Road boundary estimation method and device, vehicle and storage medium

Abstract

The invention relates to a road boundary estimation method and device, a vehicle and a storage medium. The method comprises the following steps: in an automatic driving process, taking an identified road edge as an initial road boundary; judging whether the initial road boundary has an abnormal condition or not; if the initial road boundary has the abnormal condition, acquiring traffic flow information and road surface arrow information around the vehicle; wherein a drivable area formed by the initial road boundary comprises a forward lane; estimating a target road boundary of the forward lane according to the traffic flow information and / or the road arrow information; and updating the initial road boundary by adopting the target road boundary. According to the scheme provided by the invention, the limitation that only single geometric information is used for estimating the value in the related technology is broken through, and the estimation accuracy is greatly improved, so that the situation that the vehicle enters the physical boundary of the reverse lane or the forward lane by mistake in automatic driving is avoided, and the collision risk is greatly reduced.

Description

Technical Field

[0001] The present application relates to the field of autonomous driving technology, and in particular to a road boundary estimation method, device, vehicle and storage medium. Background Art

[0002] Road boundaries are important constraints for autonomous driving. For example, in an environment where traffic rules are constrained, vehicles are prohibited from driving beyond the road boundaries of the drivable area during autonomous driving to avoid the vehicle deviating from the forward lane. Therefore, road boundaries play a vital role in autonomous driving.

[0003] The related technology mainly relies on lane lines or physical boundaries. It is a method that only relies on geometric elements. However, the actual road environment is complex, the vehicle's perception ability is insufficient, there is a certain degree of occlusion of geometric elements, there may be illegal driving when the user is teaching, and lane lines are missing on special sections. Therefore, relying only on geometric elements can easily lead to inaccurate estimation of road boundaries, making it easy for the vehicle to deviate from the forward lane during autonomous driving, causing the vehicle to mistakenly drive into the reverse lane or drive towards the physical boundary of the forward lane, which is prone to collision risks. Summary of the invention

[0004] In order to solve or partially solve the problems existing in the related art, the present application provides a road boundary estimation method, device, vehicle and storage medium, which breaks the limitation of the related art of only using single geometric information for valuation, greatly improves the valuation accuracy, thereby avoiding the vehicle from mistakenly driving into the opposite lane or driving towards the physical boundary of the forward lane during autonomous driving, and greatly reduces the risk of collision.

[0005] A first aspect of the present application provides a road boundary estimation method, comprising:

[0006] During the autonomous driving process, the identified road edge is used as the initial road boundary;

[0007] Determining whether there is an abnormality in the initial road boundary;

[0008] If there is an abnormality in the initial road boundary, the traffic flow information and road arrow information around the vehicle are obtained; wherein the drivable area formed by the initial road boundary includes the forward lane;

[0009] estimating a target road boundary of the forward lane according to the traffic flow information and / or the road arrow information;

[0010] The target road boundary is adopted to update the initial road boundary.

[0011] In one embodiment, the traffic flow information includes direction information, position information and speed information of each surrounding vehicle, and the road arrow information includes direction information and position information of each road arrow; and estimating the target road boundary of the forward lane according to the traffic flow information and / or the road arrow information includes:

[0012] Based on the orientation information and speed information of each of the surrounding vehicles, the plurality of surrounding vehicles are classified to obtain reverse vehicles and forward vehicles, and based on the orientation information of each of the road arrows, the plurality of road arrows are classified to obtain reverse arrows and forward arrows; wherein the driving direction of the reverse vehicle is opposite to the driving direction of the vehicle, the driving direction of the forward vehicle is the same as the driving direction of the vehicle, the direction indicated by the reverse arrow is opposite to the driving direction of the vehicle, and the direction indicated by the forward arrow is the same as the driving direction of the vehicle;

[0013] The target road boundary of the forward lane is estimated based on the position information and speed information of the reverse vehicle, and / or the position information and speed information of the forward vehicle, and / or the position information of the reverse arrow.

[0014] In one embodiment, the initial road boundary includes a left initial road boundary, and the target road boundary includes a left target road boundary; and estimating the target road boundary of the forward lane according to the position information and speed information of the reverse vehicle, and / or the position information and speed information of the forward vehicle, and / or the position information of the reverse arrow, includes:

[0015] When it is determined that the left initial road boundary has the abnormal situation, constructing a first reference boundary line according to the position information and speed information of the oncoming vehicle, constructing a second reference boundary line according to the position information and speed information of the forward vehicle, and constructing a third reference boundary line according to the position information of the reverse arrow;

[0016] Calculating target distances between the target driving trajectory of the vehicle and the first reference boundary line, the second reference boundary line, and the third reference boundary line respectively;

[0017] According to the preset priority and the target distance, selecting a target reference boundary line from the first reference boundary line, the second reference boundary line and the third reference boundary line as the left target road boundary of the forward lane;

[0018] The adopting the target road boundary to update the initial road boundary includes:

[0019] The left initial road boundary is updated using the left target road boundary.

[0020] In another embodiment, the initial road boundary includes a right initial road boundary, and the target road boundary includes a right target road boundary; and estimating the target road boundary of the forward lane according to the position information and speed information of the reverse vehicle, and / or the position information and speed information of the forward vehicle, and / or the position information of the reverse arrow, includes:

[0021] When it is determined that the right initial road boundary has the abnormal situation, constructing a fourth reference boundary line according to the position information and speed information of the forward vehicle;

[0022] Using the fourth reference boundary line as the right target road boundary of the forward lane;

[0023] The adopting the target road boundary to update the initial road boundary includes:

[0024] The right initial road boundary is updated using the right target road boundary.

[0025] In one embodiment, when it is determined that the left initial road boundary has the abnormal situation, a first reference boundary line is constructed according to the position information and speed information of the oncoming vehicle, a second reference boundary line is constructed according to the position information and speed information of the forward vehicle, and a third reference boundary line is constructed according to the position information of the reverse arrow, including:

[0026] When it is determined that the left initial road boundary has the abnormal situation, determining a first driving trajectory of the oncoming vehicle in the oncoming lane according to the position information and speed information of the oncoming vehicle, and determining a second driving trajectory of the oncoming vehicle in the oncoming lane according to the position information and speed information of the oncoming vehicle;

[0027] Based on the orientation information of the reverse vehicle, the first driving trajectory is moved to the left by a first preset distance to construct a first reference boundary line; and, based on the orientation information of the forward vehicle, the second driving trajectory is moved to the left by a second preset distance to construct a second reference boundary line; and, based on the orientation information of the reverse arrow, a third reference boundary line is constructed to the left at a third preset distance from the position information of the reverse arrow.

[0028] In one embodiment, when it is determined that the right initial road boundary has the abnormal situation, constructing a fourth reference boundary line according to the position information and speed information of the forward vehicle includes:

[0029] When it is determined that the abnormal situation exists at the right initial road boundary, determining a second driving trajectory of the forward vehicle in the forward lane according to the position information and speed information of the forward vehicle;

[0030] Based on the orientation information of the forward vehicle, the second driving track is moved to the right by a second preset distance to construct a fourth reference boundary line.

[0031] In one implementation, the initial road boundary includes a left initial road boundary and a right initial road boundary; and determining whether there is an abnormality in the initial road boundary includes:

[0032] Calculating a road width and a road width change of a drivable area formed by the left initial road boundary and the right initial road boundary, and counting the number of boundaries close to the left initial road boundary of the vehicle;

[0033] Determining whether the road width is greater than a first threshold, determining whether the road width change is greater than a second threshold, and determining whether the boundary quantity is greater than a third threshold;

[0034] If the road width is greater than a first threshold or the road width change is greater than a second threshold or the number of boundaries is greater than a third threshold, it is determined that an abnormality exists in the initial road boundary; wherein the abnormality includes any one of the following: the drivable area contains a reverse lane, the left initial road boundary is incomplete, and the right initial road boundary is incomplete.

[0035] A second aspect of the present application provides a road boundary estimation device, comprising:

[0036] An initial road boundary acquisition module is used to use the identified road edge as the initial road boundary during the autonomous driving process;

[0037] An abnormality judgment module, used to judge whether there is an abnormality in the initial road boundary;

[0038] A reference information acquisition module, used for acquiring traffic information and road arrow information around the vehicle if there is an abnormality in the initial road boundary; wherein the drivable area formed by the initial road boundary includes the forward lane;

[0039] A target road boundary estimation module, used for estimating the target road boundary of the forward lane according to the traffic flow information and / or the road arrow information;

[0040] The initial road boundary updating module is used to update the initial road boundary by using the target road boundary.

[0041] A third aspect of the present application provides a vehicle, comprising:

[0042] Processor; and

[0043] The memory stores executable codes thereon, and when the executable codes are executed by the processor, the processor is caused to execute the method as described above.

[0044] A fourth aspect of the present application provides a computer-readable storage medium having executable code stored thereon. When the executable code is executed by a processor of a vehicle, the processor is caused to execute the method as described above.

[0045] The technical solution provided by this application may have the following beneficial effects:

[0046] The solution provided by the present application uses the identified road edge as the initial road boundary during the autonomous driving process; determines whether there are abnormalities in the initial road boundary; if there are abnormalities in the initial road boundary, obtains the traffic information and road arrow information around the vehicle; wherein the drivable area formed by the initial road boundary includes the forward lane; estimates the target road boundary of the forward lane based on the traffic information and / or road arrow information; and uses the target road boundary to update the initial road boundary. The present application further refers to multi-dimensional information such as traffic flow and road arrows for logical inference on the basis of using geometric information for valuation, thus breaking the limitation of the related technology of using only single geometric information for valuation, greatly improving the valuation accuracy, thereby avoiding the situation where the vehicle mistakenly drives into the reverse lane or drives towards the physical boundary of the forward lane during autonomous driving, and greatly reducing the risk of collision.

[0047] It should be understood that the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the present application. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] The above and other objects, features and advantages of the present application will become more apparent by describing in more detail the exemplary embodiments of the present application in conjunction with the accompanying drawings, wherein the same reference numerals generally represent the same components in the exemplary embodiments of the present application.

[0049] Figure 1 is a flowchart of a road boundary estimation method shown in an embodiment of the present application;

[0050] Figure 2 is a comparative schematic diagram of the drivable area before and after optimization shown in the embodiment of the present application;

[0051] Figure 3 is another flowchart of a road boundary estimation method shown in an embodiment of the present application;

[0052] FIG. 4A to FIG. 4Dis a schematic diagram showing various abnormal situations in the initial road boundary according to an embodiment of the present application;

[0053] Figure 5 is a schematic diagram of different types of traffic flows shown in an embodiment of the present application;

[0054] FIG. 6A to FIG. 6D is a flow chart of using an oncoming vehicle to assist in estimating the left target road boundary of a forward lane according to an embodiment of the present application;

[0055] Figure 7 and Figure 8 is a flow chart of using a forward vehicle to assist in estimating the left target road boundary of a forward lane according to an embodiment of the present application;

[0056] Fig. 9 is a flow chart of using a reverse arrow to assist in estimating the left target road boundary of a forward lane according to an embodiment of the present application;

[0057] Fig.10 is a flow chart of using a forward vehicle to assist in estimating the right target road boundary of a forward lane according to an embodiment of the present application;

[0058] Fig.11 is a schematic diagram of the structure of a road boundary estimation device shown in an embodiment of the present application;

[0059] Fig.12 It is a schematic diagram of the structure of a vehicle shown in an embodiment of the present application. DETAILED DESCRIPTION

[0060] The embodiments of the present application will be described in more detail below with reference to the accompanying drawings. Although the embodiments of the present application are shown in the accompanying drawings, it should be understood that the present application can be implemented in various forms and should not be limited by the embodiments described herein. On the contrary, these embodiments are provided to make the present application more thorough and complete, and to fully convey the scope of the present application to those skilled in the art.

[0061] The terms used in this application are for the purpose of describing specific embodiments only and are not intended to limit this application. The singular forms of "a", "said" and "the" used in this application and the appended claims are also intended to include plural forms unless the context clearly indicates other meanings. It should also be understood that the term "and / or" used herein refers to and includes any or all possible combinations of one or more associated listed items.

[0062] It should be understood that although the terms "first", "second", "third", etc. may be used in this application to describe various information, this information should not be limited to these terms. These terms are only used to distinguish the same type of information from each other. For example, without departing from the scope of this application, the first information may also be referred to as the second information, and similarly, the second information may also be referred to as the first information. Thus, the features defined as "first" and "second" may explicitly or implicitly include one or more of the features. In the description of this application, the meaning of "multiple" is two or more, unless otherwise clearly and specifically defined.

[0063] The related technology mainly relies on lane lines or physical boundaries. It is a method that only relies on geometric elements. However, the actual road environment is complex, the vehicle's perception ability is insufficient, there is a certain degree of occlusion of geometric elements, there may be illegal driving when the user is teaching, and lane lines are missing on special sections. Therefore, relying only on geometric elements can easily lead to inaccurate estimation of road boundaries, making it easy for the vehicle to deviate from the forward lane during autonomous driving, causing the vehicle to mistakenly drive into the reverse lane or drive towards the physical boundary of the forward lane, which is prone to collision risks.

[0064] In response to the above problems, an embodiment of the present application provides a road boundary estimation method. On the basis of utilizing geometric information for valuation, it further refers to multi-dimensional information such as traffic flow and road arrows for logical inference, thereby breaking the limitation of related technologies that only use single geometric information for valuation, greatly improving the valuation accuracy, thereby avoiding the situation where the vehicle accidentally drives into the opposite lane or drives towards the physical boundary of the forward lane during autonomous driving, and greatly reducing the risk of collision.

[0065] The technical solution of the embodiments of the present application is described in detail below with reference to the accompanying drawings.

[0066] Figure 1 It is a flowchart of a road boundary estimation method shown in an embodiment of the present application.

[0067] See also Figure 1 , the road boundary estimation method of the present application comprises:

[0068] S110, during the autonomous driving process, using the identified road edge as an initial road boundary.

[0069] The execution subject of the road boundary estimation may be an automatic driving system mounted on the vehicle. In the process of controlling the automatic driving of the vehicle, the automatic driving system may obtain the road environment data sensed by the camera and the radar respectively, and identify the road edges on both sides of the vehicle through the road environment data. The road edges may include at least one of lane lines, shoulders, guardrails, fences, curbstones, traffic cones, and crash buckets. The automatic driving system may use the identified road edges as the initial road boundaries.

[0070] S120, determining whether there is any abnormality in the initial road boundary.

[0071] Due to the complexity of the actual road environment, insufficient vehicle perception, certain occlusion of geometric elements, possible illegal driving during user teaching, missing lane lines on special sections, and other issues, the initial road boundary may have abnormal conditions. For example, the drivable area formed by the initial road boundary may include the reverse lane, which will cause the vehicle to mistakenly drive into the reverse lane, which may easily lead to a collision risk. Another example is that the initial road boundary may be incomplete, that is, the initial road boundary may have missing parts, which will cause the vehicle to mistakenly drive into the reverse lane or cause the vehicle to drive towards the physical boundary of the forward lane (such as a curb), which may easily lead to a collision risk.

[0072] Therefore, after obtaining the initial road boundary, the autonomous driving system can first determine whether there are any abnormalities in the initial road boundary to avoid directly using the abnormal initial road boundary which may easily cause a collision accident.

[0073] S130: If there is an abnormality in the initial road boundary, traffic information and road arrow information around the vehicle are obtained; wherein the drivable area formed by the initial road boundary includes the forward lane.

[0074] The initial road boundary may include a left initial road boundary and a right initial road boundary.

[0075] In one example, the drivable area formed by the initial road boundary under normal circumstances only includes the forward lane and does not include the reverse lane. However, the actual road environment is complex, the vehicle perception ability is insufficient, there is a certain degree of occlusion of geometric elements, there may be illegal driving during user teaching, and lane lines are missing in special sections. There may be abnormal conditions on the left initial road boundary, such as Figure 2 As shown, the left initial road boundary belongs to the reverse lane, not the forward lane, so that the drivable area formed by the left initial road boundary and the right initial road boundary includes the reverse lane, causing the vehicle to mistakenly drive into the reverse lane, which is prone to collision risk. In this regard, the embodiment of the present application can optimize the left initial road boundary to obtain the left target road boundary, so that the drivable area formed by the left target road boundary and the right initial road boundary only includes the forward lane, not the reverse lane.

[0076] In another example, the left initial road boundary is complete and intact under normal circumstances. However, due to the complex road environment, insufficient vehicle perception, certain occlusion of geometric elements, illegal driving during user teaching, and missing lane lines on special sections, the left initial road boundary may have missing parts, causing the vehicle to easily mistakenly drive into the opposite lane, thereby easily causing a collision risk. In this regard, the embodiment of the present application can optimize the left initial road boundary to obtain a complete and intact left target road boundary.

[0077] In another example, the right initial road boundary is intact under normal circumstances. However, due to the complex road environment, insufficient vehicle perception, certain occlusion of geometric elements, illegal driving during user teaching, and missing lane lines on special sections, the right initial road boundary may have missing parts, causing the vehicle to easily drive to the right physical boundary of the forward lane (such as the right curb), which is prone to collision risks. In this regard, the embodiment of the present application can optimize the right initial road boundary to obtain a complete and intact right target road boundary.

[0078] If any of the above abnormal conditions exist at the initial road boundary, the autonomous driving system can obtain the traffic information and road arrow information sensed by the camera and radar respectively.

[0079] It should be noted that no matter whether there are any abnormalities in the initial road boundary, the drivable area formed by the initial road boundary includes the forward lane.

[0080] S140: Estimate a target road boundary of a forward lane according to the traffic flow information and / or road arrow information.

[0081] Traffic flow information is used to characterize the traffic flow formed by the collection of surrounding vehicles in the vehicle's automatic driving, wherein the traffic flow types may include forward traffic flow, reverse traffic flow, and temporarily parked traffic flow. The forward traffic flow is the traffic flow formed by the surrounding vehicles traveling in the forward lane, and the reverse traffic flow is the traffic flow formed by the surrounding vehicles traveling in the reverse lane. The automatic driving system may use the forward traffic flow to assist in estimating the left target road boundary and the right target road boundary of the forward lane. In addition, the automatic driving system may also use the reverse traffic flow to assist in estimating the left target road boundary and the right target road boundary of the reverse lane. Generally, the reverse lane is on the left side of the forward lane, so the right target road boundary of the reverse lane is the left target road boundary of the forward lane, thereby assisting in estimating the left target road boundary of the forward lane.

[0082] Road arrow information is a traffic element that describes the direction of the road, including a forward arrow and a reverse arrow. The lane indicated by the forward arrow is the forward lane, and the lane indicated by the reverse arrow is the reverse lane. Since the forward arrow cannot effectively estimate the left target road boundary and the right target road boundary of the forward lane, the embodiment of the present application can only use the reverse arrow for logical judgment. For example, the automatic driving system can use the reverse arrow to assist in estimating the left target road boundary and the right target road boundary of the reverse lane. Generally, the reverse lane is on the left side of the forward lane, so the right target road boundary of the reverse lane is the left target road boundary of the forward lane, thereby assisting in estimating the left target road boundary of the forward lane.

[0083] S150: Using the target road boundary, the initial road boundary is updated.

[0084] The initial road boundary may include a left initial road boundary and a right initial road boundary, and the target road boundary may include a left target road boundary and a right target road boundary. After obtaining the left target road boundary, the automatic driving system may use the left target road boundary to update the left initial road boundary, so that the drivable area formed by the updated left initial road boundary and the right initial road boundary only includes the forward lane and does not include the reverse lane, or the updated left initial road boundary is complete and intact, thereby avoiding the situation where the vehicle mistakenly drives into the reverse lane during automatic driving, greatly reducing the risk of collision. After obtaining the right target road boundary, the automatic driving system may use the right target road boundary to update the right initial road boundary, so that the updated right initial road boundary is complete and intact, thereby avoiding the situation where the vehicle drives towards the physical boundary of the forward lane during automatic driving, greatly reducing the risk of collision.

[0085] As can be seen from this example, the solution provided by this application uses the identified road edge as the initial road boundary during the autonomous driving process; determines whether there are abnormalities in the initial road boundary; if there are abnormalities in the initial road boundary, obtains the traffic information and road arrow information around the vehicle; wherein the drivable area formed by the initial road boundary includes the forward lane; estimates the target road boundary of the forward lane based on the traffic information and / or road arrow information; and uses the target road boundary to update the initial road boundary. This application further refers to multi-dimensional information such as traffic flow and road arrows for logical inference on the basis of using geometric information valuation, thus breaking the limitation of the related technology of using only single geometric information valuation, greatly improving the valuation accuracy, thereby avoiding the situation where the vehicle mistakenly drives into the reverse lane or drives towards the physical boundary of the forward lane during autonomous driving, and greatly reducing the risk of collision.

[0086] Figure 3is another flowchart of the road boundary estimation method shown in the present application.

[0087] See also Figure 3 , the road boundary estimation method of the present application comprises:

[0088] S310, during the autonomous driving process, using the identified road edge as an initial road boundary.

[0089] For this step, please refer to the description in S110 and will not be repeated here.

[0090] S320, determining whether there is any abnormality in the initial road boundary.

[0091] For this step, please refer to the description in S120 and will not be repeated here.

[0092] In one implementation, determining whether an initial road boundary is abnormal may include:

[0093] Calculate the road width and road width change of the drivable area formed by the left initial road boundary and the right initial road boundary, and count the number of boundaries of the left initial road boundary close to the vehicle; determine whether the road width is greater than a first threshold, determine whether the road width change is greater than a second threshold, and determine whether the number of boundaries is greater than a third threshold; if the road width is greater than the first threshold, or the road width change is greater than the second threshold, or the number of boundaries is greater than the third threshold, determine that an abnormality exists in the initial road boundary; wherein the abnormality includes any one of the following: the drivable area contains a reverse lane, the left initial road boundary is incomplete, and the right initial road boundary is incomplete.

[0094] The left initial road boundary and the right initial road boundary form a drivable area. The autonomous driving system can calculate the distance between the left initial road boundary and the right initial road boundary, which can be used as the road width of the drivable area. Then the autonomous driving system can calculate the difference between the widths of two adjacent roads, which can be used as the change in road width of the drivable area. In addition, the autonomous driving system can count the number of boundaries close to the left initial road boundary of the vehicle. When the vehicle is driving in the forward lane, the left initial road boundary close to the vehicle belongs to the forward lane; when the vehicle mistakenly drives into the reverse lane, the left initial road boundary close to the vehicle belongs to the reverse lane.

[0095] The automatic driving system can determine whether there is an abnormality in the initial road boundary by determining whether the road width is greater than a first threshold, determining whether the road width change is greater than a second threshold, and determining whether the number of boundaries is greater than a third threshold. The first threshold and the second threshold can be set based on the road width of the forward lane, and the third threshold can be set based on the number of road boundaries of the forward lane.

[0096] In one example, if Figure 4A As shown, if the road width of the drivable area is greater than the first threshold, it means that the left initial road boundary belongs to the opposite lane and not the forward lane, so that the drivable area formed by the left initial road boundary and the right initial road boundary includes the opposite lane. Therefore, it can be determined that there is an abnormality in the left initial road boundary.

[0097] In another example, Figure 4B As shown, the left initial road boundary 1 belongs to the reverse lane, and the left initial road boundary 2 belongs to the forward lane, wherein the road width of the drivable area formed by the left initial road boundary 1 and the right initial road boundary is a, and the road width of the drivable area formed by the left initial road boundary 2 and the right initial road boundary is b. The road width change ab calculated based on the road width a and the road width b is greater than the second threshold value, indicating that the left initial road boundary 2 of the forward lane has a missing part, so it can be determined that there is an abnormality in the left initial road boundary.

[0098] In another example, Figure 4C As shown, the road width c of the drivable area formed by the left initial road boundary and the right initial road boundary, the road width d of the drivable area formed by the left initial road boundary and the curbstone, and the road width change cd calculated based on the road width c and the road width d is greater than the second threshold, indicating that there is a missing part in the right initial road boundary of the forward lane, so it can be determined that there is an abnormality in the right initial road boundary.

[0099] In another example, the autonomous driving includes memory driving. During the user teaching process, the autonomous driving system can learn the environmental perception data (such as lane lines, surrounding vehicles, etc.) and driving data (such as its own driving trajectory) during the vehicle's driving process. However, the user may violate driving rules during the teaching (such as detouring and overtaking), causing the vehicle to enter the opposite lane, which in turn causes the memory driving route learned by the autonomous driving system to be abnormal. Therefore, when the subsequent vehicle drives along this memory driving route, the vehicle will mistakenly drive into the opposite lane, such as Figure 4D As shown, the vehicle first travels in the forward lane and then switches to the reverse lane. Therefore, the left initial road boundary close to the vehicle includes the left initial road boundary 1 and the left initial road boundary 2, wherein the left initial road boundary 1 belongs to the reverse lane and the left initial road boundary 2 belongs to the forward lane, so that the drivable area formed by the left initial road boundary 1, the left initial road boundary 2 and the right initial road boundary includes the reverse lane. Therefore, it can be determined that there is an abnormality in the left initial road boundary.

[0100] In summary, the abnormal situation may include any one of the following: the drivable area includes an oncoming lane, the left initial road boundary is incomplete, and the right initial road boundary is incomplete.

[0101] S330, if there is an abnormality in the initial road boundary, the traffic flow information and road arrow information around the vehicle are obtained; wherein the drivable area formed by the initial road boundary includes the forward lane, the traffic flow information includes the direction information, position information and speed information of each surrounding vehicle, and the road arrow information includes the direction information and position information of each road arrow.

[0102] If any of the above abnormal conditions exist at the initial road boundary, the autonomous driving system can obtain the traffic information and road arrow information sensed by the camera and radar respectively.

[0103] It should be noted that no matter whether there are any abnormalities in the initial road boundary, the drivable area formed by the initial road boundary includes the forward lane.

[0104] The traffic flow information may include the direction information, position information and speed information of each surrounding vehicle, wherein the direction information of the surrounding vehicles is used to characterize the driving direction of the surrounding vehicles at each moment, the position information of the surrounding vehicles is used to characterize the driving position of the surrounding vehicles at each moment, and the speed information of the surrounding vehicles is used to characterize the driving speed of the surrounding vehicles at each moment.

[0105] The road arrow information may include orientation information and position information of each road arrow. The orientation information of the road arrow is used to represent the direction indicated by the road arrow, and the position information of the road arrow is used to represent the position of the road arrow in the identified lane.

[0106] S340, based on the direction information and speed information of each surrounding vehicle, multiple surrounding vehicles are classified to obtain reverse vehicles and forward vehicles, and based on the direction information of each road arrow, multiple road arrows are classified to obtain reverse arrows and forward arrows; wherein, the driving direction of the reverse vehicle is opposite to the driving direction of the vehicle, the driving direction of the forward vehicle is the same as the driving direction of the vehicle, the direction indicated by the reverse arrow is opposite to the driving direction of the vehicle, and the direction indicated by the forward arrow is the same as the driving direction of the vehicle.

[0107] The autonomous driving system can classify multiple surrounding vehicles based on the orientation information and speed information of each surrounding vehicle. In a specific implementation, the autonomous driving system can classify surrounding vehicles whose driving speed is not zero and whose driving direction is the same as that of the vehicle as forward vehicles, and classify surrounding vehicles whose driving speed is not zero and whose driving direction is opposite to that of the vehicle as reverse vehicles, and classify surrounding vehicles whose driving speed is zero as temporarily stopped vehicles. Among them, multiple forward vehicles can form a forward traffic flow, multiple reverse vehicles can form a reverse traffic flow, and multiple temporarily stopped vehicles can form a temporarily stopped traffic flow. In order to facilitate distinction, the embodiments of the present application can use different colors to identify different types of traffic flows, such as Figure 5 As shown, forward traffic uses blue signs, reverse traffic uses red signs, and temporarily stopped traffic uses yellow signs.

[0108] The autonomous driving system can classify multiple road arrows based on the orientation information of each road arrow. In a specific implementation, if the direction indicated by a road arrow is the same as the vehicle's travel direction, the autonomous driving system can classify the road arrow as a forward arrow, and the lane marked by the forward arrow is the forward lane; if the direction indicated by a road arrow is opposite to the vehicle's travel direction, the autonomous driving system can classify the road arrow as a reverse arrow, and the lane marked by the reverse arrow is the reverse lane.

[0109] S350, estimating a target road boundary of the forward lane based on the position information and speed information of the reverse vehicle, and / or the position information and speed information of the forward vehicle, and / or the position information of the reverse arrow.

[0110] As an example, the autonomous driving system can assist in estimating the left target road boundary and the right target road boundary of the opposite lane based on the position information and speed information of the opposite vehicle. Generally, the opposite lane is on the left side of the forward lane, so the right target road boundary of the opposite lane is the left target road boundary of the forward lane, thereby assisting in estimating the left target road boundary of the forward lane.

[0111] As another example, the autonomous driving system may assist in estimating the left and right target road boundaries of the forward lane based on the position information and speed information of the forward vehicle.

[0112] As another example, the autonomous driving system can assist in estimating the left target road boundary and the right target road boundary of the reverse lane based on the position information of the reverse arrow. Generally, the reverse lane is on the left side of the forward lane, so the right target road boundary of the reverse lane is the left target road boundary of the forward lane, thereby assisting in estimating the left target road boundary of the forward lane.

[0113] In one embodiment, estimating the target road boundary of the forward lane according to the position information and speed information of the reverse vehicle, and / or the position information and speed information of the forward vehicle, and / or the position information of the reverse arrow may include:

[0114] When it is determined that there is an abnormality in the left initial road boundary, a first reference boundary line is constructed according to the position information and speed information of the oncoming vehicle, a second reference boundary line is constructed according to the position information and speed information of the forward vehicle, and a third reference boundary line is constructed according to the position information of the reverse arrow; the target distances between the target driving trajectory of the vehicle and the first reference boundary line, the second reference boundary line and the third reference boundary line are calculated respectively; according to the preset priority and target distance, the target reference boundary line is selected from the first reference boundary line, the second reference boundary line and the third reference boundary line as the left target road boundary of the forward lane.

[0115] The abnormality in the initial road boundary may be that there is an abnormality in the left initial road boundary or there is an abnormality in the right initial road boundary. Therefore, the autonomous driving system can first determine on which side the initial road boundary has the abnormality.

[0116] In a specific implementation, the road width of the drivable area includes the left road width and the right road width, wherein the left road width refers to the distance between the left initial road boundary and the target driving trajectory of the vehicle, and the right road width refers to the distance between the right initial road boundary and the target driving trajectory of the vehicle. Therefore, the automatic driving system can specifically determine whether the left road width is greater than the first threshold, and specifically determine whether the right road width is greater than the first threshold. If the left road width is greater than the first threshold, it is determined that there is an abnormality in the left initial road boundary, such as Figure 4A If the right road width is greater than the first threshold, it is determined that an abnormality exists at the right initial road boundary.

[0117] Similarly, the road width change in the drivable area includes the left road width change and the right road width change. Therefore, the automatic driving system can specifically determine whether the left road width change is greater than the second threshold, and specifically determine whether the right road width change is greater than the second threshold. If the left road width change is greater than the second threshold, it is determined that there is an abnormality in the left initial road boundary, such as Figure 4B If the width change of the right road is greater than the second threshold, it is determined that there is an abnormality in the right initial road boundary, such as Figure 4C shown.

[0118] Similarly, if the number of boundaries of the left initial road boundary close to the vehicle is greater than the third threshold, it is determined that there is an abnormality in the left initial road boundary, such as Figure 4DAs shown. Since the right side of the forward lane is a physical boundary (such as a curb), the number of boundaries on the right initial road side close to the vehicle is usually less than the third threshold. Therefore, this condition is only used to determine whether there is an abnormality in the left initial road boundary, and cannot be used to determine whether there is an abnormality in the right initial road boundary.

[0119] When it is determined that there is an abnormality in the left initial road boundary through any of the above methods, since the lanes where the oncoming vehicle, the forward vehicle and the reverse arrow are respectively located are all related to the left target road boundary of the forward lane, the automatic driving system can combine the oncoming vehicle, the forward vehicle and the reverse arrow to assist in estimating the left target road boundary of the forward lane.

[0120] In a specific implementation, there may be multiple vehicles going in the opposite direction. To ensure the driving safety of the vehicle, the automatic driving system can select the opposite vehicle with the smallest distance from the multiple opposite vehicles based on the distance between the position information of each opposite vehicle and the target driving trajectory of the vehicle. The opposite vehicle is closest to the vehicle. Therefore, the automatic driving system can construct a first reference boundary line based on the position information and speed information of the opposite vehicle. In addition, there may also be multiple vehicles going in the forward direction. To ensure the driving safety of the vehicle, the automatic driving system can select the opposite vehicle with the smallest distance from the multiple forward vehicles based on the distance between the position information of each forward vehicle and the target driving trajectory of the vehicle. The forward vehicle is closest to the vehicle. Therefore, the automatic driving system can construct a second reference boundary line based on the position information and speed information of the forward vehicle. In addition, the reverse arrow is a traffic element of the reverse lane. The embodiment of the present application can expand its scope of action according to the position of the reverse arrow. Therefore, the automatic driving system can construct a third reference boundary line based on the position information of the reverse arrow.

[0121] After obtaining the first reference boundary line, the second reference boundary line and the third reference boundary line respectively, in order to evaluate which reference boundary line is more accurate and safer, the automatic driving system can calculate the target distances between the vehicle's target driving trajectory and the first reference boundary line, the second reference boundary line and the third reference boundary line respectively, so as to select the target reference boundary line from the first reference boundary line, the second reference boundary line and the third reference boundary line as the left target road boundary of the forward lane according to the preset priority and target distance. Specifically, since the oncoming vehicle and the reverse arrow are both located in the reverse lane and the forward vehicle is located in the forward lane, the second reference boundary line constructed based on the forward vehicle may be within the range of the reverse lane, and the first reference boundary line constructed based on the oncoming vehicle and the third reference boundary line constructed based on the reverse arrow are generally the dividing line between the reverse lane and the forward lane or are within the range of the forward lane. In order to avoid a greater risk of collision caused by the vehicle mistakenly driving into the reverse lane, the preset priority can be set to: first reference boundary line = third reference boundary line > second reference boundary line. Therefore, the automatic driving system first selects the first reference boundary line and the third reference boundary line according to the preset priority, and then selects the target reference boundary line from the first reference boundary line and the third reference boundary line according to the target distance. Specifically, if the target distance corresponding to the first reference boundary line is smaller than the target distance corresponding to the third reference boundary line, it means that the first reference boundary line is closest to the vehicle, and therefore the first reference boundary line is determined as the target reference boundary line, so that the target reference boundary line (i.e., the first reference boundary line) can be used as the left target road boundary of the forward lane, thereby minimizing the possibility that the vehicle mistakenly drives into the reverse lane; if the target distance corresponding to the third reference boundary line is smaller than the target distance corresponding to the first reference boundary line, it means that the third reference boundary line is closest to the vehicle, and therefore the third reference boundary line is determined as the target reference boundary line, so that the target reference boundary line (i.e., the third reference boundary line) can be used as the left target road boundary of the forward lane, thereby minimizing the possibility that the vehicle mistakenly drives into the reverse lane.

[0122] It should be noted that if there is no oncoming vehicle and reverse arrow on the current road section, the embodiment of the present application can directly determine the second reference boundary line as the target reference boundary line, so as to use the target reference boundary line (i.e., the second reference boundary line) as the left target road boundary of the forward lane, thereby avoiding the situation where the vehicle mistakenly drives into the reverse lane as much as possible.

[0123] In one embodiment, when it is determined that there is an abnormality in the left initial road boundary, constructing a first reference boundary line according to the position information and speed information of the reverse vehicle, constructing a second reference boundary line according to the position information and speed information of the forward vehicle, and constructing a third reference boundary line according to the position information of the reverse arrow may include:

[0124] When it is determined that there is an abnormality in the left initial road boundary, the first driving trajectory of the oncoming vehicle in the oncoming lane is determined according to the position information and speed information of the oncoming vehicle, and the second driving trajectory of the oncoming vehicle in the oncoming lane is determined according to the position information and speed information of the oncoming vehicle; based on the orientation information of the oncoming vehicle, the first driving trajectory is moved to the left by a first preset distance to construct a first reference boundary line; and, based on the orientation information of the oncoming vehicle, the second driving trajectory is moved to the left by a second preset distance to construct a second reference boundary line; and, based on the orientation information of the reverse arrow, a third reference boundary line is constructed to the left at a third preset distance from the position information of the reverse arrow.

[0125] Take the reverse vehicle as an example: FIG. 6A to FIG. 6D In the example, the left initial road boundary belongs to the reverse lane and does not belong to the forward lane. Therefore, it is determined that there is an abnormality in the left initial road boundary. In this regard, after obtaining different types of traffic, the embodiment of the present application can filter out the forward traffic and the temporarily stopped traffic, so as to extract the reverse traffic (such as Fig. 6A As shown), the reverse traffic flow is a traffic flow formed by multiple reverse vehicles. In order to ensure the driving safety of the vehicle, the embodiment of the present application can select the reverse vehicle A closest to the vehicle from the reverse traffic flow (such as Figure 6B As shown), the automatic driving system tracks the oncoming vehicle A: the oncoming vehicle A travels from the position of the first frame to the position of the second frame and then to the position of the third frame, and the first driving trajectory of the oncoming vehicle A in the oncoming lane can be determined. However, the perception ability of the vehicle may be insufficient, resulting in that the perceived position information of the oncoming vehicle A may not be accurate, and thus the first driving trajectory of the oncoming vehicle A determined based on the position information may not be accurate. In order to improve the accuracy, the automatic driving system can perform position matching in combination with the speed information of the oncoming vehicle A. For example, based on the speed of the oncoming vehicle A in the first frame, the position of the oncoming vehicle A in the second frame can be roughly predicted, and then the perceived position of the second frame is matched with the predicted position. If the perceived position of the second frame is consistent with the predicted position or the distance difference between the two is less than the fourth threshold, it is determined that the perceived position of the second frame matches the predicted position, which can indicate that the perceived position of the second frame is accurate, that is, the position information of the second frame is trustworthy. In this way, the first driving trajectory of the oncoming vehicle A finally determined is more accurate. After obtaining the first driving trajectory, the automatic driving system moves the first driving trajectory of the oncoming vehicle A to the left by a first preset distance based on the orientation information of the oncoming vehicle A, thereby constructing a first reference boundary line (such as Figure 6CAs shown). In the embodiment of the present application, the first preset distance can be set based on the body size of the oncoming vehicle. Specifically, the traffic information can also include the size information of each surrounding vehicle. The automatic driving system can extract the size information of the oncoming vehicle A from the traffic information. The size information of the oncoming vehicle A is used to characterize the body length, body width and body height of the oncoming vehicle A. The automatic driving system can set the first preset distance to be more than half the body width of the oncoming vehicle A. Exemplarily, assuming that the body width of the oncoming vehicle A is 1.8m, the first preset distance can be set to 0.9m. Therefore, the automatic driving system uses the orientation information of the oncoming vehicle A as a reference, moves the first driving trajectory of the oncoming vehicle A to the left by 0.9m, and obtains the first reference boundary line.

[0126] Take the forward vehicle as an example: Figure 7 In the figure, the left initial road boundary 2 of the forward lane has a missing part, so it is determined that there is an abnormality in the left initial road boundary. In this regard, after obtaining different types of traffic, the embodiment of the present application can filter out the reverse traffic and the temporarily parked traffic to extract the forward traffic. The forward traffic is a traffic flow formed by multiple forward vehicles. In order to ensure the driving safety of the vehicle, the embodiment of the present application can select the forward vehicle B (such as Figure 7As shown), the autonomous driving system can track the forward vehicle B: the autonomous driving system can determine the second driving trajectory of the forward vehicle B in the forward lane based on the position information of the forward vehicle B and its speed information, and then the autonomous driving system uses the orientation information of the forward vehicle B as a reference to move the second driving trajectory of the forward vehicle B to the left by a second preset distance, thereby constructing a second reference boundary line. Among them, the embodiment of the present application can set the second preset distance based on the body size of the forward vehicle B and the position of the forward vehicle B relative to the self-vehicle. Specifically, the autonomous driving system can extract the size information of the forward vehicle B from the traffic information. The size information of the forward vehicle B is used to characterize the body length, body width and body height of the forward vehicle B. The autonomous driving system can set the second preset distance to be more than half the body width of the forward vehicle B. As an example, assuming that the body width of the forward vehicle B is 1.8m, and the forward vehicle B is on the left side of the vehicle, the second preset distance can be set to half the body width of the forward vehicle B, that is, the second preset distance is 0.9m, so the automatic driving system uses the orientation information of the forward vehicle B as a reference, moves the second driving track of the forward vehicle B to the left by 0.9m, and obtains the second reference boundary line. As another example, assuming that the body width of the forward vehicle B is 1.8m, and the forward vehicle B is on the right side of the vehicle, the second preset distance can be set to one body width of the forward vehicle B, that is, the second preset distance is 1.8m, so the automatic driving system uses the orientation information of the forward vehicle B as a reference, moves the second driving track of the forward vehicle B to the left by 1.8m, and obtains the second reference boundary line.

[0127] Let's take the forward vehicle as an example again: Figure 8 In the case of illegal driving during user teaching, the subsequent vehicles will enter the opposite lane when driving according to the learned memory driving route (such as Figure 8 As shown in FIG. 1 , the left initial road boundary close to the vehicle includes the left initial road boundary 1 and the left initial road boundary 2. Therefore, it is determined that there is an abnormality in the left initial road boundary. For this reason, after obtaining different types of traffic flows, the embodiment of the present application can filter out the reverse traffic flow and the temporarily parked traffic flow, so as to extract the forward traffic flow, and then select the forward vehicle B closest to the vehicle from the forward traffic flow (such as Figure 8 As shown), the autonomous driving system can track the forward vehicle B: based on the position information of the forward vehicle B and its speed information, the autonomous driving system can determine the second driving trajectory of the forward vehicle B in the forward lane, and then the autonomous driving system uses the orientation information of the forward vehicle B as a reference to move the second driving trajectory of the forward vehicle B to the left by a second preset distance, thereby constructing a second reference boundary line of the upper and lower parts, and then connecting the second reference boundaries of the upper and lower parts to obtain the final second reference boundary line.

[0128] Take the reverse arrow as an example: Fig. 9 In the figure, the reverse arrow is a traffic element of the reverse lane, and the left initial road boundary should not exceed the reverse arrow. However, due to problems such as misdetection or missed detection of lane lines, the left initial road boundary exceeds the reverse arrow. This situation indicates that the left initial road boundary belongs to the reverse lane, not the forward lane. In this regard, after obtaining different types of road surface arrows, the embodiment of the present application can filter out the forward arrow to extract the reverse arrow, and then use the direction information of the reverse arrow as a reference to construct a third reference boundary line (such as the third reference boundary line) to the left that is a third preset distance away from the position information of the reverse arrow. Fig. 9 ). In this embodiment of the present application, the third preset distance can be set based on the position of the reverse arrow in the reverse lane. Specifically, the reverse arrow is generally located in the middle of the reverse lane, so the autonomous driving system can set the third preset distance to half the lane width of the reverse lane. Exemplarily, assuming that the lane width of the reverse lane is 3.5m, the third preset distance can be set to 1.75m. Therefore, the autonomous driving system uses the direction information of the reverse arrow as a reference and constructs a third reference boundary line 1.75m away from the position information of the reverse arrow to the left.

[0129] It should be noted that, in order to avoid using the information of different surrounding vehicles to jointly construct the same reference boundary line, the traffic information may also include the vehicle model information of each surrounding vehicle, so as to determine whether the position information and speed information at different times belong to the same surrounding vehicle based on the vehicle model information. Figure 6B As shown, the automatic driving system can determine whether the model information of the reverse vehicle in the first frame is consistent with the model information of the reverse vehicle in the second frame, and determine whether the model information of the reverse vehicle in the second frame is consistent with the model information of the reverse vehicle in the third frame. If they are consistent, it means that the reverse vehicles in the first frame, the second frame, and the third frame are the same, so the first reference boundary line is constructed using the information of the same reverse vehicle. Similarly, based on the model information of the forward vehicle, it can be determined that the second reference boundary line is constructed using the information of the same forward vehicle.

[0130] In another embodiment, estimating the target road boundary of the forward lane according to the position information and speed information of the reverse vehicle, and / or the position information and speed information of the forward vehicle, and / or the position information of the reverse arrow may include:

[0131] When it is determined that there is an abnormality in the right initial road boundary, a fourth reference boundary line is constructed according to the position information and speed information of the forward vehicle; and the fourth reference boundary line is used as the right target road boundary of the forward lane.

[0132] Since the oncoming vehicle and the reverse arrow are not associated with the right target road boundary of the forward lane, only the forward vehicle is associated with the right target road boundary of the forward lane, and the forward lane contains at least one sub-lane, that is, the number of sub-lanes of the forward lane is uncertain, in order to ensure that the vehicle has enough driving space, when it is determined that there is an abnormality in the right initial road boundary through any of the above methods, the embodiment of the present application can only use the oncoming vehicle to assist in estimating the right target road boundary of the forward lane.

[0133] In a specific implementation, there may be multiple forward vehicles. In order to ensure the driving safety of the vehicle, the automatic driving system can select the forward vehicle with the smallest distance from the multiple forward vehicles based on the distance between the position information of each forward vehicle and the target driving trajectory of the own vehicle. The forward vehicle is closest to the own vehicle. Therefore, the automatic driving system can construct a fourth reference boundary line based on the position information and speed information of the forward vehicle, so as to use the fourth reference boundary line as the right target road boundary of the forward lane, thereby minimizing the situation where the vehicle drives towards the physical boundary (such as curbstone) of the forward lane.

[0134] In one embodiment, when it is determined that there is an abnormality in the right initial road boundary, constructing a fourth reference boundary line according to the position information and speed information of the forward vehicle may include:

[0135] When it is determined that there is an abnormality in the right initial road boundary, the second driving trajectory of the forward vehicle in the forward lane is determined based on the position information and speed information of the forward vehicle; based on the orientation information of the forward vehicle, the second driving trajectory is moved to the right by a second preset distance to construct a fourth reference boundary line.

[0136] Take the forward vehicle as an example: Fig.10 In the example, the right initial road boundary of the forward lane has a missing part, so it is determined that there is an abnormality in the right initial road boundary. In this regard, after obtaining different types of traffic, the embodiment of the present application can filter out the reverse traffic and the temporarily parked traffic to extract the forward traffic. The forward traffic is formed by multiple forward vehicles. In order to ensure the driving safety of the vehicle, the embodiment of the present application can select the forward vehicle B (such as Fig.10As shown), the autonomous driving system can track the forward vehicle B: the autonomous driving system can determine the second driving trajectory of the forward vehicle B in the forward lane based on the position information of the forward vehicle B and its speed information, and then the autonomous driving system moves the second driving trajectory of the forward vehicle B to the right by a second preset distance based on the orientation information of the forward vehicle B, thereby constructing a second reference boundary line. Among them, the embodiment of the present application can set the second preset distance based on the body size of the forward vehicle B and the position of the forward vehicle B relative to the self-vehicle. Specifically, the autonomous driving system can extract the size information of the forward vehicle B from the traffic information. The size information of the forward vehicle B is used to characterize the body length, body width and body height of the forward vehicle B. The autonomous driving system can set the second preset distance to be more than half the body width of the forward vehicle B. As an example, assuming that the body width of the forward vehicle B is 1.8m, and the forward vehicle B is on the left side of the vehicle, the second preset distance can be set to one body width of the forward vehicle B, that is, the second preset distance is 1.8m, so the automatic driving system uses the orientation information of the forward vehicle B as a reference, moves the second driving track of the forward vehicle B to the right by 1.8m, and obtains the fourth reference boundary line. As another example, assuming that the body width of the forward vehicle B is 1.8m, and the forward vehicle B is on the right side of the vehicle, the second preset distance can be set to half the body width of the forward vehicle B, that is, the second preset distance is 0.9m, so the automatic driving system uses the orientation information of the forward vehicle B as a reference, moves the second driving track of the forward vehicle B to the right by 0.9m, and obtains the fourth reference boundary line.

[0137] It should be noted that the above-mentioned moving mode may specifically be a translation mode, which does not change the shape, size and direction of the driving track.

[0138] S360: Update the initial road boundary using the target road boundary.

[0139] For this step, please refer to the description in S150 and will not be repeated here.

[0140] In one embodiment, updating the initial road boundary using the target road boundary may include:

[0141] The left initial road boundary is updated using the left target road boundary.

[0142] When there is an abnormality in the left initial road boundary, the embodiment of the present application can use the reverse vehicle, the forward vehicle and the reverse arrow in combination to assist in estimating the left target road boundary of the forward lane, so as to use the left target road boundary to update the left initial road boundary. The updating method of the left initial road boundary can include any one of replacement, supplementation and correction.

[0143] As an example, Fig.6D As shown, since the left initial road boundary belongs to the reverse lane and not the forward lane, the automatic driving system can use the final estimated left target road boundary (i.e., the first reference boundary line) to replace the left initial road boundary, so that the drivable area formed by the left target road boundary and the right initial road boundary only includes the forward lane and does not include the reverse lane, thereby avoiding the vehicle from mistakenly driving into the reverse lane and greatly reducing the risk of collision.

[0144] As another example, Figure 7 As shown, since there is a missing part in the left initial road boundary 2 of the forward lane, the automatic driving system can use the final estimated left target road boundary (i.e., the second reference boundary line) to supplement the missing part of the left initial road boundary 2, so that the drivable area formed by the left target road boundary and the right initial road boundary only includes the forward lane and does not include the reverse lane, thereby avoiding the vehicle from mistakenly driving into the reverse lane and greatly reducing the risk of collision.

[0145] As another example, Figure 8 As shown in the figure, due to illegal driving during the user's teaching, the subsequent self-driving vehicle mistakenly entered the opposite lane while driving according to the learned memory driving route. Therefore, the automatic driving system can use the final estimated left target road boundary (i.e., the second reference boundary line) to correct the left initial road boundary. For example, the left initial road boundary 2 is replaced by the left target road boundary, and the left initial road boundary 1 is deleted, so that the drivable area formed by the left target road boundary and the right initial road boundary only includes the forward lane, but not the opposite lane, thereby avoiding the vehicle mistakenly driving into the opposite lane again, greatly reducing the risk of collision.

[0146] As another example, Fig. 9 As shown, since the left initial road boundary belongs to the reverse lane and not the forward lane, the automatic driving system can use the final estimated left target road boundary (i.e., the third reference boundary line) to replace the left initial road boundary, so that the drivable area formed by the left target road boundary and the right initial road boundary only includes the forward lane and does not include the reverse lane, thereby avoiding the vehicle from mistakenly driving into the reverse lane and greatly reducing the risk of collision.

[0147] In another embodiment, updating the initial road boundary using the target road boundary may include:

[0148] The right initial road boundary is updated using the right target road boundary.

[0149] When there is an abnormality in the right initial road boundary, the embodiment of the present application can only use the forward vehicle to assist in estimating the right target road boundary of the forward lane, so as to use the right target road boundary to update the right initial road boundary. The updating method of the right initial road boundary can include supplementation.

[0150] As an example, Fig.10 As shown, since there is a missing part in the right initial road boundary of the forward lane, the automatic driving system can use the final estimated right target road boundary (i.e., the fourth reference boundary line) to supplement the missing part of the right initial road boundary, so that the right target road boundary is complete, thereby avoiding the situation where the vehicle drives towards the physical boundary of the forward lane (such as a curb), greatly reducing the risk of collision.

[0151] It can be seen that the embodiment of the present application is not limited to estimating the road boundary of the drivable area based on map geometric elements, but refers to multi-dimensional information for logical inference, wherein the multi-dimensional information provided by the embodiment of the present application includes traffic information and road arrow information. With the enrichment of project requirements or other information, the multi-dimensional information can be expanded so that the accuracy of the final estimation result can be further improved based on the decision tree mechanism.

[0152] It can be seen from this example that the solution provided by the present application, by further referring to multi-dimensional information such as traffic flow and road arrows for logical inference on the basis of utilizing geometric information valuation, thus breaking the limitation of related technologies that only utilize single geometric information for valuation, greatly improving the valuation accuracy, thereby avoiding the situation where the vehicle mistakenly drives into the opposite lane or drives towards the physical boundary of the forward lane during autonomous driving, and greatly reducing the risk of collision.

[0153] Furthermore, the present application makes the final constructed reference boundary line more accurate by comprehensively considering the position information, speed information and vehicle model information of surrounding vehicles.

[0154] Furthermore, traffic flow information and road arrow information can be used alone or in combination for optimization. For example, if there is an abnormality in the left initial road boundary, the present application can use the information of the reverse vehicle, the forward vehicle, and the reverse arrow to construct multiple reference boundary lines, and then select the best reference boundary line as the left target road boundary of the forward lane according to the preset priority and the target distance between each reference boundary line and the vehicle, so as to minimize the situation where the vehicle accidentally drives into the reverse lane; if there is an abnormality in the right initial road boundary, the present application can use the reverse vehicle alone to assist in estimating the right target road boundary of the forward lane, so as to ensure that the vehicle has enough driving space.

[0155] Corresponding to the aforementioned application function implementation method embodiment, the present application also provides a road boundary estimation device, a vehicle, a computer-readable storage medium and corresponding embodiments.

[0156] Fig.11 Schematic diagram of the structure of a road boundary estimation device shown in an embodiment of the present application.

[0157] See also Fig.11 The present application provides a road boundary estimation device, which may include:

[0158] An initial road boundary acquisition module 1110 is used to use the identified road edge as an initial road boundary during the autonomous driving process;

[0159] An abnormality determination module 1120, used to determine whether there is an abnormality in the initial road boundary;

[0160] The reference information acquisition module 1130 is used to acquire the traffic flow information and road arrow information around the vehicle if there is an abnormality in the initial road boundary; wherein the drivable area formed by the initial road boundary includes the forward lane;

[0161] A target road boundary estimation module 1140 is used to estimate the target road boundary of the forward lane according to the traffic flow information and / or road arrow information;

[0162] The initial road boundary updating module 1150 is used to update the initial road boundary using the target road boundary.

[0163] In one embodiment, the traffic flow information includes the direction information, position information and speed information of each surrounding vehicle, and the road arrow information includes the direction information and position information of each road arrow; the target road boundary estimation module 1140 may include:

[0164] A classification submodule is used to classify multiple surrounding vehicles based on the direction information and speed information of each surrounding vehicle to obtain reverse vehicles and forward vehicles, and to classify multiple road arrows based on the direction information of each road arrow to obtain reverse arrows and forward arrows; wherein the driving direction of the reverse vehicle is opposite to the driving direction of the vehicle, the driving direction of the forward vehicle is the same as the driving direction of the vehicle, the direction indicated by the reverse arrow is opposite to the driving direction of the vehicle, and the direction indicated by the forward arrow is the same as the driving direction of the vehicle;

[0165] The target road boundary estimation submodule is used to estimate the target road boundary of the forward lane according to the position information and speed information of the reverse vehicle, and / or the position information and speed information of the forward vehicle, and / or the position information of the reverse arrow.

[0166] In one embodiment, the initial road boundary includes a left initial road boundary, the target road boundary includes a left target road boundary; and the target road boundary estimation submodule may include:

[0167] A first construction unit is used to construct a first reference boundary line according to the position information and speed information of the reverse vehicle, a second reference boundary line according to the position information and speed information of the forward vehicle, and a third reference boundary line according to the position information of the reverse arrow when it is determined that the left initial road boundary has an abnormal situation;

[0168] A target distance calculation unit, used to calculate target distances between the target driving trajectory of the vehicle and the first reference boundary line, the second reference boundary line and the third reference boundary line respectively;

[0169] The left target road boundary estimation unit is used to select a target reference boundary line from the first reference boundary line, the second reference boundary line and the third reference boundary line as the left target road boundary of the forward lane according to a preset priority and a target distance.

[0170] In one embodiment, the initial road boundary update module 1150 may include:

[0171] The left initial road boundary updating submodule is used to update the left initial road boundary by using the left target road boundary.

[0172] In another embodiment, the initial road boundary includes a right initial road boundary, the target road boundary includes a right target road boundary; and the target road boundary estimation submodule may include:

[0173] A second construction unit is used to construct a fourth reference boundary line according to the position information and speed information of the forward vehicle when it is determined that there is an abnormality in the right initial road boundary;

[0174] The right target road boundary estimation unit is used to use the fourth reference boundary line as the right target road boundary of the forward lane.

[0175] In another embodiment, the initial road boundary update module 1150 may include:

[0176] The right initial road boundary updating submodule is used to update the right initial road boundary by using the right target road boundary.

[0177] In one embodiment, the first building block may include:

[0178] a first trajectory determination subunit, configured to determine a first driving trajectory of the oncoming vehicle in the oncoming lane according to the position information and speed information of the oncoming vehicle when it is determined that there is an abnormality at the left initial road boundary, and to determine a second driving trajectory of the oncoming vehicle in the oncoming lane according to the position information and speed information of the oncoming vehicle;

[0179] The first moving subunit is used to move the first driving trajectory to the left by a first preset distance based on the direction information of the reverse vehicle to construct a first reference boundary line; and, based on the direction information of the forward vehicle, move the second driving trajectory to the left by a second preset distance to construct a second reference boundary line; and, based on the direction information of the reverse arrow, construct a third reference boundary line to the left that is a third preset distance away from the position information of the reverse arrow.

[0180] In one embodiment, the second building block may include:

[0181] A second trajectory determination subunit is used to determine a second driving trajectory of the forward vehicle in the forward lane according to the position information and speed information of the forward vehicle when it is determined that there is an abnormality at the right initial road boundary;

[0182] The second moving subunit is used to move the second driving track to the right by a second preset distance based on the orientation information of the forward vehicle, so as to construct a fourth reference boundary line.

[0183] In one embodiment, the initial road boundary includes a left initial road boundary and a right initial road boundary; the abnormal situation determination module 1120 may include:

[0184] A parameter calculation submodule, used to calculate the road width and road width change of the drivable area formed by the left initial road boundary and the right initial road boundary, and to count the number of boundaries of the left initial road boundary close to the vehicle;

[0185] A parameter judgment submodule, used to judge whether the road width is greater than a first threshold, whether the road width change is greater than a second threshold, and whether the number of boundaries is greater than a third threshold;

[0186] The abnormal situation determination submodule is used to determine that there is an abnormal situation in the initial road boundary if the road width is greater than a first threshold value, or the road width change is greater than a second threshold value, or the number of boundaries is greater than a third threshold value; wherein the abnormal situation includes any one of the following: the drivable area contains a reverse lane, the left initial road boundary is incomplete, and the right initial road boundary is incomplete.

[0187] As can be seen from this example, the solution provided by this application uses the identified road edge as the initial road boundary during the autonomous driving process; determines whether there are abnormalities in the initial road boundary; if there are abnormalities in the initial road boundary, obtains the traffic information and road arrow information around the vehicle; wherein the drivable area formed by the initial road boundary includes the forward lane; estimates the target road boundary of the forward lane based on the traffic information and / or road arrow information; and uses the target road boundary to update the initial road boundary. This application further refers to multi-dimensional information such as traffic flow and road arrows for logical inference on the basis of using geometric information valuation, thus breaking the limitation of the related technology of using only single geometric information valuation, greatly improving the valuation accuracy, thereby avoiding the situation where the vehicle mistakenly drives into the reverse lane or drives towards the physical boundary of the forward lane during autonomous driving, and greatly reducing the risk of collision.

[0188] Regarding the device in the above embodiment, the specific manner in which each module performs the operation has been described in detail in the embodiment of the method, and will not be elaborated again here.

[0189] Fig.12 It is a schematic diagram of the structure of a vehicle shown in an embodiment of the present application.

[0190] See also Fig.12 , the vehicle 1200 includes a memory 1210 and a processor 1220 .

[0191] The processor 1220 may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor or the processor may be any conventional processor, etc.

[0192] The memory 1210 may include various types of storage units, such as system memory, read-only memory (ROM), and permanent storage. Among them, ROM can store static data or instructions required by the processor 1220 or other modules of the computer. The permanent storage device may be a readable and writable storage device. The permanent storage device may be a non-volatile storage device that does not lose the stored instructions and data even after the computer is powered off. In some embodiments, the permanent storage device uses a large-capacity storage device (such as a magnetic or optical disk, flash memory) as a permanent storage device. In some other embodiments, the permanent storage device may be a removable storage device (such as a floppy disk, optical drive). The system memory may be a readable and writable storage device or a volatile readable and writable storage device, such as a dynamic random access memory. The system memory may store some or all instructions and data required by the processor at run time. In addition, the memory 1210 may include any combination of computer-readable storage media, including various types of semiconductor memory chips (such as DRAM, SRAM, SDRAM, flash memory, programmable read-only memory), and disks and / or optical disks may also be used. In some embodiments, the memory 1210 may include a readable and / or writable removable storage device, such as a laser disc (CD), a read-only digital versatile disc (such as a DVD-ROM, a double-layer DVD-ROM), a read-only Blu-ray disc, an ultra-density optical disc, a flash memory card (such as an SD card, a mini SD card, a Micro-SD card, etc.), a magnetic floppy disk, etc. The computer-readable storage medium does not include carrier waves and transient electronic signals transmitted wirelessly or wired.

[0193] The memory 1210 stores executable codes, and when the executable codes are processed by the processor 1220 , the processor 1220 may execute part or all of the methods described above.

[0194] In addition, the method according to the present application may also be implemented as a computer program or a computer program product, which includes computer program code instructions for executing some or all of the steps in the above method of the present application.

[0195] Alternatively, the present application can also be implemented as a computer-readable storage medium (or non-temporary machine-readable storage medium or machine-readable storage medium) on which executable code (or computer program or computer instruction code) is stored. When the executable code (or computer program or computer instruction code) is executed by the vehicle's processor, the processor executes part or all of the steps of the above-mentioned method according to the present application.

[0196] The embodiments of the present application have been described above, and the above description is exemplary, not exhaustive, and is not limited to the disclosed embodiments. Many modifications and changes will be apparent to those of ordinary skill in the art without departing from the scope and spirit of the described embodiments. The selection of terms used herein is intended to best explain the principles of the embodiments, practical applications, or improvements to the technology in the market, or to enable other persons of ordinary skill in the art to understand the embodiments disclosed herein.

Claims

1. A road boundary estimation method, characterized in that: include: During the autonomous driving process, the identified road edge is used as the initial road boundary; Determining whether there is an abnormality in the initial road boundary; If there is an abnormality in the initial road boundary, the traffic flow information and road arrow information around the vehicle are obtained; wherein the drivable area formed by the initial road boundary includes the forward lane; estimating a target road boundary of the forward lane according to the traffic flow information and / or the road arrow information; The target road boundary is adopted to update the initial road boundary.

2. The method according to claim 1, characterized in that The traffic flow information includes the direction information, position information and speed information of each surrounding vehicle, and the road arrow information includes the direction information and position information of each road arrow; The estimating the target road boundary of the forward lane according to the traffic flow information and / or the road arrow information includes: Based on the orientation information and speed information of each of the surrounding vehicles, the plurality of surrounding vehicles are classified to obtain reverse vehicles and forward vehicles, and based on the orientation information of each of the road arrows, the plurality of road arrows are classified to obtain reverse arrows and forward arrows; wherein the driving direction of the reverse vehicle is opposite to the driving direction of the vehicle, the driving direction of the forward vehicle is the same as the driving direction of the vehicle, the direction indicated by the reverse arrow is opposite to the driving direction of the vehicle, and the direction indicated by the forward arrow is the same as the driving direction of the vehicle; The target road boundary of the forward lane is estimated based on the position information and speed information of the reverse vehicle, and / or the position information and speed information of the forward vehicle, and / or the position information of the reverse arrow.

3. The method according to claim 2, characterized in that The initial road boundary includes a left initial road boundary, and the target road boundary includes a left target road boundary; estimating the target road boundary of the forward lane according to the position information and speed information of the reverse vehicle, and / or the position information and speed information of the forward vehicle, and / or the position information of the reverse arrow, includes: When it is determined that the left initial road boundary has the abnormal situation, constructing a first reference boundary line according to the position information and speed information of the oncoming vehicle, constructing a second reference boundary line according to the position information and speed information of the forward vehicle, and constructing a third reference boundary line according to the position information of the reverse arrow; Calculating target distances between the target driving trajectory of the vehicle and the first reference boundary line, the second reference boundary line, and the third reference boundary line respectively; According to the preset priority and the target distance, selecting a target reference boundary line from the first reference boundary line, the second reference boundary line and the third reference boundary line as the left target road boundary of the forward lane; The adopting the target road boundary to update the initial road boundary includes: The left initial road boundary is updated using the left target road boundary.

4. The method according to claim 2, characterized in that: The initial road boundary includes a right initial road boundary, and the target road boundary includes a right target road boundary; estimating the target road boundary of the forward lane according to the position information and speed information of the reverse vehicle, and / or the position information and speed information of the forward vehicle, and / or the position information of the reverse arrow, includes: When it is determined that the right initial road boundary has the abnormal situation, constructing a fourth reference boundary line according to the position information and speed information of the forward vehicle; Using the fourth reference boundary line as the right target road boundary of the forward lane; The adopting the target road boundary to update the initial road boundary includes: The right initial road boundary is updated using the right target road boundary.

5. The method according to claim 3, characterized in that: When it is determined that the left initial road boundary has the abnormal situation, a first reference boundary line is constructed according to the position information and speed information of the oncoming vehicle, a second reference boundary line is constructed according to the position information and speed information of the forward vehicle, and a third reference boundary line is constructed according to the position information of the reverse arrow, including: When it is determined that the left initial road boundary has the abnormal situation, determining a first driving trajectory of the oncoming vehicle in the oncoming lane according to the position information and speed information of the oncoming vehicle, and determining a second driving trajectory of the oncoming vehicle in the oncoming lane according to the position information and speed information of the oncoming vehicle; Based on the orientation information of the reverse vehicle, the first driving trajectory is moved to the left by a first preset distance to construct a first reference boundary line; and, based on the orientation information of the forward vehicle, the second driving trajectory is moved to the left by a second preset distance to construct a second reference boundary line; and, based on the orientation information of the reverse arrow, a third reference boundary line is constructed to the left at a third preset distance from the position information of the reverse arrow.

6. The method according to claim 4, characterized in that When it is determined that the right initial road boundary has the abnormal situation, constructing a fourth reference boundary line according to the position information and speed information of the forward vehicle, comprises: When it is determined that the abnormal situation exists at the right initial road boundary, determining a second driving trajectory of the forward vehicle in the forward lane according to the position information and speed information of the forward vehicle; Based on the orientation information of the forward vehicle, the second driving track is moved to the right by a second preset distance to construct a fourth reference boundary line.

7. The method according to claim 1, characterized in that The initial road boundary includes a left initial road boundary and a right initial road boundary; and the determining whether there is an abnormality in the initial road boundary includes: Calculating a road width and a road width change of a drivable area formed by the left initial road boundary and the right initial road boundary, and counting the number of boundaries close to the left initial road boundary of the vehicle; Determining whether the road width is greater than a first threshold, determining whether the road width change is greater than a second threshold, and determining whether the boundary quantity is greater than a third threshold; If the road width is greater than a first threshold or the road width change is greater than a second threshold or the number of boundaries is greater than a third threshold, it is determined that an abnormality exists in the initial road boundary; wherein the abnormality includes any one of the following: the drivable area contains a reverse lane, the left initial road boundary is incomplete, and the right initial road boundary is incomplete.

8. A road boundary estimation device, characterized in that: include: An initial road boundary acquisition module is used to use the identified road edge as the initial road boundary during the autonomous driving process; An abnormality judgment module, used to judge whether there is an abnormality in the initial road boundary; A reference information acquisition module, used for acquiring traffic information and road arrow information around the vehicle if there is an abnormality in the initial road boundary; wherein the drivable area formed by the initial road boundary includes the forward lane; A target road boundary estimation module, used for estimating the target road boundary of the forward lane according to the traffic flow information and / or the road arrow information; The initial road boundary updating module is used to update the initial road boundary by using the target road boundary.

9. A vehicle, characterized in that: include: processor; as well as A memory having executable codes stored thereon, which, when executed by the processor, causes the processor to execute the method according to any one of claims 1 to 7.

10. A computer-readable storage medium having executable codes stored thereon, which, when executed by a processor of a vehicle, causes the processor to execute the method according to any one of claims 1 to 7.

Images