Driving baseline calculation method, system, vehicle, equipment and storage medium

Abstract

The present invention relates to the field of autonomous driving technology and proposes a driving baseline calculation method, system, vehicle, equipment, and storage medium, which include the following steps: obtaining current lane line data according to a preset calculation cycle; judging the existence and reliability of the current left and right lane lines based on the lane line data, and predicting the lane line data; selectively filtering the lane line data based on the judgment results on the existence and reliability of the lane lines and the predicted values ​​of the lane line data; calculating the lane line width based on the selectively filtered lane line data, and calculating the current driving baseline data based on the lane line width. The present invention selectively filters based on the reliability and existence of the lane line data of the current vehicle, which can effectively reduce computing resources, shorten computing delays, and thereby shorten the LKA and ELK reaction times, thereby improving the vehicle's driving return efficiency and improving the safety of autonomous driving.

Description

Technical Field

[0001] The present invention relates to the field of autonomous driving technology, and more specifically, to a driving baseline calculation method, system, vehicle, device, and storage medium. Background Art

[0002] With the continuous increase in the number of cars, traffic safety has been under tremendous pressure and challenges. During driving, drivers experience driving fatigue due to excessive brain fatigue and physical exhaustion caused by long-term driving, road environment and other factors. At this time, the driver's body reaction speed, attention and driving skills decrease, which can easily lead to traffic accidents.

[0003] The application of autonomous driving systems has significantly promoted the development of intelligent transportation. By collecting real-time lane markings and road object data around the vehicle, the system can determine the location of objects on the road, enabling it to quickly identify potential hazards and take effective preventative measures. Lane Keeping Assist (LKA) and Emergency Lane Keeping (ELK) are driver assistance features within autonomous driving systems, designed to improve driving safety. Currently, methods have been proposed to determine driving baselines for LKA and ELK. These methods calculate driving baseline data based on a driving baseline model, which is then used to assist in planning and calculating return trajectory parameters. However, existing baseline calculation methods involve extensive computational resources, impose high hardware requirements, and introduce computational latency, which significantly prolongs the response time of LKA and ELK, making it difficult to ensure the safety of autonomous driving in emergency situations. Summary of the Invention

[0004] In order to overcome the defects of the driving baseline calculation method described in the above-mentioned prior art, which involves a large amount of calculation and causes calculation delay, prolongs the LKA and ELK reaction time, and affects the safety of autonomous driving, the present invention provides a driving baseline calculation method, system, vehicle, equipment and storage medium.

[0005] In order to solve the above technical problems, the technical solutions of the present invention are as follows:

[0006] A driving reference line calculation method includes the following steps:

[0007] Obtain current lane line data according to a preset calculation cycle;

[0008] Determining the existence and reliability of the current left and right lane lines based on the lane line data, and predicting the lane line data;

[0009] selectively filtering the lane line data based on a determination result regarding the existence and reliability of the lane line and a predicted value of the lane line data;

[0010] The lane line width is calculated according to the lane line data that has been selectively filtered, and the current driving reference line data is calculated based on the lane line width.

[0011] This technical solution selectively filters the reliability and existence of the lane data currently being driven by the vehicle, and then performs calculations based on this filtered lane data. This effectively reduces computing resources and latency, thereby shortening the LKA and ELK response times. This technical solution is applicable when a vehicle deviates from the operating range of LKA and ELK but meets the activation conditions for these two systems. By calculating and updating the current baseline, the vehicle's return trajectory parameters are further planned to keep the vehicle within the baseline.

[0012] As a preferred solution, the lane line data includes the distance between the vehicle coordinates and the lane line, the heading angle of the lane line, the curvature of the lane line, the curvature change rate of the lane line, the starting distance of the lane line, the ending distance of the lane line and the segmentation confidence of the left lane line and the right lane line; wherein, the distance between the vehicle coordinates and the lane line includes the distance from the left lane line or the right lane line with the current vehicle coordinates as the coordinate origin; the driving baseline data includes the distance between the vehicle coordinates and the baseline, the heading angle of the baseline, the curvature of the baseline, and the curvature change rate of the baseline.

[0013] As a preferred solution, judging the existence and reliability of the current left and right lane lines based on the lane line data includes:

[0014] Determine the existence of the current left and right lane lines: If the distance between the vehicle coordinates and the lane line in the lane line data is greater than 0.01, and the lane line length is greater than the preset lane line length threshold, then the corresponding lane line is determined to exist, and the minimum confidence ParseCon that meets the lane line reliability requirements is calculated based on the current vehicle speed. min , and the minimum lane line length Range that meets the lane line reliability requirements min , further judge the reliability of the lane line; its expression is as follows:

[0015] ParseCon min =min(max(h_×3.6+10,20),50)

[0016] Range min =max(20,15×h_)

[0017] Where h_ is the current vehicle speed; otherwise, it is determined that the corresponding lane line does not exist;

[0018] The lane line length is calculated by the difference between the lane line end distance and the lane line start distance;

[0019] Determine the reliability of the current left and right lane lines: If the lane line length is greater than the minimum lane line length Range min , and the lane segmentation confidence is greater than the minimum confidence ParseCon min , the corresponding lane line is determined to be reliable; otherwise, the corresponding lane line is determined to be unreliable.

[0020] As a preferred solution, the predicting of the lane line data includes:

[0021] Based on the currently collected vehicle speed and yaw rate, as well as the distance between the vehicle coordinates and the lane line, the heading angle of the lane line, the curvature of the lane line, and the curvature change rate of the lane line obtained in the previous calculation cycle, a prediction calculation is performed on the data category that is not 0 in the lane line data obtained in the previous calculation cycle to obtain the distance between the vehicle coordinates and the lane line, the heading angle of the lane line, the curvature of the lane line, and / or the curvature change rate of the lane line predicted in the current calculation cycle; wherein, the expressions for performing the prediction calculation on the corresponding lane line data are respectively as follows:

[0022] CoefPre(,1)=

[0023] PrevCoef(,1)+sin(PrevCoef(,1))×hv_spd×SampleTime

[0024] CoefPre(,2)=

[0025] PrevCoef(,2)+(YawRate_rad+2×hv_spd×PrevCoef(s,3))×SampleTime

[0026] CoefPre(,3)=

[0027] PrevCoef(,3)+3×h_×SampleTime×PrevCoef(,4)

[0028] CoefPre(,4)=PrevCoef(,4)

[0029] Where s∈{1,2}, s represents the left lane line and the right lane line respectively; n∈{1,2,3,4}, n represents the category of the corresponding lane line data; CoefPre(,n) represents the predicted n-th corresponding side lane line data; PrevCoef(,n) represents the n-th corresponding side lane line data obtained in the previous calculation cycle; SampleTime represents the operation cycle; YawRate_rad represents the yaw angular velocity collected previously.

[0030] As a preferred solution, the selective filtering of the lane line data includes:

[0031] Detect the current vehicle lateral control activation status;

[0032] If the judgment result is that the lane line is reliable and the current vehicle lateral control is in an activated state, filtering the distance between the vehicle coordinates and the lane line, the heading angle of the lane line, the curvature of the lane line, and the curvature change rate of the lane line, and outputting the filtered lane line data;

[0033] If the judgment result is that the lane line is unreliable and the current vehicle lateral control is in an activated state, filtering the distance between the vehicle coordinates and the lane line, and outputting the filtered distance value between the vehicle coordinates and the lane line;

[0034] The lane line data is filtered, and its expression is:

[0035] LnCoef(,n)=FiltCoef×LnMkMat(s,n)+(1-FiltCoef)×CoefPre′(s′,n)

[0036] Where FiltCoef is the filter coefficient; LnMkMat(s,n) is the lane line matrix composed of the left and right lane line data; CoefPre′(′,n) represents the n-th lane line data predicted in the previous calculation cycle, s′∈{1,2} and s′≠s, s′ represents the left lane line and the right lane line respectively; LnCoef(,n) represents the filtering result of the n-th lane line data of the corresponding side lane line;

[0037] If the judgment result is that the lane line does not exist, or the current vehicle lateral control is in an inactive state, the lane line data is not filtered, that is:

[0038] LnCoef(,n)=LnMkMat(s,n).

[0039] As a preferred solution, the lane line width is calculated based on the lane line data that has been selectively filtered, including:

[0040] If the judgment result is that the lane line exists and is reliable, the lane line width LnWidthRaw is calculated based on the distance between the vehicle coordinates and the lane line, and its expression is:

[0041] LnWidthRaw=|LnCoef(1,1)-LnCoef(2,1)|

[0042] Where LnWidthRaw represents the lane width, LnCoef(1,1) represents the distance between the vehicle coordinates and the left lane line, and LnCoef(2,1) represents the distance between the vehicle coordinates and the right lane line.

[0043] If the judgment result is that the lane line does not exist or is unreliable, the lane line width LnWidthRaw is assigned to a preset fixed value.

[0044] As a preferred solution, the current driving reference line data is calculated based on the lane line width LnWidthRaw, wherein:

[0045] For n=1:

[0046] Coef(,n)=LnCoef(,1)-0.5

[0047] ×max(min(max(LnWidthRaw,K_LaneWidthMin),LnWd)

[0048] -K_BaseLineOffset×2,2.8)

[0049] For n=2 or 3 or 4:

[0050] Coef(,n)=LnCoef(,n)

[0051] Where Coef(,n),s∈{1,2},n∈{1,2,3,4},represent the distance between the vehicle coordinates and the left and right baselines, the heading angle of the baseline, the curvature of the baseline, and the curvature change rate of the baseline, respectively. K_LaneWidthMin represents the lane width after exiting LKA or ELK, LnWd represents the historical lane width, and K_BaseLineOffset represents the offset distance of the baseline.

[0052] As a preferred solution, the method further includes: accumulating the prediction time of the lane line data in the current calculation cycle, and calculating the current baseline confidence based on the accumulated prediction time to evaluate the reliability of the current driving baseline data; wherein:

[0053] If the judgment result is that the lane line is reliable, the accumulated prediction time is set to zero and the baseline confidence is assigned a value of 2;

[0054] If the judgment result is that the lane line is unreliable or does not exist, the cumulative prediction time is executed; if the cumulative prediction time is (0, 1], the baseline confidence is assigned to 1; otherwise, the baseline confidence is assigned to 0.

[0055] Furthermore, the present invention also proposes a driving reference line calculation system, which applies the driving reference line calculation method proposed in the present invention. The system includes:

[0056] A data acquisition module, configured to acquire and output lane line data of a currently traveling vehicle according to a preset calculation cycle;

[0057] a data judgment module, configured to judge the existence and reliability of the current left and right lane lines based on the lane line data, and output the judgment result on the existence and reliability of the lane lines;

[0058] a data prediction module, configured to predict the lane line data and store the predicted value of the lane line data in the current calculation cycle;

[0059] a filtering module, configured to selectively filter the lane line data according to the judgment result and the predicted value of the lane line data in the previous calculation cycle, and output a selective filtering result;

[0060] A data processing module is used to calculate the lane line width according to the selective filtering result, and calculate and output the current driving reference line data based on the lane line width.

[0061] As a preferred embodiment, the system further includes a baseline confidence calculation module connected to the data prediction module, which is used to obtain the cumulative prediction time of the current calculation cycle, and calculate the current baseline confidence based on the cumulative prediction time and output it to an external active safety system for the active safety system to evaluate the reliability of the current driving baseline data.

[0062] Furthermore, the present invention also proposes a vehicle equipped with an automatic driving system, which is connected to the driving baseline calculation system proposed in the present invention, and is used to control the driving state of the vehicle according to the current driving baseline data output by the driving baseline calculation system.

[0063] Furthermore, the present invention also proposes a computer device, including a memory and a processor, wherein the memory stores computer-readable instructions, and when the computer-readable instructions are executed by the processor, the processor executes the steps of the driving baseline calculation method proposed in the present invention.

[0064] Furthermore, the present invention also proposes a storage medium having computer-readable instructions stored thereon, which, when executed by a processor, implements the steps of the driving baseline calculation method proposed in the present invention.

[0065] Compared with the existing technology, the technical solution of the present invention has the following beneficial effects: the present invention performs selective filtering based on the reliability and existence of the lane line data currently traveled by the vehicle, which can effectively reduce computing resources and shorten computing latency, thereby shortening the reaction time of LKA and ELK, thereby improving the efficiency of vehicle driving return and enhancing the safety of autonomous driving; the present invention is applicable to the case where the vehicle deviates from the operating range of LKA and ELK and meets the activation conditions of LKA and ELK. By calculating and updating the current driving baseline, the current baseline is used to further plan the vehicle's return trajectory parameters to keep the vehicle within the baseline. BRIEF DESCRIPTION OF THE DRAWINGS

[0066] Figure 1 This is a flow chart of the driving reference line calculation method of Example 1.

[0067] Figure 2 Schematic diagram of the selective filtering process of Example 1.

[0068] Figure 3 This is an architectural diagram of the driving baseline calculation system of Example 2. DETAILED DESCRIPTION

[0069] The accompanying drawings are for illustrative purposes only and should not be construed as limiting this patent. It is understandable that some descriptions of the accompanying drawings may be omitted in order to better illustrate this embodiment.

[0070] The technical solution of the present invention is further described below with reference to the accompanying drawings and embodiments.

[0071] Example 1

[0072] This embodiment proposes a driving reference line calculation method, such as Figure 1 FIG. 1 is a flow chart of the driving reference line calculation method of this embodiment.

[0073] The driving reference line calculation method proposed in this embodiment includes the following steps:

[0074] S1. Obtain current lane line data according to a preset calculation cycle.

[0075] S2. Determine the existence and reliability of the current left and right lane lines based on the lane line data, and predict the lane line data.

[0076] S3. Selectively filter the lane line data based on the judgment result on the existence and reliability of the lane line and the predicted value of the lane line data.

[0077] S4. Calculate the lane line width according to the lane line data that has undergone selective filtering, and calculate the current driving reference line data based on the lane line width.

[0078] As an example, the vehicle can collect and obtain lane line data through the on-board unit. Optionally, the vehicle's lane line data can be output to the ECU or processor through driving environment detection devices such as cameras, millimeter-wave radars, lidars, sonars, GPS and various sensors for further calculation of the driving baseline.

[0079] This embodiment applies to situations where a vehicle deviates from the operating range of LKA and ELK and meets the activation conditions for LKA and ELK. By calculating and updating the current driving baseline, it further plans the vehicle's return trajectory parameters to keep the vehicle within the baseline. This embodiment selectively filters the reliability and existence of the lane data currently being driven by the vehicle, effectively reducing computing resources and latency. This, in turn, shortens the response time of LKA and ELK, improving the efficiency of vehicle return and enhancing autonomous driving safety.

[0080] Furthermore, in an optional embodiment, the lane data includes the distance A0 between the vehicle coordinates and the lane line, the lane line heading angle A1, the lane line curvature A2, the lane line curvature change rate A3, as well as the lane line starting distance A4, the lane line ending distance A5, and the segmentation confidence A6. Optionally, the lane line type, measurement type, and other data are also included.

[0081] The distance A0 between the vehicle coordinates and the lane line includes the distance between the current vehicle coordinates and the left lane line or the right lane line with the current vehicle coordinates as the coordinate origin.

[0082] The driving reference line data includes the distance between the vehicle coordinates and the reference line, the heading angle of the reference line, the curvature of the reference line, and the curvature change rate of the reference line.

[0083] The distance between the vehicle coordinates and the reference line includes the distance between the current vehicle coordinates and the left reference line or the right reference line, taking the current vehicle coordinates as the coordinate origin.

[0084] As an exemplary explanation, in the above lane line data and driving reference line data, the central position of the vehicle is used as the coordinate origin, the left and right directions of the vehicle are used as the x-coordinate axis, and the front and rear directions are used as the y-coordinate axis. Optionally, the x-coordinate value to the left of the origin is set to a positive value, and the y-coordinate value in front of the origin is set to a positive value.

[0085] like Figure 2 FIG. 1 is a schematic diagram of a specific process of calculating the driving reference line in this embodiment.

[0086] Furthermore, in an optional embodiment, when step S2 determines the existence and reliability of the current left and right lane lines based on the lane line data, it specifically includes the following steps:

[0087] S201. Determine the existence of the current left and right lane lines: If the distance between the vehicle coordinates and the lane line in the lane line data is greater than 0.01, and the lane line length is greater than the preset lane line length threshold, then determine that the corresponding lane line exists, and calculate the minimum confidence ParseCon that meets the lane line reliability requirements based on the current vehicle speed. min , and the minimum lane line length Range that meets the lane line reliability requirements min , further judge the reliability of the lane line; its expression is as follows:

[0088] ParseCon min =min(max(h_×3.6+10,20),50)

[0089] Range min =max(20,15×h_)

[0090] Where h_ is the current vehicle speed; otherwise, it is determined that the corresponding lane line does not exist.

[0091] The lane line length is calculated by the difference between the lane line end distance and the lane line start distance.

[0092] S202, determine the reliability of the current left and right lane lines: if the lane line length is greater than the minimum lane line length Range min , and the lane segmentation confidence is greater than the minimum confidence ParseCon min , the corresponding lane line is determined to be reliable; otherwise, the corresponding lane line is determined to be unreliable.

[0093] In this embodiment, the existence of the left and right lane lines is first judged. When the distance between the vehicle coordinates and the lane line is greater than 0.01m and the lane line length is greater than the preset lane line length threshold, the lane line on this side is considered to exist.

[0094] As an exemplary illustration, the lane line length threshold in this embodiment is set to 10m.

[0095] When it is determined that the corresponding lane line exists, it is necessary to further determine the reliability of the corresponding lane line. In this embodiment, based on the ParseCon min , and the Range that meets the lane line reliability requirementsmin , judge the reliability of the left and right lane lines, when the lane line length is greater than Range min , and the lane segmentation confidence is greater than ParseCon min When , the lane marking on this side is considered reliable.

[0096] Furthermore, in an optional embodiment, when predicting the lane line data, the following steps are specifically included:

[0097] Based on the currently collected vehicle speed and yaw rate, as well as the distance PrevCoef(,1) between the vehicle coordinates and the lane line, the heading angle PrevCoef(,2) of the lane line, the curvature PrevCoef(,3) of the lane line, and the curvature change rate PrevCoef(,4) of the lane line obtained in the previous calculation cycle, a prediction calculation is performed on the data categories that are not zero in the lane line data obtained in the previous calculation cycle to obtain the distance CoefPre(,1) between the vehicle coordinates and the lane line, the heading angle CoefPre(,2) of the lane line, the curvature CoefPre(,3) of the lane line, and / or the curvature change rate CoefPre(,4) of the lane line predicted in the current calculation cycle.

[0098] The expressions for predicting and calculating the corresponding lane line data are as follows:

[0099] CoefPre(,1)=

[0100] PrevCoef(,1)+sin(PrevCoef(,1))×hv_spd×SampleTime

[0101] CoefPre(,2)=

[0102] PrevCoef(,2)+(YawRate_rad+2×hv_spd×PrevCoef(s,3))×SampleTime

[0103] CoefPre(,3)=

[0104] PrevCoef(,3)+3×h_×SampleTime×PrevCoef(,4)

[0105] CoefPre(,4)=PrevCoef(,4)

[0106] Where s∈{1,2}, s represents the left lane line and the right lane line respectively; n∈{1,2,3,4}, n represents the category of the corresponding lane line data; CoefPre(,n) represents the predicted n-th corresponding side lane line data; PrevCoef(,n) represents the n-th corresponding side lane line data obtained in the previous calculation cycle; SampleTime represents the operation cycle; YawRate_rad represents the yaw angular velocity collected previously.

[0107] Preferably, the operation cycle SampleTime in this embodiment is set to a value of 0.02 seconds.

[0108] The above steps may be performed after any step of the driving baseline calculation method proposed in this embodiment. As an exemplary illustration, this embodiment performs the above steps after the lane line existence and reliability judgment step.

[0109] In this embodiment, corresponding cubic curve parameters can be obtained by predicting the lane line data, which can be further used to perform selective filtering calculation on the lane line data in the next calculation cycle.

[0110] Furthermore, in an optional embodiment, when selectively filtering the lane line data in step S3, the following steps are specifically included:

[0111] S301, detecting the current vehicle lateral control activation state.

[0112] S302: If the judgment result is that the lane line is reliable and the vehicle lateral control is currently activated, filtering the distance A0 between the vehicle coordinates and the lane line, the heading angle A1 of the lane line, the curvature A2 of the lane line, and the curvature change rate A3 of the lane line, and outputting the filtered lane line data;

[0113] If the judgment result is that the lane line is unreliable and the current vehicle lateral control is in an activated state, the distance A0 between the vehicle coordinates and the lane line is filtered, and the filtered distance value between the vehicle coordinates and the lane line is output; the lane line heading angle A1, the lane line curvature A2, and the lane line curvature change rate A3 use the original lane line data.

[0114] Among them, the lane line data is filtered, and its expression is:

[0115] LnCoef(,n)=FiltCoef×LnMkMat(s,n)+(1-FiltCoef)×CoefPre′(s′,n)

[0116] Where FiltCoef is the filter coefficient; LnMkMat(s,n) is the lane line matrix composed of the left and right lane line data; CoefPre′(′,n) represents the n-th lane line data predicted in the previous calculation cycle, s′∈{1,2} and s′≠s, s′ represents the left lane line and the right lane line respectively; LnCoef(,n) represents the filtering result of the n-th lane line data of the corresponding side lane line.

[0117] S303. If the judgment result is that the lane line does not exist, or the current vehicle lateral control is in an inactive state, the lane line data is not filtered, and the original lane line data is used for the distance A0 between the vehicle coordinates and the lane line, the heading angle A1 of the lane line, the curvature A2 of the lane line, and the curvature change rate A3 of the lane line, that is:

[0118] LnCoef(,n)=LnMkMat(s,n).

[0119] As an exemplary illustration, the filter coefficient FiltCoef in this embodiment is set to 0.5.

[0120] In this embodiment, the activation state of the vehicle's lateral control means that the vehicle's lateral control is in an activated state when the activation conditions of the active safety system in the automatic driving system (such as the lane departure warning system, LKA, ELK) are met, and the activation conditions of the vehicle's EPS (Electronic-Power-Steering) system are met.

[0121] Furthermore, in an optional embodiment, when calculating the lane line width based on the lane line data that has undergone selective filtering in step S4, the following steps are included:

[0122] (1) If the judgment result is that the lane line exists and is reliable, the lane line width LnWidthRaw is calculated based on the distance between the vehicle coordinates and the lane line, and its expression is:

[0123] LnWidthRaw=|LnCoef(1,1)-LnCoef(2,1)|

[0124] Where LnWidthRaw represents the lane width, LnCoef(1,1) represents the distance between the vehicle coordinates and the left lane line, and LnCoef(2,1) represents the distance between the vehicle coordinates and the right lane line.

[0125] (2) If the judgment result is that the lane line does not exist or is unreliable, the lane line width LnWidthRaw is assigned a preset fixed value.

[0126] Preferably, when the lane line does not exist or is unreliable, the lane line width LnWidthRaw is assigned a value of 2.8m.

[0127] Furthermore, when the lane line width LnWidthRaw exceeds the lane line width range of LKA and ELK and is less than the historical lane line width value LnWd, the historical lane line width value LnWd is set as the lane line width LnWidthRaw.

[0128] Among them, because the activation requirements of LKA and ELK are strongly related to the distance between the vehicle's front wheels and the lane line and the lane line width, it is necessary to use the historical lane line width to limit the distance between the vehicle's front wheels and the lane line required for LKA and ELK activation to prevent affecting the activation conditions of LKA and ELK.

[0129] This embodiment updates the historical lane line width LnWd to further update the threshold value of the distance between the front wheel of the vehicle and the lane line in the LKA and ELK activation requirements.

[0130] Furthermore, in an optional embodiment, when calculating the current driving reference line data based on the lane line width LnWidthRaw in step S4, the following steps are included:

[0131] (1) For n=1:

[0132] Coef(,n)=LnCoef(,1)-0.5

[0133] ×max(min(max(LnWidthRaw,K_LaneWidthMin),LnWd)

[0134] -K_BaseLineOffset×2,2.8)

[0135] (2) For n=2 or 3 or 4:

[0136] Coef(,n)=LnCoef(,n)

[0137] Where Coef(,n),s∈{1,2},n∈{1,2,3,4},represent the distance between the vehicle coordinates and the left and right baselines, the heading angle of the baseline, the curvature of the baseline, and the rate of change of the curvature of the baseline, respectively. K_LaneWidthMin represents the lane width for exiting LKA and ELK, and is a calibration value. LnWd represents the historical lane line width value, and K_BaseLineOffset represents the offset distance of the baseline, and is a calibration value.

[0138] Furthermore, in an optional embodiment, the method further includes: accumulating the prediction time for predicting the lane line data in the current calculation cycle, and calculating the current baseline confidence based on the accumulated prediction time to evaluate the reliability of the current driving baseline data.

[0139] Among them, if the judgment result is that the lane line is reliable, the cumulative prediction time is set to zero and the baseline confidence is assigned to 2; if the judgment result is that the lane line is unreliable or does not exist, the cumulative prediction time is executed; if the cumulative prediction time is (0,1], the baseline confidence is assigned to 1; otherwise, the baseline confidence is assigned to 0.

[0140] Further optionally, during the specific implementation process, the baseline confidence is input into an analysis and evaluation module in an active safety system (eg, lane departure warning system, LKA, ELK) for evaluating the reliability and availability of the current driving baseline data.

[0141] As an exemplary illustration, when the baseline confidence is 0, the current driving baseline data is evaluated to be unavailable; when the baseline confidence is 1, the current driving baseline data is evaluated to be available but unreliable; and when the baseline confidence is 2, the current driving baseline data is evaluated to be available and reliable.

[0142] Example 2

[0143] This embodiment proposes a driving reference line calculation system, which applies the driving reference line calculation method proposed in Example 1. Figure 3 , which is an architecture diagram of the driving baseline calculation system of this embodiment.

[0144] The driving reference line calculation system proposed in this embodiment includes:

[0145] A data acquisition module, configured to acquire and output lane line data of a currently traveling vehicle according to a preset calculation cycle;

[0146] a data judgment module, configured to judge the existence and reliability of the current left and right lane lines based on the lane line data, and output the judgment result on the existence and reliability of the lane lines;

[0147] a data prediction module, configured to predict the lane line data and output or store a predicted value of the lane line data in a current calculation cycle;

[0148] a filtering module, configured to selectively filter the lane line data according to the judgment result and the predicted value of the lane line data, and output a selective filtering result;

[0149] A data processing module is used to calculate the lane line width according to the selective filtering result, and calculate and output the current driving reference line data based on the lane line width.

[0150] Among them, the data acquisition module can be optionally connected to an on-board unit carried by the vehicle to obtain lane line data of the current driving vehicle, such as a camera, millimeter wave radar, lidar, sonar, GPS and various sensors and other on-board driving environment detection devices.

[0151] In an optional embodiment, in this embodiment, the lane line data collected in the data acquisition module includes the distance between the vehicle coordinates and the lane line in the left lane line and the right lane line, the heading angle of the lane line, the curvature of the lane line, the curvature change rate of the lane line, the starting distance of the lane line, the ending distance of the lane line and the segmentation confidence.

[0152] The distance between the vehicle coordinates and the lane line includes the distance between the current vehicle coordinates and the left lane line or the right lane line with the current vehicle coordinates as the coordinate origin.

[0153] The driving reference line data output by the data processing module includes the distance between the vehicle coordinates and the reference line, the heading angle of the reference line, the curvature of the reference line, and the curvature change rate of the reference line.

[0154] The distance between the vehicle coordinates and the reference line includes the distance between the current vehicle coordinates and the left reference line or the right reference line, taking the current vehicle coordinates as the coordinate origin.

[0155] In an optional embodiment, the data judgment module judges the existence and reliability of the left and right lane lines based on the lane line data, and makes a judgment based on the distance between the vehicle coordinates and the lane line. If the distance between the vehicle coordinates and the lane line is greater than 0.01, and the lane line length is greater than a preset lane line length threshold, the corresponding lane line is determined to exist, and the minimum confidence ParseCon that meets the lane line reliability requirement is calculated based on the current vehicle speed. min , and the minimum lane line length Range that meets the lane line reliability requirements min , further judge the reliability of the lane line; otherwise, it is determined that the corresponding lane line does not exist.

[0156] Furthermore, if the lane length is greater than the minimum lane length Range min , and the lane segmentation confidence is greater than the minimum confidence ParseCon min , the corresponding lane line is determined to be reliable; otherwise, the corresponding lane line is determined to be unreliable.

[0157] Among them, the minimum confidence ParseCon min and the minimum lane length Rangemin The expression is as follows:

[0158] ParseCon min =min(max(h_×3.6+10,20),50)

[0159] Range min =max(20,15×h_)

[0160] Where h_ is the current vehicle speed.

[0161] In an optional embodiment, when predicting the lane line data, the data prediction module performs prediction calculations on the data categories that are not 0 in the lane line data obtained in the previous calculation cycle based on the currently collected vehicle speed and yaw angular velocity, as well as the distance PrevCoef(,1) between the vehicle coordinates and the lane line, the heading angle PrevCoef(,2) of the lane line, the curvature PrevCoef(,3) of the lane line, and the curvature change rate PrevCoef(,4) of the lane line obtained in the previous calculation cycle, to obtain the distance CoefPre(,1) between the vehicle coordinates and the lane line, the heading angle CoefPre(,2) of the lane line, the curvature CoefPre(,3) of the lane line, and / or the curvature change rate CoefPre(,4) of the lane line predicted in the current calculation cycle.

[0162] The expressions for predicting and calculating the corresponding lane line data are as follows:

[0163] CoefPre(,1)=

[0164] PrevCoef(,1)+sin(PrevCoef(,1))×hv_spd×SampleTime

[0165] CoefPre(,2)=

[0166] PrevCoef(,2)+(YawRate_rad+2×hv_spd×PrevCoef(s,3))×SampleTime

[0167] CoefPre(,3)=

[0168] PrevCoef(,3)+3×h_×SampleTime×PrevCoef(,4)

[0169] CoefPre(,4)=PrevCoef(,4)

[0170] Where s∈{1,2}, s represents the left lane line and the right lane line respectively; n∈{1,2,3,4}, n represents the category of the corresponding lane line data; CoefPre(,n) represents the predicted n-th corresponding side lane line data; PrevCoef(,n) represents the n-th corresponding side lane line data obtained in the previous calculation cycle; SampleTime represents the operation cycle; YawRate_rad represents the yaw angular velocity collected previously.

[0171] In an optional embodiment, when selectively filtering lane line data, the filtering module first detects the current vehicle lateral control activation state, and then further determines: if the lane line is reliable and the vehicle lateral control is currently activated, the distance between the vehicle coordinates and the lane line, the heading angle of the lane line, the curvature of the lane line, and the curvature change rate of the lane line are filtered, and the filtered lane line data is output; if the lane line is unreliable and the vehicle lateral control is currently activated, the distance between the vehicle coordinates and the lane line is filtered, and the filtered distance value between the vehicle coordinates and the lane line is output. The expression for filtering lane line data is:

[0172] LnCoef(,n)=FiltCoef×LnMkMat(s,n)+(1-FiltCoef)×CoefPre′(s′,n)

[0173] Where FiltCoef is the filter coefficient; LnMkMat(s,n) is the lane line matrix composed of the left and right lane line data; CoefPre′(′,n) represents the n-th lane line data predicted in the previous calculation cycle, s′∈{1,2} and s′≠s, s′ represents the left lane line and the right lane line respectively; LnCoef(,n) represents the filtering result of the n-th lane line data of the corresponding side lane line.

[0174] If the lane line does not exist or the current vehicle lateral control is in an inactive state, the lane line data is not filtered, that is: LnCoef(,n) = LnMkMat(s,n).

[0175] In an optional embodiment, the data processing module performs different calculation operations according to the existence and reliability of the lane line when calculating the lane line width using the lane line data that has undergone selective filtering.

[0176] If the lane line exists and is reliable, the lane line width is calculated based on the distance between the vehicle coordinates and the lane line. The expression is:

[0177] LnWidthRaw=|LnCoef(1,1)-LnCoef(2,1)|

[0178] Where LnWidthRaw represents the lane width, LnCoef(1,1) represents the distance between the vehicle coordinates and the left lane line, and LnCoef(2,1) represents the distance between the vehicle coordinates and the right lane line.

[0179] If the lane line does not exist or is unreliable, the lane line width LnWidthRaw is assigned a preset fixed value. Optionally, the fixed value is 2.8m.

[0180] Furthermore, when the data processing module calculates the current driving reference line data based on the currently calculated lane line width LnWidthRaw,

[0181] (1) For n=1:

[0182] Coef(,n)=LnCoef(,1)-0.5

[0183] ×max(min(max(LnWidthRaw,K_LaneWidthMin),LnWd)

[0184] -K_BaseLineOffset×2,2.8)

[0185] (2) For n=2 or 3 or 4:

[0186] Coef(,n)=LnCoef(,n)

[0187] Where Coef(,n),s∈{1,2},n∈{1,2,3,4},represent the distance between the vehicle coordinates and the lane line of the left baseline and the right baseline, the heading angle of the baseline, the curvature of the baseline, and the curvature change rate of the baseline, respectively; K_LaneWidthMin represents the lane width after exiting LKA and ELK, LnWd represents the historical lane line width value, and K_BaseLineOffset represents the offset distance of the baseline.

[0188] The data processing module outputs the calculated driving baseline data for further planning the vehicle's return trajectory parameters to keep the vehicle within the baseline.

[0189] Furthermore, in an optional embodiment, the system also includes a baseline confidence calculation module, which is used to connect with the data prediction module, obtain the cumulative prediction time of the current calculation cycle, and calculate the current baseline confidence based on the cumulative prediction time and output it to an external active safety system for the active safety system to evaluate the reliability of the current driving baseline data.

[0190] In the baseline confidence calculation module of this embodiment, if the cumulative prediction time is (0, 1], the baseline confidence is assigned a value of 1; otherwise, the baseline confidence is assigned a value of 0 and output. If the judgment result is that the lane line is reliable and the current vehicle lateral control is activated, the cumulative prediction time is set to zero and the baseline confidence is assigned a value of 2.

[0191] Further, as an exemplary explanation, when the baseline confidence is 0, it is evaluated that the current driving baseline data is unavailable; when the baseline confidence is 1, it is evaluated that the current driving baseline data is available but unreliable; when the baseline confidence is 2, it is evaluated that the current driving baseline data is available and reliable.

[0192] Example 3

[0193] This embodiment proposes a vehicle equipped with an automatic driving system, which is connected to a driving baseline calculation system as proposed in Example 2, and is used to control the driving state of the vehicle based on the current driving baseline data output by the driving baseline calculation system.

[0194] Example 4

[0195] This embodiment proposes a computer device, including a memory and a processor, wherein the memory stores computer-readable instructions. When the computer-readable instructions are executed by the processor, the processor executes the steps of the driving baseline calculation method proposed in Example 1.

[0196] Example 5

[0197] This embodiment proposes a storage medium having computer-readable instructions stored thereon. When the computer-readable instructions are executed by a processor, the steps of the driving reference line calculation method proposed in Example 1 are implemented.

[0198] The description terms in the drawings are for illustrative purposes only and shall not be construed as limiting this patent;

[0199] Obviously, the above embodiments of the present invention are merely examples for the purpose of clearly illustrating the present invention, and are not intended to limit the embodiments of the present invention. Those skilled in the art will appreciate that other variations or modifications can be made based on the above description. It is not necessary and impossible to enumerate all embodiments here. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention shall be included within the scope of protection of the claims of the present invention.

Claims

1. A driving reference line calculation method, characterized in that: include: Acquire current lane line data according to a preset calculation cycle; the lane line data includes the distance between the vehicle coordinates and the lane line, the heading angle of the lane line, the curvature of the lane line, the curvature change rate of the lane line, the starting distance of the lane line, the ending distance of the lane line, and the segmentation confidence level for the left lane line and the right lane line; the driving baseline data includes the distance between the vehicle coordinates and the baseline, the heading angle of the baseline, the curvature of the baseline, and the curvature change rate of the baseline; Determining the existence and reliability of the current left and right lane lines based on the lane line data, and predicting the lane line data; selectively filtering the lane line data based on a determination result regarding the existence and reliability of the lane line and a predicted value of the lane line data; The selective filtering of the lane line data includes: Detect the current vehicle lateral control activation status; If the judgment result is that the lane line is reliable and the current vehicle lateral control is in an activated state, filtering the distance between the vehicle coordinates and the lane line, the heading angle of the lane line, the curvature of the lane line, and the curvature change rate of the lane line, and outputting the filtered lane line data; If the judgment result is that the lane line is unreliable and the current vehicle lateral control is in an activated state, filtering the distance between the vehicle coordinates and the lane line, and outputting the filtered distance value between the vehicle coordinates and the lane line; The lane line data is filtered, and its expression is: Where, is the filter coefficient; is the lane line matrix composed of the left and right lane line data; Indicates the number of predictions for the previous calculation period. n Lane line data, and , Represent the left lane line and the right lane line respectively; Indicates the lane line of the corresponding side n Filtering results of lane line data; If the judgment result is that the lane line does not exist, or the current vehicle lateral control is in an inactive state, the lane line data is not filtered, that is: The lane line width is calculated according to the lane line data that has been selectively filtered, and the current driving reference line data is calculated based on the lane line width.

2. The driving reference line calculation method according to claim 1, characterized in that: The determining of the existence and reliability of the current left and right lane lines based on the lane line data includes: Determine the existence of the current left and right lane lines: If the distance between the vehicle coordinates and the lane line in the lane line data is greater than 0.01, and the lane line length is greater than the preset lane line length threshold, then the corresponding lane line is determined to exist, and the minimum confidence level that meets the lane line reliability requirements is calculated based on the current vehicle speed. , and the minimum lane length that meets the lane reliability requirements , further judge the reliability of the lane line; its expression is as follows: Where, is the current vehicle speed; otherwise, it is determined that the corresponding lane line does not exist; The lane line length is calculated by the difference between the lane line end distance and the lane line start distance; Determine the reliability of the current left and right lane lines: If the lane line length is greater than the minimum lane line length , and the lane segmentation confidence is greater than the minimum confidence , the corresponding lane line is determined to be reliable; otherwise, the corresponding lane line is determined to be unreliable.

3. The driving reference line calculation method according to claim 2, characterized in that: The predicting of the lane line data includes: Based on the currently collected vehicle speed and yaw rate, as well as the distance between the vehicle coordinates and the lane line, the heading angle of the lane line, the curvature of the lane line, and the curvature change rate of the lane line obtained in the previous calculation cycle, a prediction calculation is performed on the data category that is not 0 in the lane line data obtained in the previous calculation cycle to obtain the distance between the vehicle coordinates and the lane line, the heading angle of the lane line, the curvature of the lane line, and / or the curvature change rate of the lane line predicted in the current calculation cycle; wherein, the expressions for performing the prediction calculation on the corresponding lane line data are respectively as follows: Where, , Represent the left lane line and the right lane line respectively; , Indicates the category of the corresponding lane line data; The predicted n Item corresponding side lane line data; Indicates the number of n Item corresponding side lane line data; Indicates the operating cycle; Indicates the yaw rate collected before.

4. The driving reference line calculation method according to claim 3, characterized in that: Calculating the lane line width based on the lane line data that has been selectively filtered includes: If the judgment result is that the lane line exists and the lane line is reliable, the lane line width is calculated based on the distance between the vehicle coordinates and the lane line. , whose expression is: Where, Indicates lane width, Indicates the distance between the vehicle coordinates and the left lane line, Indicates the distance between the vehicle coordinates and the right lane line; If the judgment result is that the lane line does not exist or is unreliable, the lane line width Assign a preset fixed value.

5. The driving reference line calculation method according to claim 4, characterized in that: Based on the lane width Calculate the current driving baseline data, where: Where, , respectively representing the distance between the vehicle coordinates and the left and right reference lines, the heading angle of the reference line, the curvature of the reference line, and the rate of change of the curvature of the reference line; Indicates the lane width for exiting LKA and ELK. Indicates the historical lane width value, Indicates the offset distance of the baseline.

6. The driving reference line calculation method according to any one of claims 2 to 5, characterized in that: The method further includes: accumulating prediction time for predicting the lane line data in a current calculation cycle, and calculating a current baseline confidence level based on the accumulated prediction time to evaluate the reliability of the current driving baseline data; wherein: If the judgment result is that the lane line is reliable, the accumulated prediction time is set to zero and the baseline confidence is assigned a value of 2; If the judgment result is that the lane line is unreliable or does not exist, the cumulative prediction time is executed; if the cumulative prediction time is , the baseline confidence is assigned a value of 1; otherwise, the baseline confidence is assigned a value of 0.

7. A driving reference line calculation system, applying the driving reference line calculation method according to any one of claims 1 to 6, characterized in that: include: A data acquisition module, configured to acquire and output lane line data of a currently traveling vehicle according to a preset calculation cycle; a data judgment module, configured to judge the existence and reliability of the current left and right lane lines based on the lane line data, and output the judgment result on the existence and reliability of the lane lines; a data prediction module, configured to predict the lane line data and output or store a predicted value of the lane line data in a current calculation cycle; a filtering module, configured to selectively filter the lane line data according to the judgment result and the predicted value of the lane line data, and output a selective filtering result; A data processing module is used to calculate the lane line width according to the selective filtering result, and calculate and output the current driving reference line data based on the lane line width.

8. The driving reference line calculation system according to claim 7, characterized in that: The system also includes a baseline confidence calculation module connected to the data prediction module, which is used to obtain the cumulative prediction time of the current calculation cycle, and calculate the current baseline confidence based on the cumulative prediction time and output it to the external active safety system.

9. A vehicle equipped with an automatic driving system, characterized in that: The automatic driving system is connected to the driving reference line calculation system as described in claim 7 or 8, and is used to control the driving state of the vehicle according to the current driving reference line data output by the driving reference line calculation system.

10. A computer device comprising a memory and a processor, wherein the memory stores computer-readable instructions, wherein: When the computer-readable instructions are executed by the processor, the processor performs the steps of the driving reference line calculation method according to any one of claims 1 to 6.

11. A storage medium having computer-readable instructions stored thereon, characterized in that: When the computer-readable instructions are executed by a processor, the steps of the driving reference line calculation method according to any one of claims 1 to 6 are implemented.

Images