Vehicle travel control method and vehicle

Abstract

This application provides a vehicle driving control method and a vehicle. The method includes: in response to the vehicle traveling to a preset distance from an intersection, acquiring surrounding lane information at the vehicle's location and intersection lane information; determining surrounding lane evaluation values ​​based on preset evaluation values ​​and the surrounding lane information, and determining intersection lane evaluation values ​​based on preset total evaluation values ​​and intersection lane information, wherein the surrounding lane evaluation values ​​represent the evaluation value of any surrounding lane, and the intersection lane evaluation values ​​represent the evaluation value of any intersection lane; determining at least one candidate lane from the surrounding lanes based on the surrounding lane evaluation values ​​and the intersection lane evaluation values; determining a target lane from the at least one candidate lane based on the traffic state of the at least one candidate lane; and controlling the vehicle to travel in the target lane. This application solves the technical problem of how to flexibly select lanes at intersections where the number of lanes changes at any time.

Description

Technical Field

[0001] This application relates to the field of vehicles, and more specifically, to a vehicle driving control method and a vehicle. Background Technology

[0002] In autonomous driving, lane selection is a core element in achieving navigation intent.

[0003] Traditional solutions typically employ high-definition maps or similar lightweight map solutions. The system relies entirely on the pre-defined lane connections within the high-definition map, allowing the vehicle to execute lane sequences according to a pre-defined global path. However, if these pre-defined lanes are temporarily altered, the autonomous driving system will fall into a "decision deadlock." Furthermore, due to the long update cycle of high-definition maps, such problems occur frequently.

[0004] There is currently no good solution to the above problems. Summary of the Invention

[0005] This application provides a vehicle driving control method and a vehicle to at least solve the technical problem of how to flexibly select lanes at intersections where the number of lanes changes at any time.

[0006] According to one aspect of the embodiments of this application, a vehicle driving control method is provided, comprising: in response to a vehicle traveling to a preset distance from an intersection, acquiring surrounding lane information at the vehicle's location and intersection lane information at the intersection; determining surrounding lane evaluation values ​​based on preset evaluation values ​​and surrounding lane information, and determining intersection lane evaluation values ​​based on preset total evaluation values ​​and intersection lane information, wherein the surrounding lane evaluation values ​​represent the evaluation values ​​of any surrounding lane, and the intersection lane evaluation values ​​represent the evaluation values ​​of any intersection lane; determining at least one candidate lane from the surrounding lanes based on the surrounding lane evaluation values ​​and intersection lane evaluation values; determining a target lane from the at least one candidate lane based on the traffic state of the at least one candidate lane; and controlling the vehicle to travel in the target lane.

[0007] Furthermore, obtaining information on surrounding lanes at the vehicle's location and information on intersection lanes at the intersection includes: obtaining surrounding lane information through the vehicle's onboard perception system, wherein the surrounding lane information includes at least the number of surrounding lanes at the vehicle's location; and obtaining intersection lane information through the vehicle's in-vehicle navigation system, wherein the intersection lane information includes at least the number of intersection lanes, the lane's driving direction attribute, and the driving intention at the intersection.

[0008] Furthermore, determining the surrounding lane assessment value based on the preset total assessment value and surrounding lane information, and determining the intersection lane assessment value based on the preset total assessment value and intersection lane information, includes: calculating the average value of the surrounding lanes based on the preset total assessment value and the number of surrounding lanes to obtain the surrounding lane assessment value; and calculating the average value of the intersection lanes based on the preset total assessment value and the number of intersection lanes to obtain the intersection lane assessment value.

[0009] Further, determining at least one candidate lane from the surrounding lanes based on the surrounding lane assessment values ​​and the intersection lane assessment values ​​includes: sequentially assigning the intersection lane assessment values ​​to the surrounding lanes according to the assignment order and the surrounding lane assessment values ​​to obtain an assessment value mapping table, wherein the assessment value mapping table is used to record the assignment relationship between the assessment value of any intersection lane and the surrounding lanes; and determining at least one candidate lane from the surrounding lanes based on the assessment value mapping table and the intersection driving intention.

[0010] Further, based on the allocation order and the evaluation values ​​of surrounding lanes, the evaluation values ​​of intersection lanes are sequentially allocated to surrounding lanes to obtain an evaluation value mapping table. This includes: in response to an intersection lane evaluation value being greater than the surrounding lane evaluation values, the first sub-evaluation value of the first intersection lane is allocated to the first surrounding lane, where the first intersection lane is the first intersection lane traversed according to the allocation order, the first surrounding lane is the first surrounding lane traversed according to the allocation order, and the first sub-evaluation value is equal to the evaluation value of the first surrounding lane; according to the allocation order, the second sub-evaluation value of the first intersection lane is sequentially allocated to the subsequent adjacent surrounding lanes until the evaluation value of the first intersection lane is allocated, where the sum of the first sub-evaluation value and the second sub-evaluation value is equal to the evaluation value of the first intersection lane; after the evaluation value of the first intersection lane is allocated, the evaluation value of the next intersection lane is allocated according to the allocation order until the evaluation values ​​of all intersection lanes are allocated, thus obtaining the evaluation value mapping table.

[0011] Furthermore, based on the allocation order and the evaluation values ​​of surrounding lanes, the evaluation values ​​of intersection lanes are sequentially allocated to surrounding lanes to obtain an evaluation value mapping table. This includes: in response to an intersection lane evaluation value being less than the evaluation values ​​of surrounding lanes, the evaluation values ​​of multiple consecutive intersection lanes traversed in the allocation order are cumulatively allocated to the first surrounding lane until the evaluation values ​​of the first surrounding lane are filled; once the evaluation values ​​of the first surrounding lane are filled, the evaluation values ​​of the unallocated intersection lanes are continuously cumulatively allocated to the next surrounding lane in the allocation order until the evaluation values ​​of all surrounding lanes are filled, thus obtaining an evaluation value mapping table.

[0012] Furthermore, based on the allocation order and the evaluation values ​​of the surrounding lanes, the evaluation values ​​of the intersection lanes are sequentially allocated to the surrounding lanes to obtain the evaluation value mapping table. This includes: in response to the intersection lane evaluation value being equal to the surrounding lane evaluation value, the evaluation values ​​of the intersection lanes are allocated to the surrounding lanes one-to-one according to the allocation order to obtain the evaluation value mapping table.

[0013] Further, determining at least one candidate lane from the surrounding lanes based on the evaluation value mapping table and the intersection driving intention includes: determining the target intersection lane from the intersection lanes based on the intersection driving intention; determining the evaluation value allocation relationship of the target intersection lane based on the evaluation value mapping table, wherein the evaluation value allocation relationship is used to indicate one or more surrounding lanes to which the evaluation value of the target intersection lane is allocated; and determining at least one candidate lane based on the evaluation value allocation relationship, wherein at least one candidate lane carries the evaluation value allocated to the target intersection lane.

[0014] Further, determining the target lane from at least one candidate lane based on the traffic conditions of at least one candidate lane includes: acquiring the traffic conditions of at least one candidate lane through the vehicle's onboard perception system, wherein the traffic conditions include at least one of the following: vehicle density in the lane, relative speed of vehicles ahead, obstacle information, and construction sign information; calculating the safety score of at least one candidate lane based on the traffic conditions; and determining the target lane from at least one candidate lane based on the safety scores of at least one candidate lane and the evaluation value of the lane allocation at the target intersection.

[0015] According to another aspect of the embodiments of this application, a vehicle driving control device is also provided, comprising: an acquisition module, configured to acquire surrounding lane information at the vehicle's location and intersection lane information at the intersection in response to the vehicle driving to a preset distance from an intersection; a first determination module, configured to determine surrounding lane assessment values ​​based on preset assessment values ​​and surrounding lane information, and to determine intersection lane assessment values ​​based on preset total assessment values ​​and intersection lane information, wherein the surrounding lane assessment values ​​represent the assessment values ​​of any surrounding lane, and the intersection lane assessment values ​​represent the assessment values ​​of any intersection lane; a second determination module, configured to determine at least one candidate lane from the surrounding lanes based on the surrounding lane assessment values ​​and intersection lane assessment values; a third determination module, configured to determine a target lane from the at least one candidate lane based on the traffic state of the at least one candidate lane; and a control module, configured to control the vehicle to drive in the target lane.

[0016] According to another aspect of the embodiments of this application, a vehicle is also provided, including: a memory storing an executable program; and a processor for running the program, wherein the program executes the methods in various embodiments of this application when it runs.

[0017] According to another aspect of the embodiments of this application, a computer-readable storage medium is also provided, the computer-readable storage medium including a stored executable program, wherein, when the executable program is running, it controls the device where the computer-readable storage medium is located to perform the methods of various embodiments of this application.

[0018] According to another aspect of the embodiments of this application, a computer program product is also provided, including a computer program that, when executed by a processor, implements the methods of various embodiments of this application.

[0019] According to another aspect of the embodiments of this application, a computer program product is also provided, including a non-volatile computer-readable storage medium storing a computer program that, when executed by a processor, implements the methods in various embodiments of this application.

[0020] According to another aspect of the embodiments of this application, a computer program is also provided, which, when executed by a processor, implements the methods of the various embodiments of this application.

[0021] In this embodiment, a vehicle driving control method is adopted. In response to the vehicle reaching a preset distance from an intersection, information about the surrounding lanes at the vehicle's location and the intersection lane information are obtained. An evaluation value for the surrounding lanes is determined based on a preset evaluation value and the surrounding lane information. An evaluation value for the intersection lane is determined based on a preset total evaluation value and the intersection lane information. The surrounding lane evaluation value represents the evaluation value of any surrounding lane, and the intersection lane evaluation value represents the evaluation value of any intersection lane. At least one candidate lane is determined from the surrounding lanes based on the surrounding lane evaluation value and the intersection lane evaluation value. A target lane is determined from the at least one candidate lane based on the traffic state of the candidate lane. The vehicle is controlled to drive in the target lane. This achieves the goal of flexible lane selection decision-making at intersections where the number of lanes changes constantly, thus realizing the technical effect of adaptive lane selection decision-making for dynamic changes in the number of lanes at intersections. This solves the technical problem of how to flexibly select lanes at intersections where the number of lanes changes constantly. Attached Figure Description

[0022] The accompanying drawings, which are included to provide a further understanding of this application and form part of this application, illustrate exemplary embodiments of this application and are used to explain this application, but do not constitute an undue limitation of this application. In the drawings:

[0023] Figure 1 This is a flowchart of a vehicle driving control method according to an embodiment of this application;

[0024] Figure 2 This is a schematic diagram of an optional lane selection method before an intersection (three lanes becoming four lanes) according to an embodiment of this application;

[0025] Figure 3 This is a schematic diagram of an optional downsampling (three lanes to four lanes) allocation strategy according to an embodiment of this application;

[0026] Figure 4 This is a schematic diagram of an optional lane selection method before an intersection (four lanes becoming three lanes) according to an embodiment of this application;

[0027] Figure 5 This is a schematic diagram of an optional upsampling (four lanes to three lanes) allocation strategy according to an embodiment of this application;

[0028] Figure 6 This is a schematic diagram of an optional lane selection method before an intersection (five lanes becoming three lanes) according to an embodiment of this application;

[0029] Figure 7 This is a schematic diagram of an optional upsampling (five lanes to three lanes) allocation strategy according to an embodiment of this application;

[0030] Figure 8 This is a structural block diagram of a vehicle driving control device according to an embodiment of this application. Detailed Implementation

[0031] To enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present application, and not all embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative effort should fall within the scope of protection of the present application.

[0032] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of this application described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0033] In the field of autonomous driving, lane selection is a core element in realizing navigation intent. Navigation systems typically only provide instructions like "turn left 300 meters ahead," or at most, tell the driver to use the leftmost lane (at the intersection). However, the lane selection module needs to concretize this into pre-planning and sequence generation, as follows:

[0034] Plan ahead: When you are far from the intersection, plan ahead the lane you need to enter so that you can directly enter the left-turn lane.

[0035] Sequence generation: Plan a series of lane change sequences required to move from the current lane to the target left-turn lane (e.g., first change from the third left lane to the second left lane, and then to the dedicated left-turn lane), and make a comprehensive decision based on road traffic conditions.

[0036] Therefore, it can be understood that without precise route selection, navigation instructions cannot be executed, and the driving task (to the navigation destination) in autonomous driving will not be achieved.

[0037] In traditional solutions, autonomous driving systems typically employ high-definition maps or similar lightweight maps when dealing with lane selection problems at intersections where the number of lanes changes. This means the system relies entirely on the pre-defined lane connections within the high-definition map. The map explicitly states that "lane A can only connect to lane B before the intersection." This allows the vehicle to simply find and execute the pre-defined "unique correct" lane sequence based on the global path. Theoretically, this traditional approach is highly efficient, but it has a fatal flaw: if the pre-defined lane is temporarily occupied by parked vehicles, obstructed by construction, or if the road is redesigned due to roadwork, the autonomous driving system will fall into a "decision deadlock" due to this freshness issue. Furthermore, because high-definition maps have a long update cycle—ranging from monthly to semi-annual—this problem occurs frequently.

[0038] According to an embodiment of this application, a method for vehicle driving control is provided. It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions. Furthermore, although a logical order is shown in the flowchart, in some cases, the steps shown or described may be executed in a different order than that shown here.

[0039] This embodiment provides a vehicle driving control method. Figure 1 This is a flowchart of a vehicle driving control method according to an embodiment of this application, such as... Figure 1 As shown, the process includes the following steps:

[0040] Step S10: In response to the vehicle traveling to a preset distance from the intersection, obtain the surrounding lane information at the vehicle's location and the intersection lane information at the intersection.

[0041] In this embodiment, when the vehicle travels to a preset distance from the intersection, information about the surrounding lanes and the lanes at the intersection is obtained. The preset distance to the intersection refers to a fixed and configurable longitudinal distance (e.g., 150 to 300 meters) before the vehicle reaches the intersection, to ensure sufficient time to complete the lane selection decision.

[0042] The vehicle location refers to the current position of the vehicle on the road.

[0043] The location of the intersection area indicated by the navigation system at the intersection.

[0044] Therefore, by obtaining information about the surrounding lanes at the vehicle's location and the lane information at the intersection, a data foundation is provided for subsequent lane matching.

[0045] Step S11: Determine the surrounding lane assessment value based on the preset assessment value and surrounding lane information, and determine the intersection lane assessment value based on the preset total assessment value and intersection lane information, wherein the surrounding lane assessment value is used to represent the assessment value of any surrounding lane, and the intersection lane assessment value is used to represent the assessment value of any intersection lane.

[0046] In this embodiment, the vehicle driving control system determines the surrounding lane evaluation value based on a preset evaluation value and surrounding lane information, and determines the intersection lane evaluation value based on a preset total evaluation value and intersection lane information. The preset total evaluation value can be 60 points or other scores. The surrounding lane evaluation value represents the evaluation value of any surrounding lane, characterizing the adaptability of each lane to the navigation intent.

[0047] The intersection lane assessment value is used to represent the assessment value of any lane at an intersection, characterizing the quantitative score of each navigation target lane.

[0048] The preset total evaluation value is the total score of all lanes at the intersection corresponding to the navigation command, which is calculated by dividing the score equally based on the number of functions of the target lane at the intersection.

[0049] For example, using 60 points as a unified scoring benchmark, the total semantic weight of the intersection target lane functions (such as straight, left turn, right turn) provided by the navigation system is set to 60 points, and the evaluation value of each intersection lane is obtained by equally distributing the weights according to the number of target lanes (e.g., 15 points for each of the 4 lanes). Alternatively, using 100 points as a unified scoring benchmark, the total semantic weight of the intersection target lane functions (such as straight, left turn, right turn) provided by the navigation system is set to 100 points, and the evaluation value of each intersection lane is obtained by equally distributing the weights according to the number of target lanes (e.g., 25 points for each of the 4 lanes).

[0050] Therefore, by dynamically aligning the intersection quantitative score with the number of lanes, a score mapping relationship is formed between the two, realizing the quantitative transmission of navigation intent when the number of lanes is inconsistent, and providing an objective and calculable matching basis for candidate lane selection.

[0051] Step S12: Based on the surrounding lane assessment values ​​and the intersection lane assessment values, determine at least one candidate lane from the surrounding lanes.

[0052] In this embodiment, the vehicle driving control system determines at least one candidate lane from the surrounding lanes based on the surrounding lane evaluation values ​​and the intersection lane evaluation values. The candidate lane refers to an optional lane selected from the currently perceived lanes to achieve the navigation objective, based on the matching results of the surrounding lane evaluation values ​​and the intersection lane evaluation values.

[0053] As can be seen, the vehicle driving control system matches the accumulated semantic evaluation value of each surrounding lane (derived from a total weight of 60 points through up-and-down sampling) with the evaluation value of the target lane at the corresponding intersection, selecting at least one surrounding lane with an evaluation value higher than a threshold as a candidate lane to ensure that the lane selection result is consistent with the navigation intent within the scoring system. For example, the number of candidate lanes is determined according to the actual situation: one when the intersection lanes correspond one-to-one with each lane, two when changing from three lanes to four lanes, and three when changing from five lanes to three lanes.

[0054] For example, when the navigation target is a left turn (15 points), three lanes are detected. After being allocated 60 points, the middle lane gets 25 points and the rightmost lane gets 35 points. The system will use the middle lane and the rightmost lane with the highest evaluation value as candidate lanes, since both of them carry effective left turn semantic weights and the total score is derived from a unified 60-point benchmark, making them comparable.

[0055] Therefore, by using a unified scoring system to screen candidate lanes, the lane selection results can stably reflect the navigation intent and improve the reliability of lane selection.

[0056] Step S13: Determine the target lane from at least one candidate lane based on the traffic status of at least one candidate lane.

[0057] In this embodiment, the vehicle driving control system determines the target lane from at least one candidate lane based on the traffic conditions of at least one candidate lane. The target lane refers to the final driving lane selected from the candidate lanes after a comprehensive evaluation based on real-time traffic conditions and lane changes at intersections, and is the most suitable for executing the navigation intent.

[0058] As can be seen, based on the selected candidate lanes, the vehicle driving control system combines real-time perception data to weight and score the traffic status of each candidate lane, and selects the lane with both high semantic matching degree and low obstruction risk as the target lane, ensuring that the selected lane is both in line with the navigation intent and adapted to the real traffic environment.

[0059] For example, if the candidate lanes are lane 2 (evaluation value 3 points) and lane 3 (evaluation value 9 points), and there are temporarily parked vehicles in front of lane 3, the system will lower its passage score, while lane 2, although with a slightly lower semantic weight, has smooth passage. In the end, lane 2 is selected as the target lane, thus achieving a balance between semantic priority and safe passage.

[0060] The selection of the target lane takes into account both navigation semantics and real-time traffic conditions, thereby improving the stability and practicality of lane selection before intersections.

[0061] Step S14: Control the vehicle to drive in the target lane.

[0062] In this embodiment, the vehicle driving control system controls the vehicle to drive in the target lane.

[0063] As can be seen, based on the target lane information output by the previous steps, lane keeping or lane changing commands are output to the vehicle's control module, enabling the vehicle to smoothly enter and drive along the lane, ensuring that the driving trajectory is consistent with the preset lane selection result.

[0064] For example, when the target lane is determined to be the second lane of the current road (e.g., its 60-point semantic score includes 10 points of left-turn weight and there are no obstacles ahead), the vehicle automatically activates steering assistance to complete the lane change with comfortable acceleration and stay in that lane until entering the intersection.

[0065] Therefore, by directly linking lane selection decisions with control, vehicles can reliably locate their lanes before intersections even without high-precision maps, thus improving driving continuity and safety.

[0066] In summary, the vehicle driving control method of this application first acquires information on the surrounding lanes and the intersection lanes at the vehicle's location when the vehicle is at a preset distance from the intersection, providing a data foundation for subsequent lane matching. Secondly, it determines the surrounding lane evaluation value based on preset evaluation values ​​and the surrounding lane information, and determines the intersection lane evaluation value based on the preset total evaluation value and the intersection lane information, providing an objective and calculable matching basis for candidate lane selection. Then, based on the surrounding lane evaluation values ​​and the intersection lane evaluation value, at least one candidate lane is determined from the surrounding lanes. A unified scoring system is used to filter candidate lanes, ensuring that the lane selection result stably reflects the navigation intent and improving the reliability of lane selection. Subsequently, based on the traffic status of at least one candidate lane, the target lane is determined from at least one candidate lane, thereby improving the stability and practicality of lane selection before the intersection. Finally, the vehicle is controlled to drive in the target lane. By directly linking the lane selection decision with the execution control, the vehicle can reliably complete lane positioning before the intersection even without a high-precision map, improving driving continuity and safety.

[0067] The above steps of this application involve obtaining information about the surrounding lanes at the vehicle's location and the intersection lanes at the intersection in response to the vehicle reaching a preset distance from the intersection; determining the surrounding lane evaluation value based on a preset evaluation value and the surrounding lane information, and determining the intersection lane evaluation value based on a preset total evaluation value and the intersection lane information, wherein the surrounding lane evaluation value represents the evaluation value of any surrounding lane, and the intersection lane evaluation value represents the evaluation value of any intersection lane; determining at least one candidate lane from the surrounding lanes based on the surrounding lane evaluation value and the intersection lane evaluation value; determining the target lane from the at least one candidate lane based on the traffic state of the at least one candidate lane; and controlling the vehicle to travel in the target lane. This achieves the technical effect of adaptive lane selection decision-making in response to dynamic changes in the number of lanes at intersections, thereby solving the technical problem of how to flexibly select lanes at intersections where the number of lanes changes at any time.

[0068] Optionally, in step S10, obtaining the surrounding lane information at the vehicle's location and the intersection lane information at the intersection may include the following steps:

[0069] Step S101: Obtain surrounding lane information through the vehicle's onboard perception system, wherein the surrounding lane information includes at least the number of surrounding lanes at the vehicle's location.

[0070] Step S102: Obtain intersection lane information through the vehicle's in-vehicle navigation system. The intersection lane information includes at least the number of lanes at the intersection, the driving direction attribute of the lanes, and the driving intention at the intersection.

[0071] In this embodiment of the application, firstly, the vehicle driving control system obtains surrounding lane information through the vehicle's on-board perception system. The surrounding lane information includes at least the number of surrounding lanes at the vehicle's location, and may also include the spatial arrangement of surrounding lanes at the vehicle's location.

[0072] The vehicle perception system can be an on-board environmental perception module composed of cameras, radar or ultrasonic sensors, used to identify lane lines, drivable areas and obstacle distribution around the vehicle in real time.

[0073] It can be seen that the vehicle driving control system uses the on-board perception system to detect the lane line boundaries and drivable areas of the road ahead of the vehicle in real time, and extracts the actual number of lanes in the road segment where the vehicle is located at the current moment, as the physical input basis for subsequent lane topology matching.

[0074] For example, when the vehicle is about 200 meters away from the intersection, the forward-facing camera identifies that there are 3 clear and continuous lane lines on the current road. Based on this, the system determines that there are 3 lanes in the surrounding area, and that there are no temporary obstructions or construction signs in any of the lanes.

[0075] Therefore, the number of real road lanes around the vehicle can be dynamically obtained through vehicle-mounted sensing.

[0076] Secondly, the vehicle driving control system obtains intersection lane information through the vehicle's in-vehicle navigation system. The in-vehicle navigation system refers to the vehicle's built-in Standard Definition Map (SD) in-vehicle navigation system, which is used to provide route guidance information.

[0077] Intersection lane information includes at least the number of lanes at the intersection, the lane's direction of travel, and the intended direction of travel at the intersection. Intersection lane information is a semantic description of lane intentions provided by the navigation system based on pre-defined path planning. It represents the target path to the upcoming intersection and its corresponding possible directions, such as straight ahead, left turn, U-turn, or right turn, as provided by the vehicle's navigation SD map. It also indicates the desired direction of travel at this intersection during the current navigation session, such as a U-turn.

[0078] It can be seen that the vehicle driving control system reads the preset lane configuration and navigation intent of the intersection ahead through the SD vehicle navigation system, obtains the total number of lanes at the intersection, the functional attributes of each lane and the target direction of the current driving task, and uses this as the input basis for the semantic scoring system.

[0079] For example, the SD vehicle navigation system prompts "Turn left 300 meters ahead" and indicates that there are 4 lanes at the intersection, namely straight, straight, left turn, and right turn. Based on this, the system extracts that the number of lanes at the intersection is 4, the driving direction attributes are straight, straight, left turn, and right turn, and the driving intention at the intersection is "left turn".

[0080] Therefore, intent-driven route selection can be achieved based on the general intersection information of basic vehicle navigation systems, without the need for high-precision map support, thus lowering the threshold for system deployment.

[0081] Optionally, in step S11, determining the surrounding lane assessment value based on the preset total assessment value and the surrounding lane information, and determining the intersection lane assessment value based on the preset total assessment value and the intersection lane information may include the following execution steps:

[0082] Step S111: Calculate the average value of the surrounding lanes based on the preset total evaluation value and the number of surrounding lanes to obtain the evaluation value of the surrounding lanes.

[0083] Step S112: Calculate the average value of the intersection lanes based on the preset total evaluation value and the number of lanes at the intersection to obtain the intersection lane evaluation value.

[0084] In this embodiment, the average value of the surrounding lanes is first calculated based on the preset total evaluation value and the number of surrounding lanes to obtain the surrounding lane evaluation value. The surrounding lane evaluation value refers to the basic score of a single lane obtained by evenly distributing the total score to each currently perceived actual lane.

[0085] It can be seen that the vehicle driving control system divides the preset total evaluation value by the number of surrounding lanes obtained by perception, and calculates the initial score of each current lane under the condition of no semantic difference, which serves as the benchmark value for subsequent upgrade and downgrade sampling allocation.

[0086] For example, when the preset total evaluation value is 60 points, and it is perceived that there are 3 lanes around the current vehicle, the evaluation value of each lane is calculated as 20 points, which is used for subsequent matching and allocation with the target lane semantics.

[0087] Then, the average value is calculated based on the preset total evaluation value and the number of lanes at the intersection to obtain the intersection lane evaluation value. The intersection lane evaluation value refers to the basic score of a single target lane obtained by distributing the total score evenly to each lane at the intersection preset by the navigation system.

[0088] It can be seen that the vehicle driving control system divides the preset total evaluation value by the number of intersection lanes provided by the vehicle navigation system to calculate the basic semantic score of each target lane under ideal conditions, which serves as the semantic benchmark for subsequent matching with the actual lane.

[0089] For example, if the preset total evaluation value is 60 points and the navigation shows that there are 4 lanes at the intersection (straight, straight, left turn, right turn), then the evaluation value of each target lane is 15 points (60 ÷ 4), which is used for semantic mapping with the perceived 3 actual lanes.

[0090] Therefore, by establishing a fair initial scoring starting point through mean distribution, we can ensure consistency between semantics and physical structure in the upgrading and downgrading sampling process, and improve the interpretability of the routing logic.

[0091] Optionally, in step S12, determining at least one candidate lane from the surrounding lanes based on the surrounding lane evaluation values ​​and the intersection lane evaluation values ​​may include the following execution steps:

[0092] Step S121: Based on the allocation order and the evaluation values ​​of the surrounding lanes, the evaluation values ​​of the intersection lanes are sequentially allocated to the surrounding lanes to obtain an evaluation value mapping table. The evaluation value mapping table is used to record the allocation relationship between the evaluation value of any intersection lane and the surrounding lanes.

[0093] Step S122: Based on the evaluation value mapping table and the driving intention at the intersection, determine at least one candidate lane from the surrounding lanes.

[0094] In this embodiment of the application, firstly, the vehicle driving control system allocates the intersection lane evaluation values ​​to the surrounding lanes in sequence according to the allocation order and the surrounding lane evaluation values, and obtains an evaluation value mapping table. The evaluation value mapping table is used to record the allocation relationship between the evaluation value of any intersection lane and the surrounding lanes. It is a corresponding table that records how the semantic score of each navigation target lane is allocated to the actual perceived lane, and is used to quantify the matching weight between semantic intent and physical lane.

[0095] The allocation order refers to the processing sequence in which the target lane score is transferred to the adjacent actual lanes one by one from left to right or from right to left, ensuring the consistency of spatial location.

[0096] It can be seen that the vehicle driving control system distributes the evaluation value of each lane at the intersection to the adjacent surrounding lanes in a left-to-right or right-to-left order until the total score is exhausted, forming a mapping table that records the source and distribution of scores for each target lane.

[0097] For example, when the intersection lane assessment value is 15 points (left turn) and the surrounding lane assessment value is 20 points, and the allocation order is from left to right, 10 points of the 15 points of the left turn lane are allocated to the first surrounding lane and 5 points are allocated to the second surrounding lane, forming a mapping record of left turn [Lane 1: 10 points, Lane 2: 5 points].

[0098] Therefore, by establishing clear semantic and physical mapping relationships through orderly allocation, the route selection decision relies on quantitative allocation logic, avoiding the uncertainty brought about by blind matching.

[0099] Then, the vehicle driving control system determines at least one candidate lane from the surrounding lanes based on the evaluation value mapping table and the driving intention at the intersection.

[0100] As can be seen, the vehicle driving control system extracts the mapping results corresponding to the intersection target lane and the surrounding actual lanes based on the score allocation relationship recorded in the evaluation value mapping relationship table, combined with the intersection driving intention specified by the current navigation (such as left turn), and selects one or more surrounding lanes with the highest score for the intention as candidate lanes.

[0101] Specifically, the number of candidate lanes is determined based on the actual situation. When the intersection lanes correspond one-to-one with each lane, there is one lane; when three lanes become four lanes, there may be two lanes; and when five lanes become three lanes, there may be three lanes.

[0102] For example, if the driving intention at the intersection is "left turn", and the evaluation value mapping table shows that 10 out of the 15 points for the left turn intention are allocated to the first surrounding lane and 5 points are allocated to the second lane, then the system will list the first and second lanes as candidate lanes, with the second lane having the highest score.

[0103] Therefore, by accurately locating the actual lane that best matches the navigation intent based on semantic mapping, the consistency between the lane selection result and the driving intent is improved.

[0104] Optionally, in step S121, the process of sequentially assigning intersection lane evaluation values ​​to surrounding lanes according to the assignment order and surrounding lane evaluation values ​​to obtain an evaluation value mapping table may include the following steps:

[0105] In step S1211, in response to the intersection lane evaluation value being greater than the surrounding lane evaluation value, the first sub-evaluation value of the first intersection lane is assigned to the first surrounding lane, wherein the first intersection lane is the first intersection lane traversed in the assignment order, the first surrounding lane is the first surrounding lane traversed in the assignment order, and the first sub-evaluation value is equal to the evaluation value of the first surrounding lane.

[0106] Step S1212: According to the allocation order, the second sub-evaluation value of the first intersection lane is sequentially allocated to the subsequent adjacent surrounding lanes until the evaluation value of the first intersection lane is allocated. The sum of the first sub-evaluation value and the second sub-evaluation value is equal to the evaluation value of the first intersection lane.

[0107] Step S1213: After the evaluation values ​​of the lanes at the first intersection are allocated, the evaluation values ​​of the lanes at the next intersection are allocated in the allocation order until the evaluation values ​​of all lanes at all intersections are allocated, and an evaluation value mapping table is obtained.

[0108] In this embodiment of the application, firstly, when the intersection lane evaluation value is greater than the surrounding lane evaluation value, it is the upsampling mode of this application. The vehicle driving control system allocates the first sub-evaluation value of the first intersection lane to the first surrounding lane. The first intersection lane is the first intersection lane traversed in the allocation order, the first surrounding lane is the first surrounding lane traversed in the allocation order, and the first sub-evaluation value is equal to the evaluation value of the first surrounding lane.

[0109] The first sub-evaluation value refers to the fact that during the allocation process, when the target lane evaluation value is higher than the current actual lane evaluation value, the latter will be given full consideration as the transfer share to ensure that the baseline weight of the physical lane is not diluted.

[0110] It can be seen that in the allocation order from left to right, when the evaluation value of the target lane at the first intersection is higher than the evaluation value of the first surrounding actual lane, the vehicle driving control system transfers all the evaluation values ​​of the actual lane as sub-evaluation values ​​to it.

[0111] For example, if the evaluation value of the first intersection lane (such as a turning lane) is 15 points and the evaluation value of the first surrounding lane is 20 points, this rule is not triggered because 15 < 20. If the evaluation value of the first intersection lane is 25 points and the evaluation value of the surrounding lane is 20 points, then the 20 points will be fully allocated to the surrounding lane as the first sub-evaluation value.

[0112] Therefore, by prioritizing the acceptance of the complete baseline score of the actual lanes, the weight of the physical lanes in the allocation is maintained, thus avoiding scoring distortion.

[0113] Then, the vehicle driving control system distributes the second sub-evaluation value of the first intersection lane to the subsequent adjacent surrounding lanes in the allocation order until the evaluation value of the first intersection lane is allocated. The sum of the first sub-evaluation value and the second sub-evaluation value is equal to the evaluation value of the first intersection lane.

[0114] The second sub-assessment value refers to the portion of the remaining unassigned assessment value of the lane at the first intersection after the first sub-assessment value has been received, which needs to be transferred to the surrounding lanes in sequence.

[0115] As can be seen, after the first sub-evaluation value has been assigned to the first surrounding lane, the vehicle driving control system will assign the remaining evaluation values ​​(i.e., the total evaluation value minus the first sub-evaluation value) of the first intersection lane to the subsequent adjacent surrounding lanes in order from left to right, until all evaluation values ​​of the target lane have been assigned.

[0116] For example, if the first lane (left turn) at the intersection is evaluated as 25 points and the first surrounding lane is evaluated as 20 points, then the first sub-evaluation is 20 points, and the remaining second sub-evaluation is 5 points. These 5 points will be allocated to the second surrounding lane, at which point the second lane receives 5 points, and the allocation of the 25 points for the left turn intention is complete.

[0117] Therefore, by sequentially allocating residual values, the target intent is distributed in a reasonable attenuation pattern within the lane sequence, thereby improving the smoothness and rationality of lane selection.

[0118] Finally, after the evaluation values ​​for the first intersection lanes have been allocated, the vehicle driving control system allocates the evaluation values ​​for the next intersection lanes in the allocation order, until the evaluation values ​​for all intersection lanes have been allocated, resulting in an evaluation value mapping table. This evaluation value mapping table records the weight distribution comparison table formed after the evaluation values ​​of all target lanes at each intersection are allocated to the actual perceived lanes in the allocation order, reflecting the mapping relationship between navigation intent and physical lanes.

[0119] It can be seen that after the evaluation value allocation of the current intersection lane is completed, the vehicle driving control system continues to process the next target lane according to the allocation order, and repeats the sub-evaluation value acceptance and residual value continuation allocation process in turn until the evaluation values ​​of all navigation target lanes are mapped and a complete evaluation value mapping relationship table is formed.

[0120] Therefore, by sequentially covering and allocating lanes, it is ensured that all navigation intentions are systematically mapped to the actual lanes, thereby improving the integrity and consistency of lane selection in multi-intention scenarios.

[0121] Optionally, in step S121, the process of sequentially assigning intersection lane evaluation values ​​to surrounding lanes according to the assignment order and surrounding lane evaluation values ​​to obtain an evaluation value mapping table may include the following steps:

[0122] In step S1214, in response to the intersection lane assessment value being less than the surrounding lane assessment value, the assessment values ​​of multiple consecutive intersection lanes traversed in the allocation order are accumulated and allocated to the first surrounding lane until the assessment value of the first surrounding lane is filled.

[0123] Step S1215: After the evaluation values ​​of the first surrounding lane are filled, continue to allocate the evaluation values ​​of the unallocated intersection lanes to the next surrounding lane in the allocation order until the evaluation values ​​of all surrounding lanes are filled, and obtain the evaluation value mapping table.

[0124] In this embodiment, firstly, when the evaluation value of the intersection lane is less than the evaluation value of the surrounding lanes, it is the downsampling mode of this application. The vehicle driving control system will accumulate and allocate the evaluation values ​​of multiple consecutive intersection lanes in the allocation order to the first surrounding lane until the evaluation value of the first surrounding lane is filled.

[0125] The cumulative allocation refers to the process where, when the assessment value of a single intersection lane is insufficient to fill the assessment values ​​of the surrounding lanes, the assessment values ​​of multiple subsequent intersection lanes are sequentially accumulated and allocated to the surrounding lanes until their assessment values ​​are completely filled.

[0126] It can be seen that when the lane evaluation value at the intersection is less than the current evaluation value of the surrounding lanes, the vehicle driving control system takes the evaluation values ​​of the subsequent lanes at the intersection in sequence from the allocation order, adds them up, and allocates them all to the current first surrounding lane until its evaluation value is completely filled, and then continues to process the next surrounding lane.

[0127] For example, if the first surrounding lane is evaluated as 20 points, while the first intersection lane (left turn) is only 15 points and the second intersection lane (straight 1) is 15 points, then the first 20 points out of the 30 points for left turn and straight 1 are allocated to the first surrounding lane, and the remaining 10 points go into the next round of allocation.

[0128] Therefore, by accumulating multiple intent evaluation values ​​to fill a single physical lane, the semantic intent can still be fully mapped when the number of lanes is reduced, thus improving robustness in lane compression scenarios.

[0129] Secondly, after the evaluation values ​​of the first surrounding lanes are filled, the vehicle driving control system continues to allocate the evaluation values ​​of the unassigned intersection lanes to the next surrounding lanes in the allocation order, until the evaluation values ​​of all surrounding lanes are filled, thus obtaining the evaluation value mapping table.

[0130] As can be seen, after the first surrounding lane assessment value is filled, the vehicle driving control system continues to accumulate the remaining unassigned intersection lane assessment values ​​from left to right and assign them to the next surrounding lane in sequence, repeating the filling process until all surrounding lanes have completed the assessment value carrying, forming the final mapping relationship table.

[0131] For example, if the first surrounding lane is filled with 30 points for left turn (15 points) and straight lane 1 (15 points), the remaining 30 points for straight lane 2 (15 points) and right turn (15 points) will be allocated to the second surrounding lane in sequence. If its evaluation value is 25 points, it will be filled with 15 points from straight lane 2 plus 10 points from right turn, and the remaining 5 points will be reserved for subsequent allocation until all surrounding lanes are filled.

[0132] Therefore, by continuously accumulating and filling lanes one by one, the intention evaluation value can still be completely distributed when the number of lanes increases, thus improving the adaptability and stability of lane selection in expanded scenarios.

[0133] Optionally, in step S121, the vehicle driving control system sequentially assigns the intersection lane evaluation values ​​to the surrounding lanes according to the allocation order and the surrounding lane evaluation values ​​to obtain the evaluation value mapping table, which may include the following execution steps:

[0134] Step S1216: In response to the intersection lane evaluation value being equal to the surrounding lane evaluation values, the evaluation values ​​of the intersection lanes are assigned to the surrounding lanes one by one according to the assignment order to obtain the evaluation value mapping table.

[0135] In this embodiment of the application, when the evaluation value of the intersection lane is equal to the evaluation value of the surrounding lanes, the vehicle driving control system allocates the evaluation values ​​of the intersection lanes to the surrounding lanes one by one according to the allocation order, and obtains the evaluation value mapping table.

[0136] It can be seen that when the number of lanes at the intersection is equal to the number of lanes around it and the evaluation values ​​of each lane are the same, the vehicle driving control system can directly assign the evaluation value of each target lane to the sensing lane at the same location in a left-to-right allocation order, forming a unique and definite evaluation value mapping table.

[0137] For example, if the navigation indicates 4 target lanes (15 points each) and the perception also detects 4 actual lanes (15 points each), then the evaluation values ​​for left turn correspond to lane 1, straight 1 corresponds to lane 2, straight 2 corresponds to lane 3, and right turn corresponds to lane 4. The evaluation values ​​are directly matched one by one without adjustment.

[0138] Thus, in a stable scenario where the number of lanes remains constant, precise and lossless intent transmission is achieved, ensuring the certainty of lane selection decisions and execution efficiency.

[0139] Optionally, in step S122, determining at least one candidate lane from the surrounding lanes based on the evaluation value mapping table and the intersection driving intention may include the following execution steps:

[0140] Step S1221: Determine the target intersection lane from the intersection lanes based on the driving intention at the intersection.

[0141] Step S1222: Determine the evaluation value allocation relationship of the target intersection lanes according to the evaluation value mapping relationship table, wherein the evaluation value allocation relationship is used to indicate the one or more surrounding lanes to which the evaluation value of the target intersection lane is allocated.

[0142] Step S1223: Determine at least one candidate lane based on the evaluation value allocation relationship, wherein at least one candidate lane carries the evaluation value of the lane allocation at the target intersection.

[0143] In this embodiment, the vehicle driving control system first determines the target intersection lane from the intersection lanes based on the driving intention at the intersection. The target intersection lane refers to the lane required for the driving task, or it can refer to a designated lane matched from all preset lanes at the intersection based on the driving intention (such as left turn, right turn, or straight ahead) provided by the vehicle navigation system, used to complete the turning action.

[0144] It can be seen that the vehicle driving control system analyzes the driving intention clearly stated in the SD navigation command, selects the lane directly related to the intention from the intersection topology, and uses it as the starting target lane for subsequent evaluation value allocation.

[0145] For example, if the navigation instruction is "turn left", the system identifies the lane marked "dedicated to left turn" or "optional to left turn" from the intersection lane information and sets it as the target intersection lane.

[0146] This allows for the precise identification of the lane corresponding to the navigation intent, providing a clear starting point for subsequent lane assessment value allocation.

[0147] Then, the vehicle driving control system determines the evaluation value allocation relationship of the target intersection lane according to the evaluation value mapping relationship table. The evaluation value allocation relationship is used to indicate the one or more surrounding lanes to which the evaluation value of the target intersection lane is allocated. The evaluation value allocation relationship can refer to the distribution record of the semantic evaluation value of the target intersection lane on the actual perceived lane, and clarify which one or more adjacent lanes and their weight ratio are allocated to the evaluation value corresponding to the intention.

[0148] It can be seen that the vehicle driving control system, based on the generated evaluation value mapping table, looks up the surrounding lanes corresponding to the target intersection lane during the allocation process and their share of evaluation value, thus forming the specific carrying path of the intention in the physical space.

[0149] For example, if the target intersection lane is "left turn" (evaluation value 15 points), and the mapping table shows that 10 points are allocated to lane 1 and 5 points are allocated to lane 2, then the evaluation value allocation relationship is "left turn: lane 2: 10 points, lane 3: 5 points".

[0150] This clarifies the distribution of target intent on the actual lanes, providing an actionable basis for lane selection in subsequent trajectory planning and improving the interpretability and stability of the decision.

[0151] Finally, the vehicle driving control system determines at least one candidate lane based on the evaluation value allocation relationship. At least one candidate lane carries the evaluation value of the lane allocation at the target intersection, and its evaluation value weight reflects the degree of matching with the target intent.

[0152] It can be seen that the vehicle driving control system extracts all surrounding lanes that receive the target intersection lane evaluation value as candidate lanes based on the evaluation value allocation relationship, and uses them for priority ranking and selection in subsequent trajectory planning.

[0153] For example, if 15 points of the left-turn target lane are allocated to lane 1 (10 points) and lane 2 (5 points), then lanes 1 and 2 are identified as candidate lanes, with lane 1 having a higher weight.

[0154] Therefore, by using evaluation value weights to filter out truly passable candidate lanes, blind lane selection is avoided, and the accuracy and safety of lane selection before intersections are improved.

[0155] Optionally, in step S13, determining the target lane from at least one candidate lane based on the traffic state of at least one candidate lane may include the following steps:

[0156] Step S131: Obtain the traffic status of at least one candidate lane through the vehicle's onboard perception system, wherein the traffic status includes at least one of the following: vehicle density in the lane, relative speed of vehicles ahead, obstacle information, and construction sign information.

[0157] Step S132: Calculate the safety score of at least one candidate lane based on the traffic conditions.

[0158] Step S133: Based on the safety score of at least one candidate lane and the evaluation value of the lane allocation at the target intersection, determine the target lane from at least one candidate lane.

[0159] In this embodiment of the application, firstly, the vehicle driving control system obtains the traffic status of at least one candidate lane through the vehicle's on-board perception system, wherein the traffic status includes at least one of the following: vehicle density in the lane, relative speed of vehicles ahead, obstacle information, and construction sign information.

[0160] It can be seen that the vehicle driving control system uses the vehicle's cameras, radar and other sensing devices to monitor environmental information such as vehicle density in the candidate lane, relative speed of vehicles ahead, and whether there are obstacles or construction signs in real time, as a supplementary basis for lane selection decision.

[0161] For example, if the candidate lanes are lanes 2 and 3, the perception system detects that there are slow-moving vehicles (relative speed <5km / h) and temporary construction signs ahead of lane 2, while lane 3 is unobstructed.

[0162] Then, the vehicle driving control system calculates a safety score for at least one candidate lane based on traffic conditions. The safety score is a numerical indicator that quantifies the current driving safety of the candidate lane based on real-time traffic conditions, taking into account factors such as vehicle density, relative speed ahead, obstacles, and construction signs.

[0163] It can be seen that the vehicle driving control system calculates the weights of each influencing factor based on the traffic state data obtained by the perception system and preset weight rules, and outputs the safety score corresponding to each candidate lane for ranking and optimization.

[0164] For example, if lane 2 has high vehicle density, low relative speed ahead, and construction signs, it scores 6 points. Lane 3 has no obstacles, sufficient distance between vehicles, and no construction information, so it scores 9 points. Therefore, lane 3 has a higher safety score.

[0165] Therefore, by quantifying safety risks to assist in lane selection, the decision-making process becomes closer to actual road conditions, improving the stability and risk avoidance capabilities of the driving process.

[0166] Finally, based on the safety score of at least one candidate lane and the evaluation value of the lane allocation at the target intersection, the vehicle driving control system determines the target lane from at least one candidate lane. The target lane is the optimal driving lane selected by weighting or accumulating the evaluation values ​​and safety scores of the candidate lanes.

[0167] As can be seen, the vehicle driving control system weights and integrates the evaluation value and safety score of each candidate lane, and selects the lane with the highest comprehensive score as the final target lane for execution by the downstream trajectory planning module.

[0168] For example, if lane 1 has an evaluation score of 10 and a safety score of 6, and lane 2 has an evaluation score of 5 and a safety score of 9, assuming the weights of the evaluation score and the safety score are 0.3 and 0.7 respectively, after weighted calculation, lane 1 has a comprehensive score of 7.2 and lane 2 has a comprehensive score of 7.8. Lane 2 has a higher comprehensive score and is selected as the target lane.

[0169] Therefore, by taking into account both navigation intent and actual road conditions, the optimal lane selection is made, improving the rationality and reliability of driving decisions.

[0170] In summary, this application achieves accurate and safe lane selection at complex intersections without the need for high-precision maps by sensing the real-time traffic status of candidate lanes, combining navigation intent with lane topology upgrade and downgrade matching evaluation, and integrating safety scores for comprehensive weighted decision-making.

[0171] This application also provides a vehicle driving control method that can not only flexibly handle the lane selection decision problem caused by the constant change of the number of lanes in front of the intersection, but also get rid of the dependence on expensive and slow-updating high-precision maps, and is based entirely on vision and traditional SD vehicle navigation.

[0172] This application also discloses a vehicle driving control method aimed at solving the problem of high uncertainty in road selection before path planning caused by lane number changes ("lane management") in intersection areas for autonomous vehicles. The core of this vehicle driving control method is as follows: When approaching an intersection, the vehicle first obtains the current number of roads and its own position through perception, and simultaneously obtains the number of lanes at the intersection ahead and the driving direction of each lane (straight, left turn, U-turn, or right turn) through the SD vehicle navigation system. Then, a "progressive sampling" strategy is introduced to match the vehicle's target lane path with the current lane topology sequence, identifying all potential passable lane sequences. These potential sequences are dynamically scored and ranked. Finally, the vehicle selects the optimal or second-best lane based on the scoring results, providing the downstream trajectory / path planning module with a smooth, safe, and traffic-efficient trajectory. This application improves the reliability and safety of autonomous driving systems in making decisions at complex intersections without high-precision maps.

[0173] Therefore, this application can address the issue of vehicle lane selection before intersections. By introducing a "leveling up and down sampling" strategy, under normal circumstances where the lane markings at the intersection remain unchanged, it can also accommodate adverse situations such as an increase or decrease in the number of lanes before the intersection. It matches the target lane path input by the vehicle's SD in-vehicle navigation system with the current lane topology sequence detected by visual perception, identifying all potential passable lane sequences. A scoring mechanism selects the optimal or second-best lane, which is then used by the downstream trajectory / path planning module to generate a smooth, safe, and traffic-efficient trajectory. This provides drivers with more efficient and safer driving assistance.

[0174] In this embodiment, it is necessary to sense and detect the current road lane situation, such as the number of drivable lanes. Simultaneously, the vehicle navigation SD map needs to be able to provide the target path to the upcoming intersection and its corresponding drivable direction, such as: straight ahead, left turn, U-turn, or right turn, and also provide the desired drivable direction for the current navigation at that intersection, such as a U-turn.

[0175] Supplementary concept: Upgrade and downgrade sampling. The upgrade and downgrade in this application refer to the number of lanes at the intersection provided by the SD vehicle navigation system relative to the current number of lanes. If the number of lanes at the intersection provided by the SD system is greater than the current number of lanes (e.g., ...), ... Figure 2 As shown, if a three-lane road becomes a four-lane road, it's a downgrade, meaning: based on the number of lanes in the navigation, all current lanes are sampled to determine the optimal driving lane. If the number of lanes at the intersection given by SD is less than the current number of lanes (e.g., ... Figure 4 As shown, the change from five lanes to four lanes is an upgrade, meaning that the current lanes are sampled based on the number of navigation lanes to determine the optimal driving lane.

[0176] Figure 2This is a schematic diagram of an optional lane selection method before an intersection (three lanes becoming four lanes) according to an embodiment of this application, as shown below. Figure 2 The following explains the downsampling scheme. Assume the SD navigation indicates a road ahead leading to intersection 2 (straight ahead), with four lanes: left turn / U-turn, straight ahead 1, straight ahead 2, and right turn. The system detects three surrounding lanes, labeled 1, 2, and 3. The following describes which lane to take using "downsampling" to reach the straight ahead 2 lane specified by the navigation.

[0177] Figure 3 This is a schematic diagram of an optional downsampling (three lanes to four lanes) allocation strategy according to an embodiment of this application, as shown below. Figure 3 As shown, the total score for all lanes at the upcoming intersection provided by SD navigation and all surrounding lanes detected by perception is set to 60 points. This total score is then evenly distributed among the lanes at each intersection. Therefore, the left-turn, straight-ahead 1, straight-ahead 2, and right-turn lanes each receive 15 points. The surrounding lanes detected by perception each receive 20 points. Now, using the allocation strategy, we analyze the source of the surrounding lanes' scores from left to right:

[0178] Lane 1, being the far left, receives 20 points, which is more than the 15 points for the leftmost lane in the navigation system. Therefore, its points come from multiple lanes: 15 points from left turns and 5 points from going straight (lane 1). The remaining 10 points from going straight (lane 1) will be allocated to lane 2, which follows directly in the same order. Similarly, lane 2 will receive 10 points from going straight (lane 1). The remaining 5 points from going straight (lane 1) will then be allocated to lane 3. Since the target lane and the current lane have the same total score, the remaining 15 points for the right turn lane will be allocated to lane 3. This completes the lane allocation strategy based on downsampling.

[0179] Based on the above allocation strategy, the target straight lane (for the driving task) is lane 2. Lanes 2 and 3 are both candidate lanes. Most of the points (10 points) are given to lane 2. Therefore, lane 2 is the best candidate lane before the intersection, while lane 3, which receives a small portion (5 points), is the second best candidate lane.

[0180] Figure 4 This is a schematic diagram of an optional lane selection method before an intersection (four lanes becoming three lanes) according to an embodiment of this application, as shown below. Figure 4 As shown, one upsampling scheme of this application is as follows:

[0181] Suppose that the SD navigation indicates a left turn is required at the upcoming intersection, and there are four lanes at the intersection: left turn / U-turn, straight ahead 1, straight ahead 2, and right turn. The system detects five surrounding lanes, labeled 1, 2, 3, 4, and 5. The following describes which lane to take through "upsampling" to reach the left turn lane specified by the navigation at the intersection.

[0182] Figure 5 This is a schematic diagram of an optional upsampling (four lanes to three lanes) allocation strategy according to an embodiment of this application, as shown below. Figure 5 As shown, with a total score of 60 points: the left turn / U-turn, straight ahead 1, straight ahead 2, and right turn lanes at the intersection still receive 15 points, while the surrounding lanes 1-5 each receive 12 points. Similarly, using the allocation strategy, we now analyze the source of points for the surrounding lanes from left to right:

[0183] Lane 1 has 12 points, fewer than the left-turn lane, so all 12 points can be understood as coming from left turns. The remaining 3 points from left turns will be allocated to Lane 2, which will then take its 9 points from Lane 1 (straight ahead). Lane 1 (straight ahead) will also have 6 points remaining, allocated to Lane 3; the same applies to Lane 3. Lane 4, after receiving its remaining 9 points from Lane 2 (straight ahead), will take its remaining 3 points from the right-turn lane. Similarly, since the target lane and the current lane have the same total score, the remaining 12 points from the right-turn lane will be allocated to Lane 5. This completes the lane allocation strategy based on upsampling.

[0184] Therefore, the target left-turn lane (as required by the driving task) is scored into 12 points and 3 points, which are assigned to lane 1 and lane 2 respectively. It can be seen that lane 1 and lane 2 are both candidate lanes. Lane 1, which has obtained most of the points for the target left-turn lane, is the optimal candidate lane, while lane 3 is the second-best candidate lane.

[0185] Figure 6 This is a schematic diagram of an optional lane selection method before an intersection (five lanes to three lanes) according to an embodiment of this application; as shown Figure 6 As shown, another upsampling scheme of this application is as follows:

[0186] Suppose that the SD navigation indicates an upcoming intersection where you need to go straight, and there are 3 lanes at the intersection: left turn / U-turn, straight ahead, and right turn. The system detects 5 surrounding lanes, labeled 1, 2, 3, 4, and 5. The following describes which lane to take through "upsampling" to reach the straight-ahead lane specified by the navigation.

[0187] Figure 7 This is a schematic diagram of an optional upsampling (five lanes to three lanes) allocation strategy according to an embodiment of this application; as shown Figure 7 As shown, with a total score of 60 points: the left turn / U-turn, straight, and right turn lanes at the intersection each receive 20 points, while the surrounding lanes (lanes 1-5) each receive 12 points. Similarly, we now analyze the source of points for the surrounding lanes from left to right using the allocation strategy:

[0188] Lane 1 has 12 points, fewer than the left-turn lane, so these 12 points can be understood as all coming from left turns. The remaining 8 points from left turns will be allocated to Lane 2, which will then take its 4 points from the straight-ahead lane. The remaining 16 points from the straight-ahead lane will be allocated to Lanes 3 and 4. Lane 3 will receive all 12 points from the straight-ahead lane, while Lane 4, after receiving its remaining 4 points from the straight-ahead lane, will take its remaining 8 points from the right-turn lane. Similarly, since the target lane and the current lane have the same total score, the remaining 12 points from the right-turn lane will be allocated to Lane 5. This completes the lane allocation strategy based on upsampling.

[0189] Therefore, the target straight lane (required for the driving task) is scored into 4 points, 12 points and 4 points, respectively, and assigned to lane 2, lane 3 and lane 4. It can be seen that lane 2, lane 3 and lane 4 are all candidate lanes. Lane 3, which has obtained most of the points for the target left-turn lane, is the best candidate lane, while lane 2 and lane 4 are the second best candidate lanes.

[0190] When the number of lanes remains unchanged, since the current lane and the target lane are in one-to-one correspondence, the current lane will receive all the points of the target lane, and thus will naturally be the optimal candidate lane.

[0191] Using the lane topology upgrade and downgrade sampling strategy described above, drivers will be able to obtain optimal and suboptimal candidate lane sequences before each intersection. Next, based on the real-time traffic conditions (such as congestion, obstacles, and temporary road closures) of the candidate lanes containing the optimal and suboptimal sequences, corresponding scoring weighting factors can be designed to flexibly and efficiently obtain the optimal lane in real time, ensuring more accurate and stable driving assistance for drivers.

[0192] For example, this embodiment introduces a multi-dimensional comprehensive scoring mechanism based on the candidate lane sequence: the vehicle driving control system obtains traffic situation information of each candidate lane in real time, such as the density of adjacent vehicles, the existence status of obstacles, the recognition result of temporary lane closure signs, and the relative motion state of the vehicle and surrounding vehicles, and calculates two indicators for each candidate lane: "evaluation value of lane allocation at target intersection" and "safety score".

[0193] The safety score is determined by the perception module based on features such as obstacle distance, vehicle speed adaptability, and lane change conflict probability, with a value ranging from 0 to 10. The system assigns fixed weight coefficients to both, with the evaluation value having a weight of 0.3 and the safety score having a weight of 0.7, indicating that traffic safety is given higher priority in intersection scenarios.

[0194] Taking a specific scenario as an example, assuming the candidate lanes include two lanes, No. 1 and No. 2, after upsampling and downsampling, the navigation intent matching evaluation value of lane No. 1 is 10 points, indicating that it is highly consistent with the left-turn target, but there is a stationary obstacle nearby, so the safety score is 6 points; the evaluation value of lane No. 2 is 5 points, because it deviates slightly from the ideal left-turn lane, but there are no obstructions around and the traffic flow is smooth, so the safety score is 9 points; the system performs a weighted fusion calculation on the two, and the comprehensive score of lane No. 1 is 10×0.3+6×0.7=7.2 points, and the comprehensive score of lane No. 2 is 5×0.3+9×0.7=7.8 points. After comparison, the comprehensive score of lane No. 2 is higher, so the system determines it as the current optimal target lane and outputs the lane information to the downstream trajectory planning module, which generates a safe lane change and driving trajectory for obstacle avoidance.

[0195] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, data stored, data displayed, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties. Furthermore, the collection, use and processing of the relevant data must comply with the relevant laws, regulations and standards of the relevant countries and regions, and corresponding operation entry points are provided for users to choose to authorize or refuse.

[0196] According to an embodiment of this application, a device for vehicle driving control is provided. It should be noted that the device can be used to execute the above-described vehicle driving control method.

[0197] Figure 8 This is a structural block diagram of a vehicle driving control device according to one embodiment of this application, such as... Figure 8 As shown, a vehicle driving control device 800 is used as an example. This device includes: an acquisition module 801, used to acquire surrounding lane information at the vehicle's location and intersection lane information at the intersection in response to the vehicle traveling to a preset distance from an intersection; a first determination module 802, used to determine surrounding lane assessment values ​​based on preset assessment values ​​and surrounding lane information, and to determine intersection lane assessment values ​​based on preset total assessment values ​​and intersection lane information, wherein the surrounding lane assessment values ​​represent the assessment values ​​of any surrounding lane, and the intersection lane assessment values ​​represent the assessment values ​​of any intersection lane; a second determination module 803, used to determine at least one candidate lane from the surrounding lanes based on the surrounding lane assessment values ​​and intersection lane assessment values; a third determination module 804, used to determine a target lane from the at least one candidate lane based on the traffic state of the at least one candidate lane; and a control module 805, used to control the vehicle to travel in the target lane.

[0198] Embodiments of this application also provide a vehicle, including: a memory storing an executable program; and a processor for running the program, wherein the program executes the methods described in various embodiments of this application when it runs.

[0199] Embodiments of this application also provide a computer-readable storage medium including a stored executable program, wherein, when the executable program is running, it controls the device where the computer-readable storage medium is located to perform the methods of various embodiments of this application.

[0200] Embodiments of this application also provide a computer program product, including a computer program that, when executed by a processor, implements the methods of various embodiments of this application.

[0201] Embodiments of this application also provide a computer program product, including a non-volatile computer-readable storage medium for storing a computer program that, when executed by a processor, implements the methods in various embodiments of this application.

[0202] Embodiments of this application also provide a computer program that, when executed by a processor, implements the methods described in the various embodiments of this application.

[0203] In the above embodiments of this application, the descriptions of each embodiment have different focuses. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0204] In the several embodiments provided in this application, it should be understood that the disclosed technical content can be implemented in other ways. The device embodiments described above are merely illustrative; for example, the division of units can be a logical functional division, and in actual implementation, there may be other division methods. For instance, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the displayed or discussed mutual coupling, direct coupling, or communication connection may be through some interfaces; the indirect coupling or communication connection between units or modules may be electrical or other forms.

[0205] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0206] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0207] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as a USB flash drive, read-only memory (ROM), random access memory (RAM), portable hard drive, magnetic disk, or optical disk.

[0208] The above description is only a preferred embodiment of this application. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the principle of this application, and these improvements and modifications should also be considered within the scope of protection of this application.

Claims

1. A vehicle driving control method, characterized in that, The method includes: In response to a vehicle traveling to a preset distance from an intersection, information on the surrounding lanes at the vehicle's location and the intersection lane information at the intersection are obtained. The surrounding lane assessment value is determined based on the preset assessment value and the surrounding lane information, and the intersection lane assessment value is determined based on the preset total assessment value and the intersection lane information, wherein the surrounding lane assessment value is used to represent the assessment value of any surrounding lane, and the intersection lane assessment value is used to represent the assessment value of any intersection lane. Based on the surrounding lane assessment values ​​and the intersection lane assessment values, at least one candidate lane is determined from the surrounding lanes; The target lane is determined from the at least one candidate lane based on the traffic status of the at least one candidate lane; Control the vehicle to travel in the target lane.

2. The method according to claim 1, characterized in that, The acquisition of surrounding lane information at the vehicle's location and intersection lane information at the intersection includes: The surrounding lane information is obtained through the vehicle's onboard perception system, wherein the surrounding lane information includes at least the number of surrounding lanes at the vehicle's location. The vehicle's in-vehicle navigation system obtains the intersection lane information, wherein the intersection lane information includes at least the number of lanes at the intersection, the lane's driving direction attribute, and the driving intention at the intersection.

3. The method according to claim 2, characterized in that, The steps of determining the surrounding lane assessment value based on the preset total assessment value and the surrounding lane information, and determining the intersection lane assessment value based on the preset total assessment value and the intersection lane information, include: The surrounding lane evaluation value is obtained by averaging the preset total evaluation value and the number of surrounding lanes. The intersection lane evaluation value is obtained by averaging the preset total evaluation value and the number of lanes at the intersection.

4. The method according to claim 2, characterized in that, The step of determining at least one candidate lane from the surrounding lanes based on the surrounding lane assessment values ​​and the intersection lane assessment values ​​includes: Based on the allocation order and the surrounding lane evaluation values, the intersection lane evaluation values ​​are sequentially allocated to the surrounding lanes to obtain an evaluation value mapping table. The evaluation value mapping table is used to record the allocation relationship between the evaluation value of any intersection lane and the surrounding lanes. Based on the evaluation value mapping table and the driving intention at the intersection, at least one candidate lane is determined from the surrounding lanes.

5. The method according to claim 4, characterized in that, The step of sequentially allocating the intersection lane evaluation values ​​to the surrounding lanes according to the allocation order and the surrounding lane evaluation values ​​to obtain the evaluation value mapping table includes: In response to the intersection lane assessment value being greater than the surrounding lane assessment value, a first sub-assessment value of the first intersection lane is assigned to a first surrounding lane, wherein the first intersection lane is the first intersection lane traversed according to the assignment order, the first surrounding lane is the first surrounding lane traversed according to the assignment order, and the first sub-assessment value is equal to the assessment value of the first surrounding lane. According to the allocation order, the second sub-evaluation value of the first intersection lane is sequentially allocated to the subsequent adjacent surrounding lanes until the evaluation value of the first intersection lane is allocated. The sum of the first sub-evaluation value and the second sub-evaluation value is equal to the evaluation value of the first intersection lane. Once the evaluation values ​​for the lanes at the first intersection have been allocated, the evaluation values ​​for the lanes at the next intersection are allocated according to the allocation order, until the evaluation values ​​for all lanes at all intersections have been allocated, thus obtaining the evaluation value mapping table.

6. The method according to claim 4, characterized in that, The step of sequentially allocating the intersection lane evaluation values ​​to the surrounding lanes according to the allocation order and the surrounding lane evaluation values ​​to obtain the evaluation value mapping table includes: In response to the intersection lane assessment value being less than the surrounding lane assessment value, the assessment values ​​of multiple consecutive intersection lanes traversed in the allocation order are cumulatively allocated to the first surrounding lane until the assessment value of the first surrounding lane is filled. Once the evaluation values ​​of the first surrounding lanes are filled, the evaluation values ​​of the unassigned intersection lanes are sequentially accumulated and assigned to the next surrounding lane according to the allocation order, until the evaluation values ​​of all surrounding lanes are filled, thus obtaining the evaluation value mapping table.

7. The method according to claim 4, characterized in that, The step of sequentially allocating the intersection lane evaluation values ​​to the surrounding lanes according to the allocation order and the surrounding lane evaluation values ​​to obtain the evaluation value mapping table includes: In response to the intersection lane assessment value being equal to the surrounding lane assessment value, the assessment values ​​of the intersection lane are assigned one-to-one to the surrounding lanes according to the allocation order, resulting in an assessment value mapping table.

8. The method according to claim 4, characterized in that, The step of determining at least one candidate lane from the surrounding lanes based on the evaluation value mapping table and the intersection driving intention includes: The target lane for the intersection is determined from the lanes at the intersection based on the driving intention at the intersection; The evaluation value allocation relationship of the target intersection lane is determined according to the evaluation value mapping relationship table, wherein the evaluation value allocation relationship is used to indicate the one or more surrounding lanes to which the evaluation value of the target intersection lane is allocated; The at least one candidate lane is determined based on the evaluation value allocation relationship, wherein the at least one candidate lane carries the evaluation value of the lane allocation at the target intersection.

9. The method according to any one of claims 1-8, characterized in that, Determining the target lane from the at least one candidate lane based on the traffic status of the at least one candidate lane includes: The traffic state of the at least one candidate lane is obtained through the vehicle's onboard perception system, wherein the traffic state includes at least one of the following: vehicle density in the lane, relative speed of vehicles ahead, obstacle information, and construction sign information; Calculate the safety score of the at least one candidate lane based on the traffic conditions; The target lane is determined from the at least one candidate lane based on the safety score of the at least one candidate lane and the evaluation value of the lane allocation at the target intersection.

10. A vehicle, characterized in that, include: Memory, which stores executable programs; A processor for running the program, wherein the program executes the vehicle driving control method according to any one of claims 1 to 9 when it runs.

Images