095] S31, combining linear filtering to smooth the deflection angle data corresponding to the vehicle body;

[0096] S32, creating a third threshold value corresponding to the longitudinal acceleration of the vehicle, and determining whether the vehicle speed data is abnormal according to the third threshold value; wherein the third threshold value is 30% of the maximum longitudinal acceleration of the vehicle.

[0097] Specifically, in an embodiment of the present invention, a method for estimating the speed of a vehicle based on millimeter-wave radar is provided. The method is based on histogram statistics and information such as vehicle speed, wheel speed, acceleration, etc. output by the vehicle body sensor, and is combined with actual scenarios to adaptively select speed estimation results calculated by different methods. The method has good adaptability and simple calculation, and can obtain the accurate speed of the vehicle in real time.

[0098] To achieve the above object, the present invention adopts the following technical solution:

[0099] Step 1: Get the RVA information (distance, speed, angle) of all detection points in the current frame, and calculate the longitudinal speed of each detection point:

[0100] V otg =V / cosθ (1)

[0101] Where: V is the radial velocity of the detection point, θ is the azimuth angle of the detection point.

[0102] Step 2: Obtain some vehicle body information from the vehicle body sensor: vehicle speed V veh , yaw angle yawRate, vehicle longitudinal acceleration longAcc, and vehicle wheel speed wheelSpeed[4].

[0103] Step 3: Use a histogram to calculate the velocity distribution of the absolute stationary point:

[0104] First, select valid data, that is, the longitudinal velocity satisfies V otg <Threshold1 detection point, Threshold1 is the threshold used to determine the dynamic and static attributes of the detection point:

[0105] Threshold1=V err +cosθ·lon+sinθ·lat (2)

[0106] Where: V err is the minimum baseline error, lon is the longitudinal velocity correction, and lat is the lateral velocity correction. The above values ​​are set through offline data statistics.

[0107] Secondly, set the histogram parameters: range, group interval and number of groups.

[0108] Set the range D:

[0109] D=V max -V min (3)

[0110] The extreme setting only needs to consider the target speed range, and the default maximum speed V max is 50m / s, minimum speed V min is 0m / s.

[0111] Set the class interval d(d∈(0,D]): The setting of the class interval needs to consider both the speed result accuracy and the amount of calculation. The smaller the class interval value, the more accurate the final speed obtained by the histogram statistics, and the correspondingly larger the amount of calculation. It can be adjusted according to actual usage.

[0112] Set the number of groups: k = D / d;

[0113] Next, determine whether the number of valid detection points meets the set threshold Threshold 2. The threshold is adjusted according to the actual application scenario. Considering that there are fewer static targets in some scenarios, it is recommended to set it to 1 to 10. If the number of valid detection points meets the threshold Threshold 2, continue processing, otherwise go to step 4.

[0114] Again, calculate the group number of each detection point:

[0115] index=V otg / d+(D-1)+0.5 (4)

[0116] Finally: count the number of detection points that fall into each group, compare the number of detection points in each group in turn, and obtain the group number index of the group with the largest number of detection points max And calculate the estimated speed:

[0117] V est1 =-(index max -(D-1))×d (5)

[0118] Step 4: Use linear filtering to smooth the deflection angle yawRate using the following formula:

[0119] yawRate - k =yawRate k ×K+yawRate - k-1 ×(1-K),...K∈[0,1] (6)

[0120] Where: yawRate k is the measurement value obtained from the vehicle body information in the current frame, yawRate - k-1 is the smoothing value of the previous frame, yawRate- k is the smoothing value of the current frame, and K is the linear coefficient (K is adjusted according to the actual usage so that the variance of the deflection angle after smoothing is within a certain range).

[0121] Step 5: Determine the validity of the wheel speed information, that is, eliminate abnormal wheel speed data. If the vehicle speed and wheel speed meet the following conditions, the wheel speed information is considered invalid:

[0122]

[0123] Where wheelSpeed[3] is the speed of the left rear wheel, wheelSpeed[4] is the speed of the right rear wheel, and le-6 is scientific notation, which means 1 times 10 to the negative sixth power, i.e. 0.000001.

[0124] The above conditions mean that if the vehicle is in motion and the left rear / right rear wheel speed is 0, the wheel speed information is considered invalid, and the process goes to step 6; otherwise, the wheel speed information is considered valid, and the process goes to step 7.

[0125] Step 6: Calculate the final estimated speed V comp If the wheel speed information is determined to be invalid in step 5, two situations are considered: 1) the valid detection point in step 3 does not meet the threshold; 2) V est1 and V est2 There is a big difference between them, namely V est1 -V est2 >10d. If any of the above two conditions is met, then V comp =V veh , otherwise V comp =V est1 .

[0126] Step 7: If the wheel speed information is determined to be valid in step 5, proceed to this step:

[0127] First, calculate the vehicle speed V based on the left rear wheel speed and the right rear wheel speed est2 :

[0128] V est2 =(wheelSpeed[3]+wheelSpeed[4]) / 2 (8)

[0129] Secondly: determine whether the longitudinal acceleration longAcc of the vehicle meets the set threshold Threshold3. The threshold needs to be set according to offline data statistics and is generally set to 30% of the maximum longitudinal acceleration of the vehicle. At this time, consider three situations: 1) The effective detection point in step 3 does not meet the threshold; 2) V est1 and V est2 There is a big difference between them, namely V est1 -V est2 >5d; 3) longAcc meets the emergency braking condition (i.e. longAcc is greater than 60% of the maximum longitudinal acceleration). If the longitudinal acceleration meets the threshold and any of the above three conditions is met, then V comp =V est2 , otherwise V comp =V est1 ;

[0130] Finally: If longAcc does not meet Threshold3, consider two situations: 1) Turning scenario, that is, yawRate is large (|yawRate|>ω, ω needs to be adjusted according to the actual application scenario); 2) V est1 and V est2 There is a big difference between them, namely V est1 -V est2 >5d. If both of the above conditions are met, then V comp =V est1 , otherwise V comp =V est2 .

[0131] Combine the following Figure 3 The example of the expressway scenario illustrates the processing flow of the solution of the present invention, and the specific steps are as follows:

[0132] Step 1: Select a frame of data with 160 detection points and calculate the longitudinal velocity of each detection point.

[0133] Step 2: Get vehicle body information:

[0134]

[0135] Step 3: Use a histogram to calculate the velocity distribution of the absolute stationary point:

[0136] First, calculate Threshold1 according to the following formula:

[0137] Threshold1=V err +cosθ·lon+sinθ·lat (2)

[0138] Where V err =0.1, lon=0.95, lat=1.37, and the dynamic and static thresholds corresponding to each detection point can be calculated in combination with the azimuth of each detection point. Statistics show that the number of effective detection points is 146.

[0139] Secondly, set the histogram parameters: range D = V max -V min =50-0=50, group distance d=0.25 and number of groups k=D / d=50 / 0.25=200.

[0140] Again, the threshold Threshold2 is set to 5, and the number of valid detection points meets the threshold Threshold2.

[0141] Finally, calculate the group number of each detection point and count the number of detection points falling into each group, compare the number of detection points in each group in turn, and obtain the group number index of the group with the largest number of detection points max=105, calculate the estimated speed: V est1 =-(index max -(D-1))×d=23.50.

[0142] Step 4: Calculate the yawRate=0.00175 of the current frame according to the smoothing filter formula.

[0143] Step 5: Determine whether the wheel speed information is valid and proceed to step 7.

[0144] Step 6: Skip.

[0145] Step 7: Calculate the final estimated speed V comp .

[0146] First, calculate the vehicle speed based on the left rear wheel speed and the right rear wheel speed (need to be converted to m / s):

[0147] V est2 =(wheelSpeed[3]+wheelSpeed[4]) / 2=(84.59+84.69) / 2 / 3.6=23.51.

[0148] Secondly, the longitudinal acceleration threshold Threshold3 is set to 3m / s 2 , and longAcc does not meet the threshold.

[0149] Finally, there are two cases: 1) Turning scenario, that is, yawRate is large (|yawRate|>ω, ω needs to be adjusted according to the actual application scenario); 2) V est1 and V est2 There is a big difference between them, namely V est1 -V est2 >5d, cannot meet the requirements at the same time, so V comp =V est2 =23.51m / s.

[0150] To achieve the above object, the present invention also provides a vehicle speed estimation system applicable to multiple scenarios, such as Figure 4 As shown, the system is applied to the vehicle speed estimation method; the system includes:

[0151] An acquisition generating unit is used to acquire first data of a detection point corresponding to the vehicle in the scene, and generate a longitudinal speed of the vehicle corresponding to the first data according to the first data; wherein the first data includes detection point distance data, detection point speed data and detection point angle data;

[0152] A first generating unit is used to generate a static point velocity distribution corresponding to the vehicle body according to the longitudinal velocity of the vehicle and second data corresponding to the vehicle body; wherein the second data includes vehicle speed data, vehicle deflection angle data, vehicle longitudinal acceleration data and vehicle wheel speed data;

[0153] The second generating unit is used to generate the vehicle speed data corresponding to the vehicle body in real time according to the static point speed distribution.

[0154] Furthermore, the calculation formula of the longitudinal speed of the vehicle is as follows:

[0155] V otg =V / cosθ (1)

[0156] Among them, V otg Z is the longitudinal velocity; V is the radial velocity of the detection point, and θ is the azimuth angle of the detection point;

[0157] The first generating unit further includes:

[0158] A first creation module is used to create a first threshold value corresponding to the dynamic and static attributes of the detection point; wherein the first threshold value is used to determine the dynamic and static attributes of the detection point;

[0159] A first determination module, used for determining and generating first detection point data corresponding to the longitudinal speed of the vehicle and less than the first threshold value;

[0160] The calculation formula of the first threshold is as follows:

[0161] Threshold1=V err +cosθ·lon+sinθ·lat (2)

[0162] Among them, V err is the minimum baseline error, lon is the longitudinal velocity correction, and lat is the lateral velocity correction;

[0163] And / or, the first generating unit further includes:

[0164] A second creation module is used to create a second threshold value corresponding to the dynamic and static attributes of the detection point according to the vehicle scene; wherein the second threshold value is a threshold value of the number of valid points that meets the first threshold value;

[0165] A second determination module is used to determine and generate second detection point data corresponding to the longitudinal speed of the vehicle and less than the second threshold value;

[0166] And / or, the first generating unit further includes:

[0167] A third creation module is used to create histogram parameters corresponding to the histogram in combination with histogram statistics; wherein the histogram parameters include range, group interval and number of groups;

[0168] A first generating module, used to generate group data corresponding to the detection points according to the histogram parameters; wherein the group data includes the group number of each detection point;

[0169] The second generating module is used to compare the number of detection points in each group in turn according to the group data, and generate group number data corresponding to the number of detection points; wherein the group number data is the group number of the detection point array;

[0170] And / or, the second generating unit further includes:

[0171] The first processing module is used to combine linear filtering to smooth the deflection angle data corresponding to the vehicle body;

[0172] The third determination module is used to create a third threshold value corresponding to the longitudinal acceleration of the vehicle, and determine whether the vehicle speed data is abnormal according to the third threshold value; wherein the third threshold value is 30% of the maximum longitudinal acceleration of the vehicle.

[0173] In the system solution embodiment of the present invention, the method steps involved in the vehicle speed estimation system applicable to multiple scenarios have been described above in detail, that is, the functional modules in the system are used to implement the steps or sub-steps in the above method embodiment, which will not be repeated here.

[0174] To achieve the above objectives, the present invention also provides a vehicle speed estimation platform applicable to multiple scenarios, such as Figure 5 As shown, it includes a processor, a memory, and a self-vehicle speed estimation platform control program applicable to multiple scenarios; wherein the self-vehicle speed estimation platform control program applicable to multiple scenarios is executed by the processor, the self-vehicle speed estimation platform control program applicable to multiple scenarios is stored in the memory, and the self-vehicle speed estimation platform control program applicable to multiple scenarios implements the self-vehicle speed estimation method steps applicable to multiple scenarios. For example:

[0175] S1, obtaining first data of a detection point corresponding to the vehicle in the scene, and generating a longitudinal speed of the vehicle corresponding to the first data according to the first data; wherein the first data includes detection point distance data, detection point speed data and detection point angle data;

[0176] S2, generating a static point velocity distribution corresponding to the vehicle body according to the longitudinal velocity of the vehicle and combining the second data corresponding to the vehicle body; wherein the second data includes the vehicle velocity data, the vehicle deflection angle data, the longitudinal acceleration data of the vehicle and the wheel speed data of the vehicle;

[0177] S3. Generate vehicle speed data corresponding to the vehicle body in real time according to the stationary point speed distribution.

[0178] The specific details of the steps have been explained above and will not be repeated here.

[0179] In the embodiment of the present invention, the built-in processor of the self-vehicle speed estimation platform applicable to multiple scenarios can be composed of integrated circuits, for example, it can be composed of a single packaged integrated circuit, or it can be composed of multiple integrated circuits with the same function or different functions, including one or more central processing units (CPU), microprocessors, digital processing chips, graphics processors and various control chips. The processor uses various interfaces and lines to connect various components, and executes or executes programs or units stored in the memory, and calls data stored in the memory to perform various functions and process data for self-vehicle speed estimation applicable to multiple scenarios;

[0180] The memory is used to store program codes and various data. It is installed in the self-driving vehicle speed estimation platform suitable for multiple scenarios, and realizes high-speed and automatic access to programs or data during operation.

[0181] The memory includes read-only memory (ROM), random access memory (RAM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), one-time programmable read-only memory (OTPROM), electronically erasable programmable read-only memory (EEPROM), compact disc read-only memory (CD-ROM) or other optical disc storage, magnetic disk storage, magnetic tape storage, or any other computer-readable medium that can be used to carry or store data.

[0182] The present invention obtains first data of a detection point corresponding to the self-vehicle in a scene, and generates the longitudinal speed of the self-vehicle corresponding to the first data according to the first data; wherein the first data includes detection point distance data, detection point speed data and detection point angle data; according to the longitudinal speed of the self-vehicle, in combination with second data corresponding to the self-vehicle body, generates the static point speed distribution corresponding to the self-vehicle body; wherein the second data includes self-vehicle speed data, self-vehicle deflection angle data, self-vehicle longitudinal acceleration data and self-vehicle wheel speed data; according to the static point speed distribution, the self-vehicle speed data corresponding to the self-vehicle body is generated in real time, as well as the system and platform corresponding to the method, so as to achieve the effect of obtaining the accurate speed of the self-vehicle in real time, that is, a more accurate vehicle speed can be obtained when the acceleration of the self-vehicle is large (starting, emergency braking, etc.).

[0183] The above-mentioned embodiments only express several implementation methods of the present invention, and the description thereof is relatively specific and detailed, but it cannot be understood as limiting the scope of the patent of the present invention. It should be pointed out that, for ordinary technicians in this field, several variations and improvements can be made without departing from the concept of the present invention, which all belong to the protection scope of the present invention. Therefore, the protection scope of the patent of the present invention shall be subject to the attached claims.

Claims

1. A vehicle speed estimation method applicable to multiple scenarios, characterized in that: The method comprises the following steps: Acquire first data of a detection point corresponding to the vehicle in the scene, and generate a longitudinal speed of the vehicle corresponding to the first data according to the first data; wherein the first data includes detection point distance data, detection point speed data, and detection point angle data; According to the longitudinal speed of the own vehicle, combined with the second data corresponding to the own vehicle body, a static point speed distribution corresponding to the own vehicle body is generated; wherein the second data includes the own vehicle speed data, the own vehicle deflection angle data, the own vehicle longitudinal acceleration data and the own vehicle wheel speed data; Generating vehicle speed data corresponding to the vehicle body in real time according to the stationary point speed distribution; Specifically, step 1: Get all detection points of the current frame Information, Information includes: distance, speed and angle, And calculate the longitudinal velocity of each detection point: Where: is the radial velocity of the detection point, is the azimuth of the detection point; Step 2: Obtain some vehicle body information from the vehicle body sensor: vehicle speed , deflection angle , longitudinal acceleration of the vehicle , wheel speed ; Step 3: Use a histogram to calculate the velocity distribution of the absolute stationary point: First, select valid data, that is, the longitudinal velocity satisfies The detection point, is the threshold used to determine the dynamic and static properties of the detection point: Where: is the minimum baseline error, is the longitudinal velocity correction, For lateral speed correction, the above values ​​are set through offline data statistics; Secondly, set the histogram parameters: range, group interval and number of groups; Setting the range : The worst setting only needs to consider the target speed range, the default maximum speed For 50 m / s , minimum speed is 0 m / s; Set group distance , : The setting of the group interval needs to consider both the speed result accuracy and the amount of calculation. The smaller the group interval value, the more accurate the final speed obtained by the histogram statistics, and the larger the corresponding amount of calculation. It can be adjusted according to actual usage; Set the number of groups: ; Again, determine whether the number of valid detection points meets the set threshold The threshold is adjusted according to the actual application scenario. Considering that there are fewer static targets in some scenes, it is recommended to set it to 1~10; if the number of valid detection points meets the threshold , then continue processing, otherwise go to step 4; Again, calculate the group number of each detection point: Finally: count the number of detection points that fall into each group, compare the number of detection points in each group in turn, and get the group number of the group with the largest number of detection points And calculate the estimated speed: Step 4: Use linear filtering to calculate the deflection angle Smoothing Use the following formula: Where: is the measurement value obtained from the vehicle body information in the current frame, is the smoothing value of the previous frame, is the smoothing value of the current frame, is the linear coefficient, where Adjust according to actual usage so that the variance of the deflection angle after smoothing is within a certain range; Step 5: Determine the validity of the wheel speed information, that is, eliminate abnormal wheel speed data; if the vehicle speed and wheel speed meet the following conditions, the wheel speed information is considered invalid: Where: is the left rear wheel speed, is the right rear wheel speed, It is scientific notation, which means 1 times 10 to the negative sixth power, that is, 0.000001; is the vehicle speed; The above conditions mean that if the vehicle is in motion and the left rear / right rear wheel speed is 0, the wheel speed information is considered invalid, and the process goes to step 6; otherwise, the wheel speed information is considered valid, and the process goes to step 7; Step 6: Calculate the final estimated ego vehicle speed ; If the wheel speed information is determined to be invalid in step 5, two situations are considered: 1) the valid detection point in step 3 does not meet the threshold; 2) and There are large differences between ; If any of the above two conditions are met, then ,otherwise ; Step 7: If the wheel speed information is determined to be valid in step 5, proceed to this step: First, calculate the vehicle speed based on the left rear wheel speed and the right rear wheel speed : Secondly: Determine the longitudinal acceleration of the vehicle Whether the set threshold is met , the threshold needs to be set according to offline data statistics, and is generally set to 30% of the maximum longitudinal acceleration of the vehicle; at this time, three situations are considered: 1) the effective detection point in step 3 does not meet the threshold; 2) and There are large differences between ; 3) Satisfy the emergency braking conditions, that is Greater than 60% of the maximum longitudinal acceleration; if the longitudinal acceleration meets the threshold and any of the above three conditions is met, then ,otherwise ; Finally: If Dissatisfied ,At this time, two situations are considered: 1) Turning scenario, that is Larger, , Need to be adjusted according to the actual application scenario; 2) and There are large differences between ; If both of the above conditions are met, then ,otherwise .

2. A vehicle speed estimation system suitable for multiple scenarios, characterized in that: The system is applied to the vehicle speed estimation method as claimed in claim 1; the system comprises: An acquisition generating unit is used to acquire first data of a detection point corresponding to the vehicle in the scene, and generate a longitudinal speed of the vehicle corresponding to the first data according to the first data; wherein the first data includes detection point distance data, detection point speed data and detection point angle data; A first generating unit is used to generate a static point velocity distribution corresponding to the vehicle body according to the longitudinal velocity of the vehicle and second data corresponding to the vehicle body; wherein the second data includes vehicle speed data, vehicle deflection angle data, vehicle longitudinal acceleration data and vehicle wheel speed data; The second generating unit is used to generate the vehicle speed data corresponding to the vehicle body in real time according to the static point speed distribution.

3. The vehicle speed estimation system applicable to multiple scenarios according to claim 2, characterized in that: The calculation formula of the longitudinal velocity of the vehicle is as follows: in, Longitudinal speed; is the radial velocity of the detection point, is the azimuth of the detection point; The first generating unit further includes: A first creation module is used to create a first threshold value corresponding to the dynamic and static attributes of the detection point; wherein the first threshold value is used to determine the dynamic and static attributes of the detection point; The first determination module is used to determine and generate first detection point data corresponding to the longitudinal speed of the vehicle and less than the first threshold value; the calculation formula of the first threshold value is as follows: in, is the minimum baseline error, is the longitudinal velocity correction, is the lateral velocity correction; And / or, the first generating unit further includes: A second creation module is used to create a second threshold value corresponding to the dynamic and static attributes of the detection point according to the vehicle scene; wherein the second threshold value is a threshold value of the number of valid points that meets the first threshold value; A second determination module is used to determine and generate second detection point data corresponding to the longitudinal speed of the vehicle and less than the second threshold value; And / or, the first generating unit further includes: A third creation module is used to create histogram parameters corresponding to the histogram in combination with histogram statistics; wherein the histogram parameters include range, group interval and number of groups; A first generating module, used to generate group data corresponding to the detection points according to the histogram parameters; wherein the group data includes the group number of each detection point; The second generating module is used to compare the number of detection points in each group in turn according to the group data, and generate group number data corresponding to the number of detection points; wherein the group number data is the group number of the detection point array; And / or, the second generating unit further includes: The first processing module is used to combine linear filtering to smooth the deflection angle data corresponding to the vehicle body; The third determination module is used to create a third threshold value corresponding to the longitudinal acceleration of the vehicle, and determine whether the vehicle speed data is abnormal according to the third threshold value; wherein the third threshold value is 30% of the maximum longitudinal acceleration of the vehicle.

4. A self-driving vehicle speed estimation platform suitable for multiple scenarios, characterized in that: The invention comprises a processor, a memory and a self-vehicle speed estimation platform control program applicable to multiple scenarios; wherein the self-vehicle speed estimation platform control program applicable to multiple scenarios is executed by the processor, the self-vehicle speed estimation platform control program applicable to multiple scenarios is stored in the memory, and the self-vehicle speed estimation platform control program applicable to multiple scenarios implements the self-vehicle speed estimation method applicable to multiple scenarios as claimed in claim 1.