Lane-level path planning method and device applied to autonomous vehicle

Abstract

Disclosed is a lane-level path planning method and device applied to an autonomous vehicle. The lane-level path planning method obtains lane network data by using a high-definition map, then converts a starting point and a target point of a task into a starting lane and a target lane based on the lane network data, and searches the lane network data by using the starting lane and the target lane, so as to convert a lane path search problem into a conventional path search problem. Moreover, in the step-by-step search from the starting lane to the target lane, the difficulty of each searched lane to reach the target lane is identified by a path cost, and the lane with the minimum path cost is selected at each step to search the next lane, so that the path planning formed gradually tends to be the path planning of the target lane.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of automatic driving, and in particular to a lane-level path planning method and device applied to an automatic driving vehicle. BACKGROUND

[0002] In recent years, with the development of artificial intelligence technology, automatic driving technology is becoming more and more mature. The automatic driving technology can be briefly divided into perception, prediction, positioning, decision-making, planning and control. At present, the path planning of the automatic driving vehicle is lane-level path planning on a high-precision map, and most of the traditional path planning algorithms take the road as the smallest path guiding unit, which is not suitable for the current automatic driving vehicle. In addition, the length, shape and bending degree of each lane are different, which brings great challenges to the design of the lane-level path planning method. SUMMARY

[0003] Therefore, the present application provides a lane-level path planning method and device applied to an automatic driving vehicle.

[0004] According to a first aspect of the present application, a lane-level path planning method applied to an automatic driving vehicle is provided, comprising:

[0005] obtaining a start lane and a target lane according to a start point coordinate and a target point coordinate of a task;

[0006] setting the start lane as a current lane;

[0007] selecting a next lane of the current lane which has not been selected to calculate a path cost thereof;

[0008] saving the next lane and the path cost thereof into a specified lane set;

[0009] judging whether the next lane is the target lane, if yes, outputting a path planning from the start lane to the target lane, if no, executing the following steps: judging whether all the next lanes of the current lane have been processed, if no, jumping to the step of selecting a next lane of the current lane which has not been selected to calculate a path cost thereof, if yes, setting a lane with a minimum path cost as the current lane in the specified lane set which has not been selected as the current lane, and jumping to the step of selecting a next lane of the current lane which has not been selected to calculate a path cost thereof.

[0010] In some embodiments, the path cost of a specified lane is positively correlated with a straight-line distance between the specified lane and the target lane and a length of the specified lane, and is negatively correlated with an average speed of the specified lane, the specified lane being the start lane, the current lane or the target lane.

[0011] In some embodiments, in the calculation of the path cost of the specified lane, the straight-line distance between the specified lane and the target lane, the included angle between the specified lane and the target lane, the length of the specified lane, and the average speed of the specified lane are represented by different weights to represent the degree of relevance of the path cost of the specified lane.

[0012] In some embodiments, the saving of the next lane and its path cost into the specified lane set is: only under the condition that the next lane is determined to be less than a set threshold, the next lane and its path cost are saved into the specified lane set.

[0013] In some embodiments, further comprising: calculating the path cost of the starting lane; and the set threshold is equal to the path cost of the starting lane.

[0014] According to a second aspect of the present application, a lane-level path planning device applied to an autonomous vehicle is provided, comprising:

[0015] A data preprocessing module is configured to obtain correlation relationship data between each lane, and retrieve the correlation relationship data by using the starting point coordinates and the target point coordinates of a task to obtain a starting lane and a target lane.

[0016] A lane-level path search module is configured to set the starting lane as a current lane; select a next lane of the current lane that has not been selected to calculate a path cost thereof; save the next lane and its path cost into a specified lane set; determine whether the next lane is the target lane, and if so, output a path planning from the starting lane to the target lane, and if not, perform the following steps: determine whether all next lanes of the current lane have been processed, and if not, jump to the step of selecting a next lane of the current lane that has not been selected to calculate a path cost thereof, and if so, set a lane with the smallest path cost among the lanes in the specified lane set that have not been selected as the current lane, and jump to the step of selecting a next lane of the current lane that has not been selected to calculate a path cost thereof.

[0017] In some embodiments, the path cost of a specified lane is positively correlated with the straight-line distance between the specified lane and the corresponding coordinate point of the target lane, and the length of the specified lane, and is negatively correlated with the average speed of the specified lane, the specified lane being the starting lane, the current lane, or the target lane.

[0018] In some embodiments, in the calculation of the path cost of the specified lane, the straight-line distance between the corresponding coordinate points of the specified lane and the target lane, the included angle between the specified lane and the target lane, the length of the specified lane, and the average speed of the specified lane are represented by different weights to represent the degree of relevance of the path cost of the specified lane.

[0019] According to a third aspect of the present application, an electronic device is provided, comprising a memory and a processor, the memory further storing computer instructions executable by the processor, the computer instructions being executed to implement the lane-level path planning method described above.

[0020] According to a fourth aspect of the present application, a computer readable medium is provided, the computer readable medium storing computer instructions executable by an electronic device, the computer instructions being executed to implement the lane-level path planning method described above.

[0021] The lane-level planning method and device provided by the present application use high-precision map to obtain lane network data, and then convert the starting point and target point of a task into a starting lane and a target lane based on the lane network data, and search the lane network data using the starting lane and the target lane, thereby converting the lane path search problem into a conventional path search problem. Moreover, in the step-by-step search from the starting lane to the target lane, the path cost is used to identify the degree of difficulty of each searched lane to reach the target lane, and the lane with the minimum path cost is selected at each step to search the next lane. The path planning formed in this way is a path planning gradually approaching the target lane.

[0022] In addition, the present application also points out the parameters positively correlated with the path cost of the specified lane, including the straight-line distance between the corresponding coordinate points of the specified lane and the target lane, the included angle between the specified lane and the target lane, the length of the specified lane, and the average speed of the specified lane, and accordingly provides a specific path cost calculation formula. BRIEF DESCRIPTION OF DRAWINGS

[0023] The above and other objects, features and advantages of the present application will become more apparent from the following description of embodiments of the present application, taken in conjunction with the accompanying drawings, in which:

[0024] Figure 1 is a flowchart of a lane-level path planning method for an autonomous vehicle according to an embodiment of the present application;

[0025] Figure 2 is a flowchart of a lane-level path planning method for an autonomous vehicle according to another embodiment of the present application

[0026] Figure 3 is a schematic diagram for explaining an embodiment of the present application;

[0027] Figure 4 is a structural diagram of a lane-level path planning device for an autonomous vehicle according to an embodiment of the present application;

[0028] Figure 5 A structural schematic diagram of an electronic device for implementing various embodiments of the present application. DETAILED DESCRIPTION

[0029] The present application will be described in more detail with reference to the accompanying drawings. In the various drawings, like reference numerals refer to like elements throughout. For the sake of clarity, each portion in the drawings is not drawn to scale. In addition, some known portions can not be shown.

[0030] Figure 1 is a flowchart of a lane-level path planning method for an autonomous vehicle according to an embodiment of the present application. Specifically, it includes steps S101 to S108.

[0031] In step S101, the starting lane and the target lane are obtained according to the starting point and the target point coordinates of the task.

[0032] In step S102, the starting lane is set as the current lane.

[0033] In step S103, the unselected next lane of the current lane is selected to calculate the path cost.

[0034] In step S104, the next lane and its path cost are saved to the designated lane set.

[0035] In step S105, it is determined whether the next lane is the target lane. If yes, step S108 is executed, and if no, step S106 is executed.

[0036] In step S106, it is determined whether all next lanes have been processed. If yes, step S107 is executed, and if no, step S103 is executed.

[0037] In step S107, the lane with the minimum path cost is set as the current lane among the unselected lanes in the designated lane set.

[0038] In step S108, the lane planning is output.

[0039] The lane network indicates the association relationship between the lanes. The lane network can usually be obtained from a high-definition map. The association relationship between the lanes indicates that each lane is bound to adjacent previous and next lanes. Each lane can have multiple adjacent previous and next lanes, for example, a next lane of a certain lane can be a straight lane, a left-turn lane, or a right-turn lane. In the lane network, each lane can have a unique ID, and record the start point coordinates, end point coordinates, lane direction (i.e., the angle between the lane and the coordinate system), lane length, and other attribute information of each lane in the Earth coordinate system. Some information related to the lane can also be recorded, such as a car. Based on the lane network, the start point coordinates and the target point coordinates of the task assigned to the autonomous vehicle are used to retrieve the lane network, respectively, to obtain the lane where the start point is located and the lane where the target point is located. In this document, the lane where the start point is located is referred to as the start lane, and the lane where the target point is located is referred to as the target lane.

[0040] Step S102 is an assignment step in the computer program. Through this step, the current lane as a variable is equal to the start lane. After assignment, the current lane has the same attribute information as the start lane.

[0041] Steps S103 to S108 of the present example are a loop body. In the loop body, a lane that has not been selected in the loop body is selected from the multiple next lanes of the current lane, the path cost of the lane is calculated according to the path cost algorithm, the lane is temporarily stored in the designated lane set (the designated lane set is initially empty), and then it is determined whether the lane is the target lane. If the lane is the target lane, the loop is exited. If the lane is not the target lane, it is determined whether all next lanes of the current lane have been processed. If all next lanes of the current lane have not been processed, the step of selecting a lane that has not been selected in the loop body from the multiple next lanes of the current lane and assigning the lane to the current lane is executed, and the step of calculating the path cost of the lane according to the path cost algorithm is executed. If all next lanes of the current lane have been processed, the lane with the minimum path cost is set as the current lane from the lanes that have not been selected for the current lane in the designated lane set, i.e., the current lane is equal to the lane that has not been selected and has the minimum path cost, and then the loop is continued to step S103.

[0042] Figure 3is a network diagram for explaining embodiments of the present application. When the autonomous vehicle is to reach R10 from R1, the path cost of R1 is calculated as cost1, and the next lanes of R1, R2 and R3 are found. The path costs of R2 and R3 are calculated as cost2 and cost3, and stored in set. At this time, set is {(R2, cost2), (R3, cost3)}. Under the condition that cost2 is less than cost3, the next lanes of R2, R4, R5 and R6 are found. The path costs of R4, R5 and R6 are calculated as cost4, cost5 and cost6, and stored in set. At this time, set is {(R2, cost2), (R3, cost3), (R4, cost4), (R5, cost5), (R6, cost6)}. At this time, cost5 is less than cost3, cost4 and cost6, and the next lanes of R5, R7 to R9 are found. The path costs of R7 to R9 are calculated as cost7 to cost9 (but if cost3 is less than cost4, cost5 and cost6, return to R3 for calculation). At this time, set is {(R2, cost2), (R3, cost3), (R4, cost4), (R5, cost5), (R6, cost6), (R7, cost7), (R8, cost8), (R9, cost9)}. Under the condition that cost8 is less than cost3, cost4, cost6, cost7 and cost9, the next lane R10 is found, and the target cost of R10 is calculated as cost10. Through the diagram, the path planning of R1 to R10 is finally obtained as R1->R2->R4->R8->R10.

[0043] Figure 2 is a flowchart of a lane-level path planning method for an autonomous vehicle according to an embodiment of the present application. The path cost of the starting lane is calculated, and the path cost of the starting lane is set as the threshold value. The path costs of the next lanes of the starting lane are calculated, and the path costs of the next lanes are stored in set. At this time, set is {(R2, cost2), (R3, cost3)}. Under the condition that cost2 is less than cost3, the next lanes of R2, R4, R5 and R6 are found. The path costs of R4, R5 and R6 are calculated as cost4, cost5 and cost6, and stored in set. At this time, set is {(R2, cost2), (R3, cost3), (R4, cost4), (R5, cost5), (R6, cost6)}. At this time, cost5 is less than cost3, cost4 and cost6, and the next lanes of R5, R7 to R9 are found. The path costs of R7 to R9 are calculated as cost7 to cost9 (but if cost3 is less than cost4, cost5 and cost6, return to R3 for calculation). At this time, set is {(R2, cost2), (R3, cost3), (R4, cost4), (R5, cost5), (R6, cost6), (R7, cost7), (R8, cost8), (R9, cost9)}. Under the condition that cost8 is less than cost3, cost4, cost6, cost7 and cost9, the next lane R10 is found, and the target cost of R10 is calculated as cost10. Through the diagram, the path planning of R1 to R10 is finally obtained as R1->R2->R4->R8->R10. Figure 2 Figure 1 It can be found by comparison that, Figure 2 The step S204 is added, and specifically, Figure 2 The embodiment shown only temporarily stores the information of the current lane with a path cost less than the set threshold value in the specified lane set, and the information of the current lane that does not meet the condition is not stored in the specified lane set. In this way, the number of lanes to be searched can be reduced, and the search calculation amount can be reduced. In some embodiments, the path cost of the starting lane is calculated, and the set threshold value is set as the path cost of the starting lane. That is, only the information of the current lane with a path cost less than the path cost of the starting lane is temporarily stored in the specified lane set.

[0044] ​As mentioned above, the path cost of a designated lane is calculated. The path cost of a designated lane is a quantitative indicator proposed in this application to measure the ease of reaching the target lane from the designated lane. The path cost of a designated lane is positively correlated with parameters such as the straight-line distance between the designated and target lanes, the length of the designated lane, and the angle between the two lanes, and negatively correlated with the average speed of the designated lane. A positive correlation means that a larger parameter value results in a larger path cost, and a smaller parameter value results in a smaller path cost; conversely, a negative correlation has the opposite effect. In other words, the greater the straight-line distance between the designated and target lanes, the larger the angle between them, the longer the designated lane, and the lower the average speed of the autonomous vehicle in the designated lane, the greater the path cost. Here, the straight-line distance between the designated and target lanes can be calculated by selecting a point on each lane and using the straight-line distance between these two points as the straight-line distance between the designated and target lanes. For example, the straight-line distance between the start point of the current lane and the start point of the target lane can be calculated and used as the straight-line distance between the designated and target lanes. However, it is preferable to use the straight-line distance between the end point of the current lane and the end point of the target lane as the straight-line distance between the current and target lanes. Multiple lanes can share the same point to calculate straight-line distances, such as multiple lanes at an intersection. In one processing method, the angle between the specified lane and the target lane refers to the angle between the line connecting the starting point coordinates and the target point coordinates of the task and the line connecting the selected point (e.g., the endpoint) on the specified lane and the target point coordinates of the task.

[0045] An algorithmic formula for the path cost of a specified lane can be constructed based on positively correlated parameters, and this formula can be fitted using experimental data. Furthermore, different weights can be used to characterize the correlation between the straight-line distance between corresponding coordinate points of the specified lane and the target lane, the angle between the specified lane and the target lane, the length of the specified lane, and the average speed of the specified lane with the path cost of the specified lane.

[0046] After research, the formula for calculating a specific path cost is formula (1):

[0047] cost = w1 * sqrt((x currentlane ―x targetlane ) 2 +(y currentlane ―y targetlane ) 2 )+w2*abs(heading currentlane ―heading targetlane )+w3*lanelength currentlane -w4*lanespeed currentlane Formula (1)

[0048] Where x currentlane and xtargetlane y represents the x-coordinate of a point in the current lane and the corresponding point in the target lane within the same coordinate system. currentlane and y targetlane The heading is the ordinate value of a point in the current lane and the corresponding point in the target lane within the same coordinate system. currentlane ―heading targetlane The angle between the line connecting the starting point to a point in the target lane and the line connecting a point in the current lane to a point in the target lane, where sqrt is the square root, abs is the absolute value, and lanelength is the absolute value. currentlane The lane speed refers to the length of the current lane. currentlane The average speed of the current lane is represented by w1 to w4, which are weights greater than 0 and can be adjusted according to the actual situation. In some embodiments, when calculating the included angle, a point in the target lane is selected as the target point, and a point in the current lane is selected as the end point of the current lane.

[0049] Figure 4 This is a structural diagram of a lane-level path planning device for autonomous vehicles provided according to an embodiment of this application. As shown in the figure, the device 400 includes a data preprocessing module 401 and a lane-level path search module 402.

[0050] The data preprocessing module 401 is used to acquire the correlation data between lanes and retrieve the correlation data using the starting point coordinates and target point coordinates of the task to obtain the starting lane and target lane. Since high-precision maps can be provided by different vendors, it is necessary to first organize the correlation data between lanes using existing high-precision maps. The correlation data can use a unique ID to identify each lane and record attribute information such as the starting point coordinates, ending point coordinates, lane direction (i.e., the angle between the lane and the coordinate system), and lane length of each lane in the Earth coordinate system. With the correlation data of each lane, the starting point coordinates and target point coordinates of the task can be used to retrieve the correlation data to obtain the starting lane and target lane.

[0051] Lane-level path search module 402 is used to perform Figure 1The steps S102-S108 include: setting the starting lane as the current lane; selecting a next lane of the current lane which has not been selected to calculate the path cost of the next lane; saving the next lane and the path cost of the next lane into the designated lane set; determining whether the next lane is the target lane, if yes, outputting the path planning from the starting lane to the target lane, if no, performing the following steps: determining whether all the next lanes of the current lane have been processed, if no, jumping to selecting a next lane of the current lane which has not been selected to calculate the path cost of the next lane, if yes, setting the lane with the minimum path cost as the current lane from the lanes in the designated lane set which have not been selected as the current lane, and jumping to selecting a next lane of the current lane which has not been selected to calculate the path cost of the next lane.

[0052] In some embodiments, the path cost of the designated lane is a quantitative index for measuring the difficulty of reaching the target lane from the designated lane, and the path cost of the designated lane can be calculated using the coordinates of the corresponding points of the designated lane and the target lane, the included angle between the designated lane and the target lane, the length of the designated lane, and the average speed of the autonomous vehicle on the designated lane. In a more specific embodiment, the path cost of the designated lane can be calculated using formula (1) above.

[0053] In some embodiments, the lane-level path search module 402 can only temporarily store the information of the current lane with a path cost less than a set threshold in the designated lane set, and the information of the current lane not meeting the condition will not be stored in the designated lane set. That is, only the lane with a path cost less than the set threshold is searched in the next stage, so as to reduce the search calculation amount. In addition, if it is found in the search that the path cost of all the lower-level lanes is too large, the search can be returned to the same-level lane or the upper-level lane of the current lane, as described above. Figure 2

[0054] The lane-level planning method and device provided in the present application use the high-precision map to obtain lane network data, and then convert the starting point and the target point of the task into a starting lane and a target lane based on the lane network data, and search the lane network data using the starting lane and the target lane, so as to convert the lane path search problem into a conventional path search problem. Moreover, in the step-by-step search from the starting lane to the target lane, the path cost is used to identify the difficulty of reaching the target lane from each searched lane, and the lane with the minimum path cost is selected at each step to search the next lane. The path planning formed in this way is a path planning gradually approaching the target lane.

[0055] ​In addition, the application also points out a positive correlation parameter of the path cost of the specified lane, including the straight line distance of the corresponding coordinate points of the specified lane and the target lane, the included angle between the specified lane and the target lane, the length of the specified lane and the average speed of the specified lane, and provides a specific path cost calculation formula accordingly.

[0056] Reference Figure 5 The electronic device 500 includes a processor 501, a memory 502 and an input / output device 503 connected by a bus. The memory 502 includes a read-only memory (ROM) and a random access memory (RAM), and the memory 502 stores various computer instructions and data required for system functions, and the processor 501 reads various computer instructions from the memory 502 to perform various appropriate actions and processes. The input / output device includes an input part of a keyboard, a mouse, etc.; an output part including a cathode ray tube (CRT), a liquid crystal display (LCD), etc., and a speaker, etc.; a storage part including a hard disk, etc.; and a communication part including a network interface card such as a LAN card, a modem, etc. The memory 502 also stores computer instructions that, when executed, complete the steps in the above embodiments.

[0057] Accordingly, the embodiments of the application provide a computer readable storage medium storing computer instructions, which, when executed, implement the steps in the embodiments.

[0058] The computer readable medium can be a computer readable signal medium or a computer readable storage medium. The computer readable storage medium, for example, and without limitation, is an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus or device, or any suitable combination of the above. More specific examples of the computer readable storage medium include an electrical connection having one or more wires, a portable computer diskette, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In this context, a computer readable storage medium can be any tangible medium that contains or stores a program for use by or in connection with a processing unit, apparatus or device.

[0059] A computer readable signal medium can include a propagated data signal with computer readable program code embodied therein, for example, in baseband or as part of a carrier wave. Such a propagated signal can take any of a variety of forms, including, but not limited to, electro-magnetic, optical, or any suitable combination thereof. A computer readable signal medium can be any computer readable medium that can be involved in

[0060] The computer readable program code embodied on a computer readable medium can be transmitted using any appropriate medium, including but not limited to wireless, wire line, optical fiber cable, RF, etc., and any suitable combination thereof.

[0061] Computer program code for carrying out operations of embodiments of the present application can be written in any combination of one or more programming languages, including an object oriented programming language such as JAVA, C++, or a conventional procedural programming language, such as the C programming language. The program code can execute entirely on the user's computer, partly on the user's computer, as a stand-alone software package, partly on the user's computer and partly on a remote computer or entirely on the remote computer or server. In the latter scenario, the remote computer can be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or the connection can be made to an external computer (for example, through the Internet using an Internet Service Provider). In an embodiment of the application, the remote computer can be a server or another desktop computer.

[0062] It is to be appreciated that the terms "first" and "second" and the like are used herein only to distinguish one element from another, and do not necessarily have an ordinal or chronological significance. Also, the terms "comprises", "comprising", or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but can include other elements not expressly listed or inherent to such process, method, article, or apparatus. Furthermore, the terms "comprises", "comprising", or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but can include other elements not expressly listed or inherent to such process, method, article, or apparatus. Furthermore, the terms "comprises", "comprising", or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but can include other elements not expressly listed or inherent to such process, method, article, or apparatus. Furthermore, the terms "comprises", "comprising", or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but can include other elements not expressly listed or inherent to such process, method, article, or apparatus.

[0063] In the drawings and specification, there have been disclosed typical embodiments of the application and, although specific terms are employed, they are used in a generic and descriptive sense only and not for purposes of limitation, the scope of the inventive teachings being set forth in the following claims.

Claims

1. A lane-level path planning method applied to an autonomous vehicle, comprising: obtaining a start lane and a target lane according to a start point coordinate and a target point coordinate of a task; setting the start lane as a current lane; selecting a next lane of the current lane which has not been selected before to calculate a path cost of the next lane, the path cost of the next lane representing a degree of difficulty of the next lane to reach the target lane; saving the next lane and the path cost of the next lane to a designated lane set; judging whether the next lane is the target lane, and if yes, outputting a path planning from the start lane to the target lane, and if no, executing the following steps: judging whether all the next lanes of the current lane have been processed, and if no, jumping to the step of selecting a next lane of the current lane which has not been selected before to calculate a path cost of the next lane, and if yes, setting a lane with a minimum path cost as the current lane from the lanes in the designated lane set which have not been selected before as the current lane, and jumping to the step of selecting a next lane of the current lane which has not been selected before to calculate a path cost of the next lane.

2. The lane-level path planning method of claim 1, wherein, The path cost of a designated lane is positively correlated with a straight-line distance between the designated lane and the target lane, a length of the designated lane, and negatively correlated with an average speed of the designated lane, the designated lane being the start lane, the current lane or the target lane.

3. The lane-level path planning method of claim 2, wherein, In the calculation of the path cost of the designated lane, different weights are used to represent a correlation degree of the straight-line distance between the designated lane and the target lane, an included angle between the designated lane and the target lane, the length of the designated lane, and the average speed of the designated lane to the path cost of the designated lane.

4. The lane-level path planning method of claim 1, wherein, The saving of the next lane and the path cost of the next lane to the designated lane set is that the next lane and the path cost of the next lane are saved to the designated lane set only when the path cost of the next lane is less than a set threshold.

5. The lane-level path planning method of claim 4, further comprising: calculating a path cost of the start lane; the set threshold is equal to the path cost of the start lane. 6.A lane-level path planning device applied to an autonomous vehicle, comprising: a data preprocessing module configured to obtain correlation relationship data between lanes, and to retrieve the correlation relationship data according to a start point coordinate and a target point coordinate of a task to obtain a start lane and a target lane; a lane-level path searching module configured to set the start lane as a current lane; select a next lane of the current lane which has not been selected before to calculate a path cost of the next lane; save the next lane and the path cost of the next lane to a designated lane set, the path cost of the next lane representing a degree of difficulty of the next lane to reach the target lane; If the next lane is the target lane, a path from the start lane to the target lane is outputted; otherwise, the following steps are performed: if all the next lanes of the current lane have been processed, the path cost of the next lane is calculated by selecting the unselected next lane of the current lane; otherwise, the lane with the minimum path cost is set as the current lane from the unselected lanes of the designated lane set, and the path cost of the next lane is calculated by selecting the unselected next lane of the current lane.

7. The lane-level path planning device according to claim 6, wherein The path cost of a designated lane is positively related to the straight-line distance between the corresponding coordinate points of the designated lane and the target lane, and the length of the designated lane, and is negatively related to the average speed of the designated lane, wherein the designated lane is the start lane, the current lane or the target lane.

8. The lane-level path planning device according to claim 7, wherein In the calculation of the path cost of the designated lane, the straight-line distance between the corresponding coordinate points of the designated lane and the target lane, the included angle between the designated lane and the target lane, the length of the designated lane and the average speed of the designated lane are represented by different weights to represent the degree of influence of the above factors on the path cost of the designated lane. 9.An electronic device comprising a memory and a processor, wherein the memory further stores computer instructions executable by the processor, and the computer instructions, when executed, implement the lane-level path planning method according to any one of claims 1 to 5. 10.A computer readable medium storing computer instructions executable by an electronic device, and the computer instructions, when executed, implement the lane-level path planning method according to any one of claims 1 to 5.

Images