Vehicle guidance data generation method and device, electronic equipment and computer storage medium

Abstract

The embodiment of the application provides a vehicle induction data generation method, comprising: generating a plurality of map feature data of a target road section according to map data of the target road section, a plurality of vehicle behavior data and a plurality of traffic flow interaction data; determining a map feature data set of each vehicle behavior data according to a mapping relationship between each vehicle behavior data and each map feature data, wherein each map feature data set comprises a plurality of map feature data; and performing matching on the map feature data set of each vehicle behavior data and the map data of the target road section to generate target vehicle induction data of the target road section. The embodiment of the application provides a vehicle induction data generation method, device and electronic equipment, which can generate vehicle induction data to improve the accuracy of the map, thereby improving the humanized driving function of the autonomous vehicle.

Description

Technical Field

[0001] This application relates to the field of high-precision map technology, and in particular to a method, apparatus, electronic device, and computer storage medium for generating vehicle guidance data. Background Technology

[0002] To ensure that autonomous vehicles can drive safely and effectively within the target road segment, especially with good adaptability and interactivity in complex road environments, and to highlight a human-like driving control approach, it is necessary to increase the effective control input at the vehicle end.

[0003] Currently, high-precision maps mainly consist of layers such as road, lane, and infrastructure, which can adequately represent the road environment around the vehicle and provide some guidance for vehicle path planning. However, the map elements introduced are primarily static layers, with little or no consideration given to safety and comfort-related control constraints on the vehicle side. Based on this, current vehicle control relies heavily on the fusion of radar, laser scanning systems, and cameras to observe target object features and make control decisions. However, due to the lack of guidance control points and inputs, real-time and effective decision-making and control are difficult, easily leading to delayed autonomous driving behavior decisions or overly aggressive control inputs.

[0004] Currently, common high-precision map applications mainly focus on global and lane-level planning of target paths, but lack modeling of guidance control points based on the road environment, making it difficult to provide effective local path planning recommendations. Therefore, to improve the real-time and effective control of autonomous vehicles, it is necessary to implement multi-level and phased control of the vehicle based on the advance input of safety and comfort-related guidance control points. This would improve the vehicle's ability to predict and match the driving environment in advance, achieving the goal of anthropomorphic vehicle behavior. Summary of the Invention

[0005] In view of the above, embodiments of this application provide a method, apparatus, electronic device, and computer storage medium for generating vehicle guidance data, in order to at least partially solve the above problems.

[0006] According to a first aspect of the embodiments of this application, a method for generating vehicle guidance data is provided, comprising:

[0007] Based on the map data of the target road segment, multiple vehicle behavior data, and multiple traffic flow interaction data, multiple map feature data of the target road segment are generated.

[0008] Based on the mapping relationship between each of the vehicle behavior data and each of the map feature data, a map feature dataset for each of the vehicle behavior data is determined, wherein each of the map feature datasets includes multiple of the map feature data;

[0009] Matching is performed between the map feature dataset of each vehicle behavior data and the map data of the target road segment to generate target vehicle guidance data for the target road segment.

[0010] According to a second aspect of the embodiments of this application, a vehicle guidance data generation apparatus is provided, characterized in that it includes:

[0011] The first generation module is used to generate multiple map feature data of the target road segment based on the map data of the target road segment, multiple vehicle behavior data, and multiple traffic flow interaction data.

[0012] The determining module is configured to determine a map feature dataset for each vehicle behavior data based on the mapping relationship between each vehicle behavior data and each map feature data, wherein each map feature dataset includes multiple map feature data.

[0013] The second generation module is used to perform matching between the map feature dataset of each vehicle behavior data and the map data of the target road segment to generate target vehicle guidance data for the target road segment.

[0014] According to a third aspect of the present application, an electronic device is provided, comprising: a processor, a memory, a communication interface, and a communication bus, wherein the processor, the memory, and the communication interface communicate with each other via the communication bus; the memory is used to store at least one executable instruction, wherein the executable instruction causes the processor to perform an operation corresponding to the method for generating high-precision vehicle guidance control points as described in the first aspect.

[0015] According to a fourth aspect of the present application, a computer storage medium is provided that stores a computer program thereon, which, when executed by a processor, implements the method for generating high-precision vehicle guidance control points as described in the first aspect.

[0016] The vehicle guidance data generation method provided in this application can add a guidance control point layer based on the static layers of existing high-precision maps, improving the information on warning and control points in the high-precision map and enhancing the robustness of autonomous vehicles to the external environment. Furthermore, the introduction of guidance control points increases the input for decision-making and control of autonomous vehicles, enabling segmented and multi-level control of the vehicle in advance, achieving smooth, safe, and efficient driving. It can also abstract driver dynamic data into parameters, making the driving experience of autonomous vehicles more human-like. Attached Figure Description

[0017] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings.

[0018] Figure 1 The following is a flowchart of the steps of a vehicle guidance data generation method according to the first aspect of this application;

[0019] Figure 2 Here is a step-by-step flowchart of step S1 of a vehicle guidance data generation method according to the first aspect of this application;

[0020] Figure 3 Here is a step-by-step flowchart of step S2 of a vehicle guidance data generation method according to the first aspect of this application;

[0021] Figure 4 Here is a step-by-step flowchart of step S3 of a vehicle guidance data generation method according to the first aspect of this application;

[0022] Figure 5 This is a structural diagram of a vehicle guidance data generation device according to the second aspect of this application;

[0023] Figure 6 This is a schematic diagram of the structure of an electronic device according to a third aspect of this application. Detailed Implementation

[0024] To enable those skilled in the art to better understand the technical solutions in the embodiments of this application, the technical solutions in the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art should fall within the protection scope of the embodiments of this application.

[0025] The specific implementation of the embodiments of this application will be further described below with reference to the accompanying drawings.

[0026] A method for generating vehicle guidance data according to a first aspect embodiment of this application, such as Figure 1 As shown, it includes the following steps:

[0027] S1: Generate multiple map feature data for the target road segment based on map data, multiple vehicle behavior data, and multiple traffic flow interaction data;

[0028] In this step, by coordinating map data of the target road segment, multiple vehicle behavior data, and multiple traffic flow interaction data, as well as matching the data, the down-look feature data of the target road segment is obtained. The map feature data may include navigation information of the target road segment, such as lane information, navigation points, speed limit signs, or ramp exit signs, etc.

[0029] S2: Based on the mapping relationship between each vehicle behavior data and each map feature data, determine the map feature dataset for each vehicle behavior data, wherein each map feature dataset includes multiple map feature data.

[0030] In this step, by mapping vehicle behavior data to map feature data, vehicle behavior data on the target road segment can be effectively determined, including driving behavior data. After mapping with the map data of the target road segment, a mapping relationship is obtained, allowing for a better understanding of the road conditions or environmental factors corresponding to the vehicle's driving behavior on that target road segment. Therefore, it is possible to more accurately collect driving behavior data when facing different road conditions. Based on the mapping relationship, vehicle behavior data is recorded to form a map feature dataset.

[0031] Each map feature dataset includes multiple map data, meaning each map feature dataset can include map data for multiple road segments traveled by vehicles.

[0032] S3: Match the map feature dataset of each vehicle behavior data with the map data of the target road segment to generate target vehicle guidance data for the target road segment.

[0033] In this step, the map feature dataset obtained in the previous step is matched with the map data. As mentioned earlier, the map data can include, for example, lane information, navigation points, speed limit signs, or ramp exit signs. Therefore, if there is a data mismatch between the map feature dataset and the map data, the data information added to the map feature dataset can be obtained. In other words, data that exists in the map feature dataset but not in the map data can be added as target vehicle guidance data, thereby adding more accurate guidance points to the map.

[0034] Therefore, with the addition of target vehicle guidance data to the map, autonomous vehicles can perform autonomous driving better and have better human-like driving capabilities.

[0035] The vehicle guidance data generation method of this application can be based on existing maps, such as high-precision maps. By acquiring existing information from high-precision maps, such as static layers, target vehicle guidance data is added, improving the warning and control point information of the high-precision map and enhancing the robustness of autonomous vehicles to the external environment. Furthermore, the introduction of target vehicle guidance data increases the input for decision-making and control of autonomous vehicles, enabling segmented and multi-level control of the vehicle in advance, achieving smooth, safe, and efficient driving. Driver dynamic data can also be parameterized, making the driving experience of autonomous vehicles more human-like.

[0036] Furthermore, step S1 also includes, as follows: Figure 2 The following steps are shown:

[0037] Step S101: Select multiple initial paths on the map based on the initial location and target location of the target road segment, and confirm the map data of the multiple initial paths;

[0038] Specifically, in this step, after selecting the initial location and target location of the target road segment for the vehicle's operation, multiple paths can be obtained from the map. It should be noted that the map mentioned in this application can be a high-definition map, and will be used as the reference hereafter. The paths mentioned above provide multiple route selection options for the vehicle to travel from the initial location to the target location. Therefore, multiple paths can all be used as the initial path.

[0039] For example, if the initial location is point A and the destination is point B, there are multiple routes available from point A to point B. Therefore, based on a map and a global path encompassing multiple routes, overall route planning can be performed.

[0040] It should be noted that the initial path includes at least the initial location and the target location, and may also include multiple points along the route from the initial location to the target location. For example, from point A to point B, one path may pass through points C and D, etc. Therefore, the path from point A to point B can be understood as from point A to point C, then from point C to point D, and finally from point D to point B.

[0041] After obtaining the initial path from the map, multiple map data for the initial path can be confirmed. The map data includes road information, lane information, navigation point information, and lane traffic rule information.

[0042] This provides basic information about the target path, facilitating subsequent steps.

[0043] Step S102: Obtain the vehicle's driving behavior on multiple initial paths, as well as vehicle behavior data and traffic flow interaction data generated by the driving behavior;

[0044] In this step, driving behavior data of the vehicles can be obtained based on their driving behavior along the initial path. It should be noted that "vehicles" here refers to a large number of vehicles passing along the path. For example, if one hundred vehicles pass along the initial path, each vehicle has its own driving behavior. Therefore, driving behavior data for one hundred vehicles can be obtained, which constitutes the first characteristic behavior data.

[0045] Furthermore, since each vehicle on the initial path exhibits its own driving behavior, and each driving behavior influences the others, for example, if two vehicles are traveling in adjacent lanes on the initial path, and one vehicle needs to change lanes to the adjacent lane, the vehicle in the adjacent lane may need to slow down to allow it to pass. Alternatively, the vehicle needing to change lanes may first slow down to allow the vehicle in the adjacent lane to pass first before changing lanes. Therefore, every driving behavior can potentially affect surrounding vehicles, thus generating traffic flow interaction data.

[0046] Traffic flow interaction data is generated through collection of typical traffic flow scenario conditions. The collection of traffic flow interaction data can be achieved by dedicated data collection engineering vehicles or by obtaining data through roadside monitoring cameras.

[0047] Therefore, vehicle behavior data and traffic flow interaction data generated by the driving behavior of vehicles passing through the target path can be obtained through the above methods, thereby enabling the statistical analysis of the vehicle operation mode on the target path.

[0048] Step S103: Based on vehicle behavior data and traffic flow interaction data, establish map feature data. The map feature data includes first point data corresponding to vehicle behavior data and second point data corresponding to traffic flow interaction data.

[0049] In this step, the first location data is obtained from vehicle behavior data, and the second location data is obtained from traffic flow interaction data. The first and second location data together constitute the location setting data for multiple locations.

[0050] It is understandable that after a vehicle performs a driving action, vehicle behavior data is generated. The corresponding action occurs at a certain point on the initial path, and that point is recorded as a location.

[0051] In other words, it can be understood as the vehicle performing a certain driving action at that point. Therefore, the vehicle will perform multiple driving actions during its journey along the initial path, and the corresponding vehicle behavior data can include data on a large number of actions. Thus, multiple points can be set on the initial path to correspond to these multiple actions, thereby generating the first point data.

[0052] Correspondingly, vehicles generate various behavioral data, namely traffic flow interaction data, due to their own timely actions and those of other vehicles, while responding to traffic flow. Therefore, it is necessary to set multiple points based on the traffic flow interaction data, i.e., to generate secondary point data. Furthermore, it can be understood that traffic flow interaction data represents data generated by the interaction with surrounding traffic, and is largely characterized by dynamic data.

[0053] In addition, vehicle behavior data and traffic flow interaction data mainly refer to the driving conditions such as lane changing, overtaking, turning, cutting in, and U-turns caused by path planning during the driving process.

[0054] In summary, based on the vehicles on the initial path, vehicle behavior data and traffic flow interaction data generated by the driving behavior of multiple vehicles and their mutual influence on traffic flow can be generated. Point information can be set on the initial path to record the occurrence points of vehicle driving behavior, namely the first point data and the second point data.

[0055] Furthermore, step S2 also includes, for example, Figure 3 The following steps are shown:

[0056] Step S201: Calculate the mapping relationship between vehicle behavior data and first location data and second location data using a first predetermined algorithm, and assign feature attributes to the first location data and second location data using the mapping relationship to obtain first target location data and second target location data;

[0057] In this step, vehicle behavior data generated from the vehicle's driving behavior is mapped to designated locations, that is, the driving behavior performed by the vehicle is mapped to the point of occurrence, thus obtaining a mapping relationship. In other words, a mapping relationship is obtained by linking the first behavioral feature with the first and second location data.

[0058] By assigning feature attributes to the first and second point data through a mapping relationship, the first and second target point data can be obtained. That is, after establishing a mapping relationship between the vehicle's driving behavior and the points where it occurs, feature attributes can be assigned to the corresponding points based on the vehicle's behavior. In other words, the mapping relationship mainly refers to the characteristic behavior generated by the vehicle at that point data. Since the vehicle data collection frequency is high, multiple feature points can correspond to one characteristic behavior.

[0059] For example, at a certain point, some vehicles slow down, some change lanes, and some turn. Statistical analysis shows that 50% of vehicles slow down at this point, 20% change lanes, and 30% turn. Therefore, it can be concluded that most vehicles slow down at this point. Thus, this point can be assigned the characteristic attribute of "deceleration point."

[0060] Meanwhile, some vehicles changed their driving behavior within a certain distance before or after this point. If all of these changes involved deceleration, then they can be considered as a single point. Therefore, valid point information is obtained, resulting in the first target location data and the second target point data.

[0061] It should be noted that the vehicle behaviors mentioned here mainly include lane changing, overtaking, turning, cutting in, and U-turns in both highway and urban driving conditions.

[0062] Step S202: Calculate the global and local paths of multiple initial paths by weights to obtain the target path, and decompose the vehicle behavior data based on the target path to filter from the first target point and the second target point and obtain the map feature dataset.

[0063] In this step, by analyzing the global and local paths of all initial paths, the path selection of most vehicles from the initial location to the target location can be obtained, thus yielding a path, namely the target path.

[0064] The first behavioral feature is obtained based on the driving behavior of vehicles on the target path, and the vehicle behavior data is decomposed to obtain multiple local behavioral data of the vehicle. For example, the vehicle performs multiple driving behaviors such as lane changes, deceleration, or turns during its operation on the initial path. Therefore, each change in driving behavior can be decomposed into a local behavioral data.

[0065] This allows us to obtain information on the points corresponding to the optimal path of the vehicle from the initial location to the target location, and to filter the points in the first target point data and the second target point data to obtain an effective map feature dataset.

[0066] This step requires considering the global path and local path of multiple initial paths, as well as the driving behavior that the vehicle can reach its destination, such as lane changing, overtaking, turning, cutting in, and U-turns in highway and urban driving conditions.

[0067] Based on the driving behavior in this step, a statistical comparison will be performed with the behavioral feature data in the previous step S201, retaining the same items. At the same time, for different feature behaviors in S201 and S202, weights will be assigned based on the vehicle's state parameters, retaining the behavioral features with higher weights and discarding the behavioral features with lower weights.

[0068] Finally, the mapping relationship between feature point data and feature behavior is updated based on S201. It should be noted that the method for calculating the optimal weights of the global path and local trajectory is provided by the vehicle planning and control module.

[0069] Furthermore, step S3 also includes, for example, Figure 4 The following steps are shown:

[0070] Step S301: Filter the map feature dataset and the second point data to obtain the target point dataset;

[0071] In this step, because the selected points may have issues such as duplication or excessive density, it is necessary to filter the points. Therefore, the map feature dataset obtained in the previous step is compared with the second set of point data, and duplicate or redundant points are deleted. Alternatively, points with lower weights can be removed based on weight calculation methods, while points with higher weights are retained. This ensures the appropriate density of the selected points.

[0072] In principle, location information generated from traffic flow interaction feature data should be prioritized for elimination. Since vehicle behavior data, primarily generated by the road environment, is static data with relatively stable and reliable properties, it carries a larger weight. The specific weight allocation needs to be statistically analyzed and calibrated based on the road environment and traffic flow model.

[0073] For example, after comparison, it was found that vehicles slowed down on a certain section of the route. Although the deceleration points were different, they were spaced a certain distance apart. The interval could be ten or twenty meters, which would result in the deceleration points being set up too densely. However, vehicles only need to decelerate at one point on the route and do not need to decelerate again at the next point. Therefore, the redundant points need to be removed.

[0074] Therefore, reasonable density needs to be set based on conventional vehicle state parameters to ensure that the vehicle can travel smoothly, reliably, and efficiently from one point to the next adjacent point.

[0075] Step S302: Match the target point dataset with the map data, and generate initial vehicle guidance data based on the unmatched data in the target point dataset and the map data;

[0076] In this step, the retained valid points constitute the target map feature dataset. By comparing them with the map's navigation points, existing navigation points and those not yet set on the map can be identified. Therefore, initial vehicle guidance data can be set at locations where no navigation points are set on the map. This reduces the workload of setting initial vehicle guidance data while ensuring the accuracy of the point settings and avoiding duplicate points. It also ensures smooth vehicle operation. It should be noted that existing navigation points on the map can be, as mentioned above, navigation points, ramp exit indicators, or speed limit indicators, etc.

[0077] It should be noted that navigation points mainly refer to nodes in high-precision maps where lane topology changes, including ramp points, road termination points, road forks, and road generation points.

[0078] Step S303: Map the initial vehicle guidance data to the lanes of the target path based on the map data;

[0079] In this step, the initial vehicle guidance data set in the previous step is further refined to a specific lane. For example, if a vehicle slows down at a certain point, but the route has three lanes and the vehicle in the middle lane slows down, then the initial vehicle guidance data for the slowdown is set to the middle lane of that route.

[0080] In other words, by binding the initial vehicle guidance data to a specific lane relationship, the initial vehicle guidance data of that lane can be used to directly remind or control the lane in which the vehicle is located.

[0081] Initial vehicle guidance data is typically bound to lane-related elements, representing a portion of lane attributes, and is more suitable for lane-level planning.

[0082] Step S304: Establish the logical relationship and local path model of the global path based on the initial vehicle guidance data, map data and target path, form local path guidance through the initial vehicle guidance data, and correct the initial vehicle guidance data to generate target vehicle guidance data.

[0083] In this step, establishing the logical relationship between initial vehicle guidance data, navigation points, and global path planning data refers to the cost function of global path planning, such as optimizing the objective function. Trajectory generation and optimization are then performed between the initial vehicle guidance data and navigation points. The trajectory generation and optimization are categorized as comfortable, moderate, or aggressive based on the cost function of global path planning.

[0084] The initial vehicle guidance data and navigation points will undergo local trajectory matching to slightly correct the coordinates of the initial vehicle guidance data, thus meeting the requirements of vehicle dynamics and trajectory driving.

[0085] After setting the initial vehicle guidance data, a map with a layer containing the initial vehicle guidance data can be created on the map. To make the setting of the initial vehicle guidance data more precise, autonomous driving real-vehicle tests can be conducted based on the layer of the initial vehicle guidance data, and approximation optimizations can be performed by comparing with the real vehicle trajectory.

[0086] In some embodiments of this application, the map data for the initial path includes road information, lane information, navigation point information, and lane traffic rule information. Specifically, the map data for the initial path obtained from the map is generated by a high-precision data module device. Road information includes road topology, traffic rules, lane sets, guardrails, and road edges. Lane information includes lane edge lines, lane center lines, and lane widths. Navigation point information includes ramp points, road termination points, road separation points, and road generation points. Lane traffic rule information includes straight, left turn, right turn, and combinations of left and right turns, left and straight turns, and right and straight turns.

[0087] In some embodiments of this application, vehicle behavior data includes vehicle control change data, lateral parameter change data, longitudinal parameter change data, guidance information change data, and traffic rule guidance data.

[0088] Specifically, drastically changing vehicle-side data refers to data that exceeds or approaches the range of normal driving parameters, and these data will be mapped to their corresponding feature points. Vehicle lateral parameters mainly refer to parameters such as wheel or steering wheel angle, steering rate, steering torque, lateral rate, lateral displacement, yaw angle, and heading angle. Vehicle longitudinal control parameters mainly refer to longitudinal acceleration, speed, displacement, and pitch angle. Data showing significant changes in induced driving information mainly refers to data influenced by the road environment. Traffic rule guidance data mainly refers to data generated due to traffic rule restrictions, such as a vehicle in a left-turn lane needing to turn right, requiring the generation of steering parameters. Surrounding traffic flow interaction data mainly refers to data generated by the interaction with surrounding traffic, which is largely reflected in dynamic data.

[0089] It should be noted that the above data can be collected using specialized data collection vehicles, which should be equipped with sensing and data collection devices such as map data generators, laser scanning systems, radar, cameras, and vehicle status parameter acquisition devices.

[0090] In some embodiments of this application, corresponding first location data and second location data are established based on vehicle behavior data and traffic flow interaction data, including extracting the first location data using at least one of Bayesian classifiers, decision trees, and neural networks, and extracting the second location data using a traffic flow scene data recognition algorithm. The specific data types are as described above and will not be repeated here.

[0091] In some embodiments of this application, the first predetermined algorithm includes at least one of probability theory or AI network algorithms.

[0092] It can also be used for other intelligent algorithms, which can effectively improve the speed of computation. Furthermore, intelligent algorithms can effectively correlate vehicle behaviors, such as lane changing, overtaking, turning, cutting in, and U-turns in highway and urban driving conditions, with actual locations, thereby enhancing the human-like intelligence of autonomous vehicles.

[0093] In some embodiments of this application, the target map feature dataset is compared with the navigation points of the map, and point data in the target map feature dataset that do not overlap with the navigation points of the map are filtered out and initial vehicle guidance data is established, including:

[0094] If the distance from the point data in the target map feature dataset that does not coincide with the navigation point of the map to the navigation point is less than the trace-before-detect (TBD) distance, then the navigation point is retained;

[0095] If the distance from the point data in the target map feature dataset that does not coincide with the map's navigation point to the navigation point is greater than the TBD distance, then the navigation point and the initial vehicle guidance data are retained.

[0096] Specifically, after comparing the target map feature dataset with the navigation points, if there are points that do not overlap with the navigation points, the distance between these points and the navigation points needs to be compared with the trace-before-detect (TBD) distance. If the distance exceeds the TBD distance, both the point and the navigation point are retained; otherwise, only the navigation point is retained.

[0097] Alternatively, feature point selection can be performed using the navigation point as the center and the trace-before-detect (TBD) distance as the radius. If a point not coinciding with the navigation point is within the circular area, only the navigation point is retained. If a point not coinciding with the navigation point is not within the circular area, that point is retained.

[0098] This application is not limited to this. The above method for setting initial vehicle guidance data based on the map needs to be implemented within the Operational Design Domain (ODD) defined by the map, that is, the area where vehicles can drive as shown on the map. However, there is no need to set initial vehicle guidance data in areas where vehicles cannot drive or do not support vehicle driving.

[0099] A vehicle guidance data generation apparatus according to a second aspect embodiment of this application, such as Figure 5 As shown, it includes:

[0100] The first generation module is used to generate multiple map feature data of the target road segment based on the map data of the target road segment, multiple vehicle behavior data, and multiple traffic flow interaction data.

[0101] The determination module is used to determine the map feature dataset for each vehicle behavior data based on the mapping relationship between each vehicle behavior data and each map feature data. Each map feature dataset includes multiple map feature data.

[0102] The second generation module is used to perform matching between the map feature dataset of each vehicle behavior data and the map data of the target road segment to generate vehicle guidance data for the target road segment.

[0103] According to the vehicle guidance data generation device of this application, an initial vehicle guidance data layer can be added based on existing static layers of high-precision maps, improving the information on warning and control points on the high-precision map and enhancing the robustness of autonomous vehicles to the external environment. Furthermore, by introducing the initial vehicle guidance data, the input for decision-making and control of autonomous vehicles can be increased, allowing for segmented and multi-level control of the vehicle in advance, achieving smooth, safe, and efficient driving. Driver dynamic data can also be parameterized, making the driving experience of autonomous vehicles more human-like.

[0104] In some embodiments of this application, the first generation module is further configured to select multiple initial paths on the map based on the initial location of the target road segment and the target location, and to confirm the map data of the multiple initial paths;

[0105] The first generation module is also used to acquire the driving behavior of the vehicle on multiple initial paths, as well as vehicle behavior data and traffic flow interaction data generated by the driving behavior;

[0106] The first generation module is also used to establish map feature data based on vehicle behavior data and traffic flow interaction data. The map feature data includes first point data corresponding to vehicle behavior data and second point data corresponding to traffic flow interaction data.

[0107] In some embodiments of this application, the determining module is further configured to calculate the mapping relationship between vehicle behavior data and first location data and second location data through a first predetermined algorithm, and assign feature attributes to the first location data and second location data through the mapping relationship to obtain first target location data and second target location data;

[0108] The determination module is also used to calculate the global and local paths of multiple initial paths by weights to obtain the target path, and decompose the vehicle behavior data based on the target path to filter from the first target location and the second target location and obtain the map feature dataset.

[0109] In some embodiments of this application, the second generation module is further used to filter the map feature dataset and the second point data to obtain the target point dataset;

[0110] The second generation module is also used to match the target point dataset with the map data, and generate initial vehicle guidance data based on the unmatched data in the target point dataset and the map data.

[0111] The second generation module is also used to map the initial vehicle guidance data to the lanes of the target path based on the map data;

[0112] The second generation module is also used to establish the logical relationship and local path model of the global path based on the initial vehicle guidance data, map data and target path, form local path guidance through the initial vehicle guidance data, and correct the initial vehicle guidance data to generate target vehicle guidance data.

[0113] The vehicle guidance data generation device of this embodiment is used to implement the corresponding deadlock handling methods in the foregoing method embodiments and has the beneficial effects of the corresponding method embodiments, which will not be repeated here. Furthermore, the functional implementation of each module in the vehicle guidance data generation device of this embodiment can be referred to the description of the corresponding part in the foregoing method embodiments, which will also not be repeated here.

[0114] An electronic device according to a third aspect of this application includes: a processor, a memory, a communication interface, and a communication bus. The processor, the memory, and the communication interface communicate with each other through the communication bus. The memory is used to store at least one executable instruction, which causes the processor to perform an operation corresponding to the vehicle guidance data generation method of the first aspect of this application.

[0115] Reference Figure 6 The diagram shows a structural schematic of an electronic device according to Embodiment 5 of this application. The specific embodiments of this application do not limit the specific implementation of the electronic device.

[0116] like Figure 6 As shown, the electronic device may include: a processor 602, a communications interface 604, a memory 606, and a communications bus 608.

[0117] in:

[0118] The processor 602, communication interface 604, and memory 606 communicate with each other via communication bus 608.

[0119] Communication interface 604 is used for communication with other electronic devices or servers.

[0120] The processor 602 is used to execute program 610, specifically to execute the relevant steps in the above-described vehicle guidance data generation method embodiment.

[0121] Specifically, program 610 may include program code that includes computer operation instructions.

[0122] The processor 602 may be a central processing unit (CPU), an application-specific integrated circuit (ASIC), or one or more integrated circuits configured to implement the embodiments of this application. The smart device includes one or more processors, which may be processors of the same type, such as one or more CPUs; or they may be processors of different types, such as one or more CPUs and one or more ASICs.

[0123] Memory 606 is used to store program 610. Memory 606 may include high-speed RAM memory, and may also include non-volatile memory, such as at least one disk storage device.

[0124] Specifically, program 610 can be used to cause processor 602 to perform the following operations:

[0125] The specific implementation of each step in program 610 can be found in the corresponding steps and units described in the above-described vehicle guidance data generation method embodiments, and will not be repeated here. Those skilled in the art will understand that, for the sake of convenience and brevity, the specific working processes of the devices and modules described above can be referred to the corresponding process descriptions in the aforementioned method embodiments, and will not be repeated here.

[0126] The electronic device in this embodiment can effectively implement the aforementioned vehicle guidance data generation method. Based on existing high-precision map static layers and other information, a guidance control point layer can be added, improving the warning and control point information of the high-precision map and enhancing the robustness of autonomous vehicles to the external environment. Furthermore, the introduction of guidance control points increases the input for decision-making and control of autonomous vehicles, enabling segmented, multi-level control of the vehicle in advance, achieving smooth, safe, and efficient driving. Driver dynamic data can also be parameterized, making the driving experience of autonomous vehicles more human-like.

[0127] The vehicle guidance data generation method of this embodiment can be executed by any suitable electronic device with data processing capabilities, including but not limited to: servers, mobile terminals (such as mobile phones, PADs, etc.) and PCs.

[0128] A computer storage medium according to a fourth aspect of the present application stores a computer program that, when executed by a processor, implements the vehicle guidance data generation method of the first aspect of the present application.

[0129] It should be noted that, depending on the implementation needs, the various components / steps described in the embodiments of this application can be broken down into more components / steps, or two or more components / steps or parts of the operation of components / steps can be combined into new components / steps to achieve the purpose of the embodiments of this application.

[0130] The methods described in the embodiments of this application can be implemented in hardware, firmware, or as software or computer code that can be stored in a recording medium (such as a CD-ROM, RAM, floppy disk, hard disk, or magneto-optical disk), or as computer code downloaded over a network that is originally stored in a remote recording medium or a non-transitory machine-readable medium and will be stored in a local recording medium. Thus, the methods described herein can be processed by software stored on a recording medium using a general-purpose computer, a dedicated processor, or programmable or dedicated hardware (such as an ASIC or FPGA). It is understood that the computer, processor, microprocessor controller, or programmable hardware includes storage components (e.g., RAM, ROM, flash memory, etc.) capable of storing or receiving software or computer code. When the software or computer code is accessed and executed by the computer, processor, or hardware, the vehicle guidance data generation method described herein is implemented. Furthermore, when a general-purpose computer accesses code used to implement the vehicle guidance data generation method shown herein, the execution of the code transforms the general-purpose computer into a dedicated computer for executing the vehicle guidance data generation method shown herein.

[0131] Those skilled in the art will recognize that the units and method steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of the embodiments of this application.

[0132] The above embodiments are only used to illustrate the embodiments of this application, and are not intended to limit the embodiments of this application. Those skilled in the art can make various changes and modifications without departing from the spirit and scope of the embodiments of this application. Therefore, all equivalent technical solutions also fall within the scope of the embodiments of this application, and the patent protection scope of the embodiments of this application should be defined by the claims.

Claims

1. A method for generating vehicle guidance data, characterized in that, include: Based on the map data of the target road segment, multiple vehicle behavior data, and multiple traffic flow interaction data, multiple map feature data of the target road segment are generated. Based on the mapping relationship between each of the vehicle behavior data and each of the map feature data, a map feature dataset for each of the vehicle behavior data is determined, wherein each of the map feature datasets includes multiple of the map feature data; Matching is performed between the map feature dataset of each vehicle behavior data and the map data of the target road segment to generate target vehicle guidance data for the target road segment; The step of generating multiple map feature data for the target road segment based on map data, multiple vehicle behavior data, and multiple traffic flow interaction data includes: selecting multiple initial paths on the map based on the initial location and the target location of the target road segment, and confirming the map data of the multiple initial paths; acquiring the driving behavior of vehicles on the multiple initial paths, as well as the vehicle behavior data and the traffic flow interaction data generated by the driving behavior; and establishing the map feature data based on the vehicle behavior data and the traffic flow interaction data, wherein the map feature data includes first point data corresponding to the vehicle behavior data and second point data corresponding to the traffic flow interaction data. The step of determining a map feature dataset for each vehicle behavior data based on the mapping relationship between each vehicle behavior data and each map feature data, wherein each map feature dataset includes multiple map feature data, includes: calculating the mapping relationship between the vehicle behavior data and the first point data and the second point data using a first predetermined algorithm, and assigning feature attributes to the first point data and the second point data using the mapping relationship to obtain first target point data and second target point data; calculating the global path and local path of multiple initial paths using weights to obtain a target path, and decomposing the vehicle behavior data based on the target path to filter among the first target point and the second target point and obtain a map feature dataset.

2. The method according to claim 1, characterized in that, Matching is performed between the map feature dataset of each vehicle behavior data point and the map data of the target road segment to generate target vehicle guidance data for the target road segment, including: The map feature dataset and the second point data are filtered to obtain the target point dataset; The target location dataset is matched with the map data, and initial vehicle guidance data is generated based on the unmatched data in the target location dataset that does not match the map data. Based on the map data, the initial vehicle guidance data is mapped to the lanes of the target path; Establish a logical relationship and local path model for the global path based on the initial vehicle guidance data, the map data, and the target path. Form local path guidance through the initial vehicle guidance data and correct the initial vehicle guidance data to generate target vehicle guidance data.

3. The method according to claim 1, characterized in that, The map data for the initial path includes road information, lane information, navigation point information, and lane traffic rule information.

4. The method according to claim 1, characterized in that, The vehicle behavior data includes vehicle control change data, lateral parameter change data, longitudinal parameter change data, guidance information change data, and traffic rule guidance data.

5. The method according to claim 1, characterized in that, The map feature data is established based on the vehicle behavior data and the traffic flow interaction data. The map feature data includes first location data corresponding to the vehicle behavior data and second location data corresponding to the traffic flow interaction data, including: The first location data is extracted using at least one of Bayesian classifier, decision tree, and neural network, and the second location data is extracted using a traffic flow scene data recognition algorithm.

6. The method according to claim 1, characterized in that, The first predetermined algorithm includes at least one of probabilistic theory or AI network algorithms.

7. A vehicle guidance data generation device, characterized in that, include: The first generation module is used to generate multiple map feature data of the target road segment based on the map data of the target road segment, multiple vehicle behavior data, and multiple traffic flow interaction data. The determining module is configured to determine a map feature dataset for each vehicle behavior data based on the mapping relationship between each vehicle behavior data and each map feature data, wherein each map feature dataset includes multiple map feature data. The second generation module is used to perform matching between the map feature dataset of each vehicle behavior data and the map data of the target road segment to generate vehicle guidance data for the target road segment; The first generation module is further configured to select multiple initial paths on the map based on the initial location and the target location of the target road segment, and confirm the map data of the multiple initial paths; acquire the driving behavior of the vehicle on the multiple initial paths, as well as the vehicle behavior data and the traffic flow interaction data generated by the driving behavior; and establish the map feature data based on the vehicle behavior data and the traffic flow interaction data, wherein the map feature data includes first point data corresponding to the vehicle behavior data and second point data corresponding to the traffic flow interaction data; The determining module is further configured to calculate the mapping relationship between the vehicle behavior data and the first point data and the second point data using a first predetermined algorithm, and assign feature attributes to the first point data and the second point data through the mapping relationship to obtain the first target point data and the second target point data; calculate the global path and local path of multiple initial paths through weights to obtain the target path, and decompose the vehicle behavior data based on the target path to filter among the first target point and the second target point and obtain a map feature dataset.

8. The apparatus according to claim 7, characterized in that, The second generation module is also used to filter the map feature dataset and the second point data to obtain the target point dataset; The second generation module is further configured to match the target point dataset with the map data, and generate initial vehicle guidance data based on the unmatched data in the map data; The second generation module is further configured to map the initial vehicle guidance data to the lanes of the target path based on the map data; The second generation module is also used to establish the logical relationship and local path model of the global path based on the initial vehicle guidance data, the map data and the target path, form local path guidance through the initial vehicle guidance data, and correct the initial vehicle guidance data to generate target vehicle guidance data.

9. An electronic device, comprising: The processor, memory, communication interface, and communication bus are provided, wherein the processor, memory, and communication interface communicate with each other via the communication bus. The memory is used to store at least one executable instruction, which causes the processor to perform the operation corresponding to the vehicle guidance data generation method as described in any one of claims 1-6.

10. A computer storage medium having a computer program stored thereon, which, when executed by a processor, implements the vehicle guidance data generation method as described in any one of claims 1-6.

Images