Method for detecting ground geometry continuity and system for a vehicle

Abstract

The invention provides a method for detecting road geometrical shape continuity, and the method comprises the steps: continuously detecting a road geometrical shape F (t) determined by a carrying tool in the driving process of the carrying tool, the geometrical shape detected at the moment of ti is Fi (ti), and i is greater than or equal to 0 and is an integer; judging whether the geometrical shape F (t) meets a continuity condition or not according to a judgment strategy; and if the judgment result shows that the geometrical shape F (t) does not meet the continuity condition, outputting an abnormal prompt. The invention also provides a method for monitoring the driving condition of the vehicle, a system for the vehicle and a storage medium.

Description

Technical Field

[0001] The present application relates to technologies related to autonomous driving, and more specifically, to technologies for detecting the continuity of the geometric shape of roads on which vehicles are traveling. Background Art

[0002] With the development of sensor technology, control technology and automobile intelligence, autonomous driving technology, including assisted driving, has also made great progress. For autonomous driving (including assisted driving), the correct prediction of the driving road is an important consideration for driving safety and has always been the focus of the industry. Summary of the invention

[0003] According to one aspect of the present application, a method for detecting the continuity of road geometry is provided. The method for detecting the continuity of road geometry includes: during the driving process of a vehicle, continuously detecting the road geometry F(t) determined by the vehicle, wherein at t i The geometry of the moment detection is F i (t i ), i ≥ 0 and is an integer; judging whether the geometric shape F(t) meets the continuity condition according to the judgment strategy; if the judgment result shows that the geometric shape F(t) does not meet the continuity condition, triggering a safety response.

[0004] According to some examples of the present application, a method for detecting the continuity of road geometry is provided, for example, the judgment strategy is to judge whether the absolute value of the difference between the geometric shapes F(t) at adjacent detection moments meets the first continuity condition; and wherein the continuous detection of the road geometry F(t) determined by the vehicle is to detect the road geometry F(t) every time length ΔT. Further, for example, the method also includes: if the absolute value of the difference between the geometric shapes F(t) at the adjacent detection moments does not meet the first continuity condition, and the situation of not meeting the first continuity condition occurs more than m times in a number of detection moments that are sequentially consecutive to any moment in the adjacent detection moments, then the safe driving mechanism is triggered, m≥1 and is an integer; wherein the several moments that are sequentially consecutive to any moment in the adjacent detection moments include the detection moment before any moment and / or the detection moment after any moment.

[0005] According to some examples of the present application, a method for detecting the continuity of road geometry is provided. For example, the continuous detection of the road geometry F(t) determined by the vehicle includes: during the driving process of the vehicle, based on the road geometry F(t) determined by the vehicle from t i Time to t i +ΔT1 time, determine the shape function F1(ΔT1) of the road under the first shape function F1(t), and the iThe time from moment ti+ΔT1 to moment ti+ΔT1 is the first time period ΔT1; in the second time period ΔT2 of the vehicle's travel, the second shape function F2(ΔT2) of the road under the second shape function F2(t) is determined based on the detection; wherein the start time of the second time period is moment ti+ΔT1. Further, illustratively, judging whether the geometric shape F(t) meets the second continuity condition according to the judgment strategy includes: predicting the first shape function F1(ΔT2) during the second time period ΔT2 based on the first shape function F1(t); comparing the second shape function F2(ΔT2) with the first shape function F1(ΔT2) to obtain the absolute value of the difference between the two shape functions at corresponding moments in the second time period ΔT2; comparing the maximum absolute value of the absolute values ​​with the second continuity condition.

[0006] According to another aspect of the present application, a method for monitoring the driving condition of a vehicle is also provided, the method comprising monitoring the driving condition of the vehicle by executing any one of the methods described above.

[0007] According to another aspect of the present application, an autonomous driving method is also provided, which includes any one of the methods for detecting the continuity of road geometry described herein.

[0008] According to another aspect of the present application, a system for a vehicle is also provided, the system comprising: a memory for storing program instructions; a processor for executing the program instructions and implementing any one of the above-described methods when executing the program instructions.

[0009] According to another aspect of the present application, a non-transitory readable storage medium is also provided, which stores program instructions configured to be executable by a vehicle, and when the program instructions are executed, the vehicle can implement any of the methods described above. BRIEF DESCRIPTION OF THE DRAWINGS

[0010] The following will describe the embodiments of the present application in detail with reference to the accompanying drawings so that the present application can be more fully understood, wherein:

[0011] Figure 1 is a flow chart of a method for detecting continuity of road geometry according to an example of the present application;

[0012] Figure 2 is a flow chart of a method for detecting road geometry continuity according to a first example of the present application;

[0013] Figure 3 is a flow chart of a method for detecting the continuity of the curvature radius of a lane line according to an example of the present application;

[0014] Figure 4 is a flow chart of a method for detecting road geometry continuity according to a second example of the present application;

[0015] Figure 5 is a schematic diagram of the vehicle in the first coordinate system based on the first shape function F1(t) during driving;

[0016] Figure 6 is a schematic diagram of the vehicle in the second coordinate system based on the second shape function F2(t) during driving;

[0017] Figure 7 yes Figure 5 and Figure 6 The curve of the second time period ΔT2 shown is a schematic diagram of the curve in the same coordinate system;

[0018] Figure 8 is a flowchart of a method for detecting continuity of road geometry according to some other examples of the present application; and

[0019] Fig. 9 It is a schematic diagram of the structure of a system for a vehicle according to an example of the present application. DETAILED DESCRIPTION

[0020] In order to make the purpose, technical scheme and advantages of the embodiments of the present application clearer, the following will be combined with the accompanying drawings to clearly and completely describe the implementation methods of the present application. It should be noted that the described implementation methods are only part of the implementation methods of the technical solution of the present application, not all. All other implementation methods obtained by ordinary technicians in this field based on the implementation methods recorded in the application documents without paying creative work are covered by the protection scope of the present application.

[0021] The terms "first", "second", etc. in the specification, claims and drawings of this application are used to distinguish similar objects, rather than to describe a specific order or sequence. It should be understood that the data used in this way can be interchangeable where appropriate, so that the embodiments of the invention described herein can be implemented in an order other than those illustrated or described herein. In addition, "first" and "second" are used for descriptive purposes only and are not to be understood as expressly or implicitly indicating relative importance or implicitly indicating the number of technical features indicated. Thus, the features defined as "first" and "second" may explicitly or implicitly include one or more of the features, and in the description of the embodiments of the present application, unless otherwise stated, "multiple" means two or more.

[0022] As used hereinafter, the words “example” or “exemplarily” or “exemplarily” mean “serving as an example, embodiment, or illustration.” Any embodiment described as “example” or “exemplarily” or “exemplarily” is not necessarily to be construed as preferred or advantageous over other embodiments.

[0023] Each example of the present application may be implemented in a vehicle capable of performing autonomous driving. Autonomous driving includes assisted driving and autonomous driving. For example, each example of the present application may be implemented in a vehicle including an advanced driver assistance system (ADAS), preferably in a vehicle including an ADAS system above L2. By way of illustration and not limitation, the ADAS system will be directly used as an example for explanation below.

[0024] In each example of the present application, a vehicle is used as an example of a vehicle, but the vehicle is not limited to a vehicle. It is also possible to implement each example of the present application in other driving tools with assisted driving and automatic driving. By way of example and not limitation, the present application uses the geometry of the lane centerline as an example of the road geometry.

[0025] Figure 1 1 is a flow chart of a method for detecting road geometry continuity according to an example of the present application. When the road geometry detection method according to each example of the present application is applied to an ADAS system, it can be implemented by the ADAS system, or by other systems or devices capable of communicating with the ADAS.

[0026] Reference Figure 1 In step S100, during the driving process of the vehicle, the road geometry F(t) determined by the vehicle is continuously detected, wherein at t i The geometry of the moment detection is F i (t i ), i≥0 and is an integer. In step S102, it is determined whether the geometric shape F(t) meets the continuity condition according to the judgment strategy. If the judgment result shows that the geometric shape F(t) does not meet the continuity condition, a safety response is triggered, as shown in step S104. If the judgment result shows that the geometric shape F(t) meets the continuity condition, step S102 can be continued. The safety response includes outputting an abnormal prompt so that the user can understand the abnormality of the vehicle's road prediction. In some cases, the safety response may also include triggering a safe driving mechanism, which is intended to change the automatic driving to manual driving to ensure driving safety and / or take other measures to make the vehicle safer.

[0027] The method may be performed by an autonomous driving system of the vehicle, such as an ADAS system, or by other electronic systems or components of the vehicle (such as a controller such as the vehicle's ECU), or by a separate electronic device that can communicate with or be coupled to the vehicle.

[0028] Specifically, during the vehicle's driving process, the vehicle continuously detects the geometric shape of the road ahead F(t),t i The geometry of the moment detection is F i (t i). Continuous detection, for example, is detection once every certain period of time. For the road geometry that has been determined, it is determined whether the geometry meets the continuity condition based on the judgment strategy. In theory, during vehicle driving, for example, the geometry of the road ahead calculated by its automatic driving system according to the model should be continuous without jump values, so if discontinuity is detected, it means that the current prediction of the road may be abnormal. According to the present application, the judgment strategy includes multiple implementation methods, which will be described in conjunction with specific examples below. When the judgment result shows that the determined geometry does not meet the continuity condition, a safety response is triggered.

[0029] Figure 2 1 is a flow chart of a method for detecting the continuity of road geometry according to the first example of the present application, wherein the road geometry is a lane line geometry. In this example, the judgment strategy is to judge whether the absolute value of the difference between the geometry F(t) at adjacent detection moments meets the first continuity condition.

[0030] Reference Figure 2 In step S200, during the driving process of the vehicle, the geometric shape F(t) of the road on which the vehicle is driving is detected at intervals of ΔT. i The geometry of the moment detection is F i (t i ), at t i-1 The geometry of the moment detection is F i-1 (t i-1 ), ΔT is the time between adjacent detection times, for example, t i With t i-1 The time difference is i ≥ 0 and is an integer. It can be understood that t i The time can be any detection time, and the previous detections are t i-1 ,t i-2 At the same time, the subsequent tests are t i+1 ,t i+2 The geometric shape of the lane centerline F(t) at the detection time t i The geometric shape of F i (t i ), it should be understood that F i (t i ) is t i The value of the geometry at the time, and the previous detection time t i-1 ,t i-2 The corresponding geometric shape is F i-1 (t i-1 ), F i-2 (t i-2 ) etc., and at the subsequent detection time t i+1 ,t i+2The geometric shapes corresponding to the same moments are F i+1 (t i+1 ), F i+2 (t i+2 ) etc., the difference in geometric shape between the two detection moments is ΔF.

[0031] Road geometry is the geometry of the road determined by the vehicle (e.g., by an ADAS system) while the vehicle is traveling. Data for determining road geometry may come from images of the road acquired by an on-board camera. In the case where a high-precision map is available, data related to road geometry may also be obtained from the high-precision map. The on-board camera may be an image acquisition device such as a camera fixed to the vehicle, or an image acquisition device that can communicate with the vehicle but is not fixed to the vehicle, or an electronic device with a camera such as a mobile phone, a tablet such as an iPAD, etc. that can communicate with the vehicle, or a combination thereof.

[0032] According to the method of the example of the present application, the geometric shape F(t) of the road is detected every ΔT time, so that the continuous geometric shape of a series of moments is obtained. It should be noted that the geometric shape is represented as F(t) here to illustrate the correspondence between the road geometry determined by the vehicle and the detection time, and it does not limit the road geometry to be calculated based on time, or to be a function model that changes with time. For example, in some of the examples below, the model of the geometric shape is the radius of curvature, F(t) is the radius of curvature R, and the radius of curvature R is the same at each detection time t i The value is expressed as R(t i ), but it can be understood that R is not calculated based on time t.

[0033] In step S202, it is determined whether the absolute value of the difference between the geometric shapes at adjacent detection moments meets the first continuity condition, for example, i (t i ) and F i-n (t i-n ) meets the first continuity condition. The first continuity condition is a pre-set condition that can indicate that the determined road geometry presents continuous changes. For example, the difference in geometry at two detection moments should be within a certain range. If it exceeds this range, it indicates that an abnormal value has occurred.

[0034] For example, t i With t i-1 The corresponding geometric shape F i (t i ) and F i-1 (t i-1 ) is F i-1 (t i-1 ) and F i (t i) and T i The difference of n ΔT is the geometry of the engraving and t i The difference in geometry at the moment is F i-1 (t i-n ) and F i (t i In this example, n≥0 and is an integer.

[0035] If the absolute value of the difference in geometry between adjacent detection moments, such as F i (t i ) and F i-n (t i-n ) meets the first continuity condition, then the detected geometric shape F(t) is continuous, and the road geometric shape detected every ΔT time period is continued to be judged according to the judgment strategy.

[0036] If the absolute value of the difference in geometry between adjacent detection moments, F i (t i ) and F i-n (t i-n ) does not meet the first continuity condition, then the detected change in the geometric shape shows discontinuous characteristics. Here, triggering a safety response may be issuing an abnormal prompt, as shown in step S204.

[0037] According to some examples of the present application, step S205 is further included. When the absolute value of the difference between the geometric shapes F(t) at adjacent detection moments does not meet the first continuity condition, it is further determined whether several detection moments that are sequentially consecutive to any of the adjacent detection moments do not meet the first continuity condition for more than m times, where m≥1 and is an integer. If the judgment result is that the first continuity condition is not met for more than m times, triggering a safety reaction may be triggering a safe driving mechanism, which is intended to allow the driver to take over the automatic driving to ensure driving safety. In addition, triggering the safe driving mechanism may also include issuing a warning in a visually visible manner, an audibly perceptible manner, or a combination of the two. Among them, several moments that are sequentially consecutive to any of the adjacent detection moments include the detection moment before any moment and / or the detection moment after any moment. For example, F i (t i ) and F i-n (t i-n ) does not meet the first continuity condition, then it is judged that i Time or t i-n Whether the situation that does not meet the first continuity condition occurs more than m times in a number of consecutive moments, if so, the safe driving mechanism is triggered. i Time or t i-nIf there is no situation that does not meet the first continuity condition within a number of consecutive moments, the safe driving mechanism will not be triggered, and only an abnormal prompt will be output.

[0038] For a vehicle traveling on a road, the automatic driving system determines the geometric shape of the road ahead according to its algorithm and controls the vehicle accordingly. Compared with the existing automatic driving system that only determines the geometric shape of the road ahead without judging abnormal conditions of the geometric shape determined, the vehicle executing the disclosed method can judge whether the road geometric shape determined by the vehicle is continuous by continuously detecting the geometric shape at different times, and take corresponding measures in time when discontinuity occurs to ensure driving safety.

[0039] According to some examples of the present application, the road geometry F(t) determined by the vehicle includes: the road geometry determined by the vehicle based on the image of the road collected during the driving process, which is referred to as the real road geometry in this disclosure; and / or the road geometry determined by the vehicle based on the high-precision map used during the driving process, which is referred to as the map road geometry in this disclosure. The road geometry determined by the vehicle based on the image of the road collected during the driving process, that is, the real road geometry, is, for example, the road geometry determined by the ADAS system based on the image taken by the image acquisition device.

[0040] According to an example of the present application, in the case where the road geometry determined by the vehicle includes the real road geometry and the map road geometry, the above is combined with Figure 1 and Figure 2 The described continuity check process uses real road geometry. However, in cases where the real road geometry is not available, it is possible to obtain the road geometry from a high-precision map.

[0041] According to some examples of the present application, when the road geometry F(t) determined by the vehicle includes both the real road geometry and the map road geometry, according to Figure 2 The method shown may further include step S220, i The real road geometry at time t iCompare the map road geometry at the moment. If the absolute value of the difference between the two is greater than the comparison threshold, enter step S222 and issue a prompt to indicate that the current real road geometry is different from the map road geometry. Otherwise, continue to step S220. For the same road, the geometry of the real road and the map road should theoretically be the same. Therefore, considering the fault tolerance mechanism, if the difference between the two is greater than the comparison threshold, it means that one of them is abnormal. It can be understood that the smaller the comparison threshold may be, the more accurate the comparison or judgment. In this way, two-way monitoring of the high-precision map and the real road geometry determined based on the image of the road taken can be achieved, which provides the possibility of further improving the safety of autonomous driving and improving map accuracy.

[0042] According to various examples of the present application, the geometry of the road may be a lane geometry. In some specific examples, the lane geometry is modeled as a lane radius of curvature or a lane curvature, wherein the lane radius of curvature and the lane curvature are reciprocals of each other.

[0043] The following will briefly describe the method according to the example of this application using lane line curvature radius and lane line curvature respectively.

[0044] Figure 3 is a flow chart of a method for detecting the continuity of the curvature radius of a lane line according to an example of the present application. Figure 3 As shown, in step S300, during the driving process of the vehicle, the curvature radius R of the lane line determined by the vehicle is detected every ΔT time to obtain t i The radius R of the moment i , that is, F i (t i ) is R i In step S302, it is determined whether the absolute value of the difference between the curvature radii at adjacent detection moments meets the first continuity condition. If the judgment result is that the first continuity condition is not met, a safety reaction is triggered, as shown in step S304. In the case of meeting the first continuity condition, continue with step S302.

[0045] As an example, determining whether the absolute value of the difference between the radii of curvature at adjacent detection moments meets the first continuity condition may include different judgment methods. For example, the radii of curvature at adjacent moments are subtracted to determine the absolute value of their difference. For another example, the radius of curvature at each moment in the adjacent moments is compared with the radius of curvature at a previous moment, such as with the radius of curvature R0 at time t0. The latter is described in detail as an example.

[0046] The change of the radius of curvature R in the longitudinal distance, that is, the travel distance S in the direction of vehicle travel, complies with the limitation of inequality (1):

[0047] dR / dS≤h (1)

[0048] Where h is the gradient.

[0049] In some specific examples, the value of the radius of curvature R may be limited to a certain range, such as a radius of curvature less than 1800 meters. Under this restriction, the change of the radius of curvature over time complies with inequality (2):

[0050] ΔR≌dR / dS(2)

[0051] When the vehicle speed v = dS / dt, inequality (3) is derived by combining inequality (2):

[0052] (dS / dt)×ΔT≤h×v×ΔT (3)

[0053] In practical applications, the vehicle speed v is a constant value within a short time interval. If the vehicle speed is not constant within a short time interval, the maximum value of the vehicle speed can be used as the value of the speed v in formula (3).

[0054] Take time t0, time t1 and time t2 as examples. The curvature radius corresponding to time t0 is R0, the curvature radius corresponding to time t1 is R1, the curvature radius corresponding to time t2 is R2, and t i The radius of curvature corresponding to the moment is R i , t1 is t0+1×ΔT, t2 is t0+2×ΔT, t i It is t0+i×ΔT. As mentioned above, ΔT is the duration between adjacent detection moments.

[0055] If the following situations occur, it is considered that the curvature radius predicted by the automatic driving does not meet the first continuity condition, and a discontinuity occurs, and an abnormal prompt is issued:

[0056] |R1–R0|>h×v×ΔT, or

[0057] |R2–R0|>2×h×v×ΔT or

[0058] |R i –R0|>i×h×v×ΔT

[0059] When the absolute value of the difference between R1 and R0 does not meet the first continuity condition, that is, |R1–R0|>h×v×ΔT, further determine the R value at i consecutive moments after time T1. i Whether the first continuity condition is satisfied, that is, whether |R i –R0|≤i×h×v×ΔT, if yes, then R1 can be regarded as an outlier that only needs to issue a fault prompt. It should be understood that the value of i×h×v×ΔT needs to be within the fault tolerance interval FTTI. If not, the safe driving mechanism is triggered. For example:

[0060] |R1–R0|>h×v×ΔT, and

[0061] |R2–R0|>2×h×v×ΔT

[0062] In this example, m=1 (see Figure 2 ). If m=3, the condition for triggering the safe driving mechanism is:

[0063] |R1–R0|>h×v×ΔT, and

[0064] |R2–R0|>2×h×v×ΔT and

[0065] |R3–R0|>3×h×v×ΔT

[0066] In this case, the fault tolerance interval is relatively large, and in some cases, it may be larger, especially when a certain error in the detection of the curvature radius and the vehicle speed is allowed. This may lose some accuracy, but the application may be more extensive.

[0067] As for the gradient h, it can be obtained through the map, positioning system, etc. used by the vehicle. In some cases, the gradient h can adopt a local limit value. For example, for the position boundary (bounds) that can be obtained from the map, in the area of ​​±2km, for example, the maximum gradient dR / dS is less than the global gradient limit value h g , limited to h l , then the h l Used as the gradient h used in the above calculation (instead of h g In other cases, the global gradient limit value h g Used as the gradient h.

[0068] In an alternative embodiment, the curvature K can be used as the lane line geometry. In this example, dK / dS=q is the gradient. Figure 3 Compared with the process shown in FIG. 1 , only the gradient value is different. Therefore, the process of this example will not be described in detail, and only a specific example will be used for explanation. This specific example uses time t0, time t1, and time t2 as the detection time. The curvature corresponding to time t0 is K0, the curvature corresponding to time t1 is K1, the curvature corresponding to time t2 is K2, and t i The curvature corresponding to the moment is K i , t1 is t0+1×ΔT, t2 is t0+2×ΔT, t i = t0+i×ΔT. If the following situations occur, it is considered that the curvature determined by the autonomous driving system is discontinuous and an abnormal prompt is issued.

[0069] |K1–K0|>g×v×ΔT, or

[0070] |K2–K0|>2×g×v×ΔT or

[0071] |K i –K0|>i×g×v×ΔT

[0072] When the absolute value of the difference between K1 and K0 does not meet the first continuity condition, further determine the K values ​​of i consecutive moments after time T1. i Whether the first continuity condition is satisfied, that is, whether |K i –K0|≤i×g×v×ΔT, if yes, then K1 can be regarded as an abnormal value that only needs to issue a fault prompt. If not, the safe driving mechanism is triggered.

[0073] Figure 4 1 is a flow chart of a method for detecting the continuity of road geometry according to the second example of the present application, in which the road geometry is modeled as a shape function, while the lane line geometry is still used as an example of the road geometry. As is known, the shape function can be constructed as a cubic polynomial or a shape point array, which can be constructed as needed, as long as the shape function can be used to express the lane line geometry.

[0074] In step S400, during the vehicle driving process, based on the i Time to t i +ΔT1 time, determine the shape function F1(ΔT1) of the lane line under the first shape function F1(t), and i The first period ΔT1 is from time t to time ti+ΔT1. i The vehicle started driving at time ΔT1 and the distance traveled during this time period can be expressed as ΔS.

[0075] In step S402, during the second period ΔT2 of vehicle travel, a second shape function F2(ΔT2) of the road under the second shape function F2(t) is determined based on real-time detection. The second period ΔT2 is a period of ΔT2 starting from the time ti+ΔT1.

[0076] In step S404 , based on the first shape function F1 (t), the first shape function F1 ( ΔT2 ) during the second period ΔT2 is predicted.

[0077] In step S406, the second shape function F2 (ΔT2) is compared with the first shape function F1 (ΔT2) to obtain the absolute value of the difference between the two shape functions at corresponding moments in the second time period ΔT2.

[0078] In step S408, the maximum absolute value among the absolute values ​​is compared with the second continuity condition to determine whether it meets the second continuity condition. If the comparison result is that the maximum absolute value does not meet the second continuity condition, a safety reaction is triggered, as shown in step S410. If the second continuity condition is met, continue with step S408. According to some examples of the present application, the first shape function F1(t) and the second shape function F2(t) are functions in different coordinate systems. The first shape function F1(t) is a function in a first coordinate system, and the second shape function F2(t) is a function in a second coordinate system. By way of example and not limitation, the first coordinate system and the second coordinate system may be, for example, one of a vehicle coordinate system and a world coordinate system.

[0079] Figure 5 is a schematic diagram of the vehicle in the first coordinate system based on the first shape function F1(t) during driving. Figure 5 As shown, from point x a1 To point x b1 The length represents the vehicle from time t i Travel to time t i +ΔT1 distance from point x a1 To point x b1 The corresponding function value, that is, y, forms a curve F1(ΔT1), which is shown as a first period curve 500. The second period ΔT2 is predicted according to the first shape function F1(t), that is, from t i +ΔT1 time to t i At the moment of +ΔT1+ΔT2, the shape of the lane line should be curve F1(ΔT2), which is from point x on the x-axis. b12 To point x e1 , indicated by the dotted line as the second time period curve 502.

[0080] Figure 6 is a schematic diagram of the vehicle in the second coordinate system based on the second shape function F2(t) during driving. Figure 6 As shown, from point x b2 To point x e2 The length represents the vehicle from time t i +ΔT1 Travel to time t i +ΔT1+ΔT2, the distance from point x b2 To point x e2 The corresponding function value, that is, y, forms a curve F2 (ΔT2), which is illustrated as a curve 602 of the second time period ΔT2 in the second coordinate system.

[0081] Figure 7 yes Figure 5 and Figure 6The curve 502 and the curve 602 of the second time period ΔT2 shown in FIG. 5 are schematic diagrams in the same coordinate system. In the example of the present application, the first shape function in the first coordinate system is converted to the second coordinate system. Figure 7 As shown, the converted curve 502 is represented by a dotted curve 702. Theoretically, for the vehicle, during the driving process in the second period, the F2 (ΔT2) determined in real time and the F1 (ΔT2) predicted according to the first shape function should coincide after being converted to the same coordinates, which makes it feasible to determine whether the prediction of the lane shape by the automatic driving system is abnormal by judging whether the difference of the function values ​​at the same time is within the second continuous condition.

[0082] It should be noted that in combination Figure 4 In the second example described, the x-axis in the coordinate system represents the vehicle's travel distance, but this is only for reference. For example, the vehicle may have traveled 40 meters, but the representation on the x-axis may be Figure 5 Midpoint x a1 To point x b1 Length.

[0083] In addition, in the second example, the second continuity condition is to compare the absolute value of the difference between the first shape function and the second shape function at the corresponding time (which can also be regarded as the gradient B of the two functions) with a preset threshold. Specifically, after comparing the second shape function F2 (ΔT2) with the first shape function F1 (ΔT2) to obtain the absolute value of the difference between the two shape functions at the corresponding time in the second period ΔT2, in step S408 (see Figure 4 ), take the maximum absolute value among the calculated absolute values, such as Figure 7 The value indicated by the double arrow 700 in the example is compared with the preset threshold value. If the maximum absolute value is greater than the preset threshold value, it indicates that the second continuity condition is not satisfied, that is, the comparison result does not meet the second continuity condition.

[0084] The preset threshold can be set according to the actual application scenario, etc. The preset threshold can be determined by considering both safety and availability, or by striking a balance between them. For example, if the true knowledge measurement value at time t is known, the absolute value of the difference between F1(ΔT2) and the true knowledge measurement value g0 represents the measurement error of F1(ΔT2), where g0 is the value of the true knowledge function. The error depends on the interval [x a , x e ], taking the above example as a reference, then for example, depending on the interval [x a1 , x e1 ]. For a given shape function, the error needs to be bounded and should satisfy the bound, and the exact same bound can be used as the preset threshold. In addition, considering the inaccuracy of different coordinate system conversions or the measurement of the vehicle's travel distance, it may be necessary to appropriately expand the preset threshold in actual use.

[0085] For the second example, if the shape function and the coordinate system are well selected, it can be considered to calculate only the end of the second time period ΔT2, that is, close to x e ( Figure 5 Indicated as x e1 , Figure 6 Indicated as x e2 ) to reduce the amount of calculation. In theory, the point with the largest absolute value appears near x e Place.

[0086] Regarding the conversion from one coordinate axis to another, the functions in the two coordinate systems can be aligned to the same coordinate system in different ways, as long as the shape comparison in the same time period required by this application can be achieved. The following is a simple example:

[0087] As is known, when each shape is described as a function in its own coordinate system, the functions in different coordinates can be converted to the same coordinate system by coordinate system transformation such as translation, and alignment can be achieved simply.

[0088] According to some other examples of the present application, it also includes, for example Figure 8 The method shown is, for example, Figure 1 Another specific example of the method shown, in which the judgment strategy is combined with the above Figure 2 and Figure 4 The examples described vary.

[0089] In this example, during the vehicle driving process, based on the i Time to t i +ΔT1 time detection, determine the lane line shape function F(ΔT1) under the shape function F(t), from t i The time from time ti+ΔT1 is the first time period ΔT1, as shown in step S800. In the first time period ΔT1, the vehicle travels, for example, 30 meters. In step S802, the shape function F(ΔT1) in the first time period ΔT1 is used to predict the earlier time t i-n The function value F(t) under the shape function F(t) i-n ). In step S803, the time t is determined from the determined actual lane line geometry function. i-n In step S804, F(t)_real and F(t i-n) are projected into the same coordinate system, and it is determined whether the absolute value of the difference between the two in the coordinate system is within the allowable error. If the absolute value of the difference is not within the allowable error, a safety reaction is triggered, as shown in step S806. If the absolute value of the difference is not within the allowable error for multiple times at consecutive detection moments, triggering a safety reaction may include triggering a safe driving mechanism.

[0090] The present application also provides a method for monitoring the driving condition of a vehicle. The monitoring method monitors the driving condition of a vehicle by executing the method for detecting the continuity of road geometry described in any of the examples above. By executing the method, during the automatic driving process, it is possible to find possible calculation errors of the vehicle, such as the vehicle, on the road geometry ahead, thereby avoiding possible accidents or failures.

[0091] The present application also provides an autonomous driving method, which includes the method for detecting the continuity of road geometry described in any one of the above examples.

[0092] Fig. 9 1 is a schematic diagram of a system for a vehicle according to an example of the present application. The system includes a memory 90 and a processor 92. The memory stores instructions, and the processor 92 is configured to execute the program instructions to implement the method for detecting the continuity of road geometry described in any of the examples above.

[0093] The system can be implemented in an existing system in the vehicle, such as an autonomous driving system, for example, through software or a combination of software and hardware; or can be implemented as an independent system that can be configured to communicate with the vehicle's autonomous driving system.

[0094] The present application also provides a non-transitory readable storage medium storing program instructions configured to be executable by a vehicle. When the program instructions are executed, the vehicle can implement the method for detecting the continuity of road geometry as described in any of the examples above.

[0095] By executing the method for detecting the continuity of road geometry described in this application, or adopting the system according to the example of this application, when the vehicle implements autonomous driving such as ADAS, the vehicle can promptly detect abnormalities in the road determination by the vehicle by continuously detecting the road shape, such as the lane line geometry, determined by the vehicle's autonomous driving system, thereby triggering a safety response and effectively improving the safety of autonomous driving.

[0096] Although various embodiments of the present application have been shown and described in detail to illustrate the principles of the present application, it should be understood that the various embodiments and their features can be combined with each other without conflict to achieve new implementations, and such implementations should be included in the scope claimed by the present application. In addition, the present application can also be implemented in other ways that do not deviate from the spirit of the present application.

Claims

1. A method for detecting the continuity of road geometry, characterized in that: The method comprises: During the vehicle's travel, the road geometry F(t) determined by the vehicle is continuously detected, wherein at t i The geometry of the moment detection is F i (t i ), i ≥ 0 and is an integer; Determining whether the geometric shape F(t) meets the continuity condition according to the judgment strategy; If the determination result indicates that the geometric shape F(t) does not meet the continuity condition, a safety response is triggered.

2. The method according to claim 1, characterized in that The judgment strategy is to judge whether the absolute value of the difference between the geometric shapes F(t) at adjacent detection moments meets the first continuity condition; and The continuously detecting the road geometry F(t) determined by the vehicle is to detect the road geometry F(t) at intervals of a time period ΔT.

3. The method according to claim 2, characterized in that The method further comprises: If the absolute value of the difference between the geometric shapes F(t) at the adjacent detection moments does not meet the first continuity condition, and the situation that the first continuity condition is not met occurs more than m times at a number of detection moments that are consecutive to any of the adjacent detection moments, then the safe driving mechanism is triggered, where m ≥ 1 and is an integer; The plurality of detection moments that are sequentially continuous with any of the adjacent detection moments include the detection moment before and / or the detection moment after any of the adjacent detection moments.

4. The method according to claim 2, characterized in that: The vehicle-determined road geometry F(t) includes: The actual road geometry is the road geometry determined by the vehicle based on the collected images of the road during driving; and / or Map road geometry is the road geometry determined by the vehicle based on the high-precision map used during its driving process.

5. The method according to claim 4, characterized in that When the vehicle determined road geometry F(t) includes both real road geometry and map road geometry, the continuously detecting the vehicle determined road geometry F(t) is detecting the real road geometry.

6. The method according to claim 5, characterized in that When the road geometry F(t) determined by the vehicle includes both real road geometry and map road geometry, the method further comprises: t i The geometry of the real road at time t i At this moment, the geometric shapes of the road on the map are compared. If the absolute value of the difference between the two is greater than a comparison threshold, a prompt is issued to indicate that the geometric shape of the current real road is different from the geometric shape of the road on the map.

7. The method according to any one of claims 2 to 6, characterized in that: The road geometry is the geometry of lane lines.

8. The method according to claim 7, characterized in that The geometry of the lane line is modeled as the curvature or the radius of curvature of the lane line.

9. The method according to claim 1, characterized in that: The continuously detecting the road geometry F(t) determined by the vehicle comprises: During the driving process of the vehicle, based on the i Time to t i +ΔT1 time, determine the shape function F1(ΔT1) of the road under the first shape function F1(t), and the i From time ti+ΔT1 is the first period ΔT1; During a second period ΔT2 of travel of the vehicle, determining a second shape function F2(ΔT2) of the road under the second shape function F2(t) based on detection; The start time of the second time period is time ti+ΔT1.

10. The method according to claim 9, characterized in that The step of judging whether the geometric shape F(t) meets the second continuity condition according to the judging strategy includes: Based on the first shape function F1(t), predicting the first shape function F1(ΔT2) during the second period ΔT2; Compare the second shape function F2 (ΔT2) with the first shape function F1 (ΔT2) to obtain the absolute value of the difference between the two shape functions at corresponding moments in the second time period ΔT2; comparing the maximum absolute value among the absolute values ​​with a second continuity condition; and If the judgment result indicates that the geometric shape F(t) does not meet the second continuity condition, triggering a safety response includes: If the comparison result of the maximum absolute value among the absolute values ​​with the second continuity condition does not meet the second continuity condition, a safety reaction is triggered.

11. The method according to claim 10, characterized in that The first shape function F1(t) is a function in a first coordinate system, and the second shape function F1(t) is a function in a second coordinate system, and the first coordinate system is different from the second coordinate system; as well as The comparing the second shape function F2 (ΔT2) with the first shape function F1 (ΔT2) to obtain the absolute value of the difference between the two shape functions at corresponding moments in the second time period ΔT2 includes: Converting the second shape function F2 (ΔT2) and the first shape function F1 (ΔT2) to the same coordinate system; The second shape function F2 (ΔT2) converted to the same coordinate system is compared with the first shape function F1 (ΔT2) to obtain the absolute value of the difference between the two converted shape functions at corresponding moments in the second time period ΔT2.

12. The method according to claim 11, characterized in that The first coordinate system and the second coordinate system are respectively one of a vehicle coordinate system and a world coordinate system.

13. The method according to any one of claims 9 to 12, characterized in that The lane line geometry is the lane line geometry.

14. The method according to claim 1, characterized in that The continuously detecting the road geometry F(t) determined by the vehicle comprises: During the driving process of the vehicle, based on the i Time to t i +ΔT1 time, determine the shape function F(ΔT1) of the road under the shape function F(t), and the i The time from time ti+ΔT1 is the first time period ΔT1.

15. The method according to claim 14, characterized in that The step of judging whether the geometric shape F(t) meets the continuity condition according to the judging strategy includes: Based on the shape function F(ΔT1), predict the earlier time t i-n The function value F(t) under the shape function F(t) i-n ); Determine t from the determined actual lane geometry function F(t) i-n The actual shape function value F( i-n )_real; F( i-n )_real and F(t i-n ) are projected into the same coordinate system and the absolute value of the difference between the two in the coordinate system is compared to see whether it is within the permissible error; and If the judgment result shows that the geometric shape F(t) does not meet the continuity condition, an abnormal prompt is issued, including: If F( i-n )_real and F(t i-n ) are projected into the same coordinate system, and the absolute value of the difference between the two in the coordinate system is not within the allowable error, a safety response is triggered.

16. A method for monitoring the driving condition of a vehicle, characterized in that: The method comprises monitoring the driving condition of the vehicle by executing the method described in any one of claims 1 to 15.

17. An automatic driving method, characterized in that: The method comprises the method according to any one of claims 1 to 15.

18. A system for a vehicle, characterized in that: The system comprises: A memory for storing program instructions; A processor, configured to execute the program instructions, and implement the method according to any one of claims 1 to 15 when executing the program instructions.

19. The system according to claim 18, characterized in that The system is implemented in the vehicle and is configured to communicate with the autonomous driving system; or the system is implemented in the autonomous driving system.

20. A non-transitory readable storage medium storing program instructions configured to be executable by a vehicle, wherein when the program instructions are executed, the vehicle can implement the method according to any one of claims 1 to 15.

Images