Data mining method, vehicle control method, device, equipment and storage medium

Abstract

The present disclosure provides a data mining method, device and equipment and storage medium, relates to the technical field of data processing, and particularly relates to the technical field of intelligent driving, smart city, artificial intelligence, big data and the like. The specific implementation scheme is as follows: extracting road position data in a target scene from electronic map data, the target scene being a scene in which the probability of a sudden event is greater than a set probability; determining a road range to be subjected to data mining according to the road position data; obtaining historical driving data in the road range; and performing data mining on the historical driving data to obtain driving feature data corresponding to the road range, wherein the driving feature data is used to assist automatic driving of a vehicle. Thus, by mining driving feature data in a specific scene, the safety and intelligent level of automatic driving of the vehicle are improved.

Description

Technical Field

[0001] This disclosure relates to the fields of intelligent driving, smart cities, artificial intelligence, big data, etc. in the field of data processing technology, and in particular to a data mining method, vehicle control method, device, equipment and storage medium. Background Technology

[0002] With the development of autonomous driving technology, its application scenarios have evolved from assisted driving to autonomous driving under certain conditions.

[0003] In related technologies, autonomous vehicles use sensors, such as laser sensors and cameras, to sense and identify road signs and pedestrians in real time, and make driving decisions in conjunction with static high-precision maps.

[0004] However, the inventors discovered during their research that the aforementioned decision-making methods could easily lead to sudden braking or acceleration, affecting the safety of autonomous vehicles. Therefore, the level of intelligence in autonomous driving still needs to be improved. Summary of the Invention

[0005] This disclosure provides a data mining method, vehicle control method, apparatus, device, and storage medium for improving the safety and intelligence of autonomous driving.

[0006] According to a first aspect of this disclosure, a data mining method is provided, comprising: extracting road location data of a target scenario from electronic map data, wherein the target scenario is a scenario in which the probability of an emergency occurring is greater than a set probability; determining the road range to be data mined based on the road location data; obtaining historical driving data within the road range; and performing data mining on the historical driving data to obtain driving feature data corresponding to the road range, wherein the driving feature data is used to assist in the autonomous driving of the vehicle.

[0007] According to a second aspect of this disclosure, a vehicle control method is provided, comprising: acquiring driving characteristic data, the driving characteristic data being obtained using the data mining method described in the first aspect; generating a driving decision based on the driving characteristic data and perception data of an autonomous vehicle; and controlling the autonomous vehicle based on the driving decision.

[0008] According to a third aspect of this disclosure, a data mining apparatus is provided, comprising: a location extraction unit for extracting road location data of a target scene from electronic map data, wherein the target scene is a scene in which the probability of an emergency occurring is greater than a set probability; a range determination unit for determining the range of roads to be data mined based on the road location data; a data acquisition unit for acquiring historical driving data within the road range; and a data mining unit for performing data mining on the historical driving data to obtain driving feature data corresponding to the road range, wherein the driving feature data is used to assist in the autonomous driving of vehicles.

[0009] According to a fourth aspect of this disclosure, a vehicle control device is provided, comprising: an acquisition unit for acquiring driving characteristic data, the driving characteristic data being obtained according to the data mining method described in the first aspect; a decision unit for generating a driving decision based on the driving characteristic data and perception data of an autonomous vehicle; and a control unit for controlling the autonomous vehicle based on the driving decision.

[0010] According to a fifth aspect of this disclosure, an electronic device is provided, comprising:

[0011] At least one processor; and

[0012] A memory that is communicatively connected to the at least one processor;

[0013] The memory stores instructions executable by the at least one processor, which are executed by the at least one processor to enable the at least one processor to perform the data mining method described in the first aspect and / or the vehicle control method described in the second aspect.

[0014] According to a sixth aspect of this disclosure, a non-transitory computer-readable storage medium is provided storing computer instructions, wherein the computer instructions are used to cause the computer to perform the data mining method of the first aspect and / or the vehicle control method of the second aspect.

[0015] According to a seventh aspect of this disclosure, a computer program product is provided, the computer program product comprising: a computer program stored in a readable storage medium, at least one processor of an electronic device being able to read the computer program from the readable storage medium, the at least one processor executing the computer program causing the electronic device to perform the steps of the data mining method of the first aspect and / or the steps of the vehicle control method of the second aspect.

[0016] According to the technical solution provided in this disclosure, road location data for a target scenario is extracted from electronic map data. Based on the road location data, the road range to be mined is determined. The target scenario is one where the probability of an emergency is greater than a set probability. The road range determined based on the road location data of the target scenario can more accurately delineate the area prone to emergencies within the target scenario. Historical driving data within the road range is obtained. While improving the accuracy of the road range, the historical driving data can accurately reflect vehicle driving behavior in the target scenario, especially human-driven vehicle driving behavior. The historical driving data within the road range is mined to obtain driving feature data for assisting autonomous driving, improving the accuracy of the driving feature data. This allows the driving feature data to accurately reflect the vehicle driving characteristics within the road range, especially the driving characteristics of human-driven vehicles. Consequently, autonomous driving can make safer driving decisions that are closer to human-driven vehicle driving characteristics based on an understanding of the vehicle driving characteristics within the road range, reducing sudden braking, sudden acceleration, and other autonomous driving behaviors, improving the intelligence level of autonomous driving, and enhancing the autonomous driving experience.

[0017] It should be understood that the description in this section is not intended to identify key or essential features of the embodiments of this disclosure, nor is it intended to limit the scope of this disclosure. Other features of this disclosure will become readily apparent from the following description. Attached Figure Description

[0018] The accompanying drawings are provided to better understand this solution and do not constitute a limitation of this disclosure. Wherein:

[0019] Figure 1 This is a schematic diagram illustrating an application scenario to which this disclosure applies;

[0020] Figure 2 This is a flowchart illustrating the data mining method provided according to embodiments of the present disclosure. Figure 1 ;

[0021] Figure 3 This is a flowchart illustrating the data mining method provided according to embodiments of the present disclosure. Figure 2 ;

[0022] Figure 4 Examples of road entrance / exit scenarios provided in embodiments of this disclosure Figure 1 ;

[0023] Figure 5 This is an example diagram of a long solid line scenario provided in an embodiment of this disclosure;

[0024] Figure 6 Example diagram of a speed measurement scenario at a non-intersection provided in this embodiment of the disclosure;

[0025] Figure 7 Example diagram of a testing device at a non-intersection location provided in an embodiment of this disclosure;

[0026] Figure 8 Examples of road entrance / exit scenarios provided in embodiments of this disclosure Figure 2 ;

[0027] Figure 9 This is a flowchart illustrating the data mining method provided according to embodiments of the present disclosure. Figure 3 ;

[0028] Figure 10 Examples of POI scenarios provided in embodiments of this disclosure Figure 1 ;

[0029] Figure 11 Examples of POI scenarios provided in embodiments of this disclosure Figure 2 ;

[0030] Figure 12 This is a flowchart illustrating the data mining method provided according to embodiments of the present disclosure. Figure 4 ;

[0031] Figure 13 Example diagram of a speed-limited road section scenario provided in the embodiments of this disclosure;

[0032] Figure 14 This is a schematic flowchart of a vehicle control method provided according to an embodiment of the present disclosure;

[0033] Figure 15 Schematic diagram of the data mining apparatus provided in the embodiments of this disclosure Figure 1 ;

[0034] Figure 16 Schematic diagram of the data mining apparatus provided in the embodiments of this disclosure Figure 2 ;

[0035] Figure 17 This is a schematic diagram of the structure of the vehicle control device provided in the embodiments of this disclosure;

[0036] Figure 18 This is a schematic block diagram of an example electronic device 1800 that can be used to implement embodiments of the present disclosure. Detailed Implementation

[0037] The exemplary embodiments of this disclosure are described below with reference to the accompanying drawings, including various details of the embodiments to aid understanding, and should be considered merely exemplary. Therefore, those skilled in the art will recognize that various changes and modifications can be made to the embodiments described herein without departing from the scope and spirit of this disclosure. Similarly, for clarity and brevity, descriptions of well-known functions and structures are omitted in the following description.

[0038] In related technologies, autonomous vehicles use sensors to detect and identify road signs and pedestrians in real time, and make driving decisions based on static high-precision maps. However, when autonomous vehicles sense a vehicle ahead suddenly accelerating or decelerating, or when they recognize a speed limit sign, they are prone to sudden braking or acceleration, such as suddenly decelerating to the speed limit indicated on the sign, which affects the safety of the autonomous vehicle and the passenger experience.

[0039] If autonomous driving algorithms can learn in advance about the fixed behaviors or habits of people and vehicles at certain locations in public roads, then the autonomous driving algorithms can effectively avoid people and vehicles at these locations, and can even learn the driving behaviors of other people and vehicles, reducing autonomous driving behaviors that affect driving safety and passenger experience, such as sudden braking and sudden acceleration.

[0040] Based on the above considerations, this disclosure provides a data mining method, vehicle control method, apparatus, device, and storage medium. In the data mining method, electronic map data is used to determine the road range to be mined in road scenarios with a high probability of sudden events. Based on historical driving data within the road range, driving characteristic data corresponding to the road range is mined. This driving characteristic data is used to assist in the autonomous driving of vehicles. The road range in a road scenario is a key area, and the probability of vehicles experiencing sudden events is high within this range. The driving characteristic data within the road range can reflect the driving characteristics of vehicles within that range, especially the driving habits of drivers and vehicles. This allows the autonomous driving algorithm to learn the driving habits of drivers and vehicles within the road range, and also enables effective responses to sudden events within the road range. This reduces sudden braking and acceleration behaviors of autonomous vehicles, improves the safety and passenger experience of autonomous vehicles, and enhances the intelligence level of autonomous vehicles.

[0041] The vehicle control method utilizes driving characteristic data obtained through data mining to control autonomous vehicles, thereby improving their safety, passenger experience, and level of intelligence. The implementation principles and technical effects of the aforementioned devices and methods are consistent and will not be elaborated upon further.

[0042] Figure 1 This diagram illustrates an application scenario applicable to an embodiment of the present disclosure. In this scenario, the devices involved include a data mining device 101 and a database 102. The data mining device 101 can be a server or a terminal. Figure 1 Taking the server as an example; database 102 stores electronic map data and historical driving data.

[0043] The data mining device 101 can obtain electronic map data and historical driving data from the database 102, and use the electronic map data and historical driving data to mine driving feature data corresponding to the road range in a specific road scenario, so as to improve the safety of autonomous vehicles driving in road scenarios, passenger riding experience and intelligence level.

[0044] Optionally, the application scenario also includes an autonomous driving computing device 103 and an autonomous driving vehicle 104, with an on-board unit 105 installed on the autonomous driving vehicle 104. The autonomous driving computing device 103 can acquire driving characteristic data from the data mining device 101, train or run an autonomous driving algorithm based on the driving characteristic data, and after running the autonomous driving algorithm, send driving decisions to the on-board unit 105 to control the autonomous driving behavior of the autonomous driving vehicle 104. The autonomous driving computing device 103 can be a server or a terminal. Figure 1 Take a server as an example.

[0045] The technical solutions of this disclosure and how they solve the aforementioned technical problems will be described in detail below with specific embodiments. These specific embodiments can be combined with each other, and the same or similar concepts or processes may not be repeated in some embodiments. The embodiments of this disclosure will now be described with reference to the accompanying drawings.

[0046] For example, the execution subject of this disclosure embodiment can be an electronic device, which can be a server or a terminal. The server can be a centralized server, a distributed server, or a cloud server. The terminal can be a personal digital assistant (PDA) device, a handheld device with wireless communication capabilities (e.g., a smartphone, tablet), a computing device (e.g., a personal computer, PC), an in-vehicle device, a wearable device (e.g., a smartwatch, smart bracelet), and a smart home device (e.g., a smart speaker, smart display device), etc.

[0047] Figure 2 This is a flowchart illustrating the data mining method provided according to embodiments of the present disclosure. Figure 1 .like Figure 2 As shown, data mining methods include:

[0048] S201, Extract road location data for the target scene from electronic map data.

[0049] In the electronic map data, multiple roads, traffic signs on the roads, roadside equipment, and other road information are identified by assigning attribute values. Traffic signs include road markings and road signs, while roadside equipment includes speed cameras.

[0050] The target scenario is a scenario where the probability of an unexpected event occurring is greater than a predetermined probability. One or more target scenarios can be pre-defined by professionals, or they can be determined from multiple road scenarios by statistically analyzing the number or frequency of unexpected events. Unexpected events include sudden driving behavior events, such as sudden braking, sudden acceleration, or sudden lane changing.

[0051] In this embodiment, roads have different characteristics in different road scenarios. For example, in the main-auxiliary road entrance / exit scenario, it includes the main road, the auxiliary road, and the entrances / exits between the main and auxiliary roads. Similarly, in the temporary parking scenario, it includes temporary parking signs and parking lines. The areas of roads prone to emergencies also differ in different road scenarios. Therefore, based on the road characteristics and the potential emergencies in the target scenario, road location data for the target scenario can be extracted from electronic map data. This allows the road location data to be correlated with the potential emergencies in the target scenario, enabling a more accurate determination of the road range prone to emergencies within the target scenario.

[0052] S202, Based on road location data, determine the scope of roads to be data mined.

[0053] The road range to be data mined includes road ranges in the target scenario where sudden events are likely to occur. There may be one or more road ranges to be data mined.

[0054] In this embodiment, after obtaining the road location data in the target scene, the starting and ending points of the road range can be determined based on the road location data. The road range can be represented by the starting and ending points of the road range. In this way, one or more road ranges in the target scene to be data mined can be obtained.

[0055] S203, obtain historical driving data within the road area.

[0056] Historical driving data may include driving data of multiple vehicles within the road area over a past period. This data may include at least one of the following: driving speed, driving time, and driving trajectory. "Past time" can be, for example, the past month or the past week, and can be a time period from a past moment to the present moment, or a time period between two past moments. There are no specific limitations on the definition of "past time."

[0057] In this embodiment, historical driving data within a road area can be queried and obtained from a database storing vehicle driving data based on past time and road area. For example, driving data for the past time can be retrieved from the database first, and then historical driving data within the road area can be retrieved from the past time driving data based on the road area.

[0058] S204. Data mining is performed on historical driving data to obtain driving characteristic data corresponding to the road range. The driving characteristic data is used to assist the autonomous driving of vehicles.

[0059] In this embodiment, after obtaining historical driving data, at least one of the information in the historical driving data, such as driving speed, driving time, and driving trajectory, can be analyzed and mined to obtain driving feature data corresponding to the road range. The driving feature data may include the driving characteristics of vehicles within the road range in at least one aspect of driving speed, driving time, and driving trajectory.

[0060] In this embodiment, the driving feature data is mined from historical driving data within the road area. Historical driving data contains a large amount of data reflecting the driving characteristics and habits of vehicles within the road area, especially the driving characteristics and habits of human-driven vehicles. The driving feature data mined from historical driving data can more clearly, concisely, and accurately reflect the driving characteristics and habits of vehicles. Using driving feature data to assist in autonomous driving allows the autonomous driving algorithm to learn more about the driving characteristics and habits of human-driven vehicles, and to better understand the driving situation of vehicles within the road area, enabling effective avoidance and reducing sudden braking and acceleration behaviors of autonomous vehicles in road scenarios with a high probability of unexpected events. This improves the safety of autonomous driving, the passenger experience, and the level of intelligence of autonomous driving.

[0061] In some embodiments, the target scenario includes at least one of the following: road entrance / exit scenario, road marking scenario, speed measurement scenario at non-intersection locations, map point of interest (POI) scenario, and speed-limited road section scenario.

[0062] Among them, the road entrance / exit scenario refers to a road scenario where vehicles can exit from one road and enter another; the road marking scenario refers to a road scenario that includes specific road markings (such as long solid lines, stop lines, and double yellow lines); the non-intersection speed measurement scenario refers to a road scenario that is not at an intersection but has a speed measurement device installed; the POI scenario refers to a road scenario that is near a POI (such as a school, shopping mall, or tourist attraction); and the speed-limited section scenario refers to a road scenario that restricts driving speed.

[0063] In road entrance / exit scenarios and road marking scenarios, sudden speed changes and / or lane changes are likely to occur. For example, when the road marking is a long solid line, a vehicle may change lanes in advance to reach a certain destination. In non-intersection speed measurement scenarios and speed-limited road sections, vehicles may suddenly decelerate or accelerate to the speed limit (such as the maximum speed limit). In POI scenarios, mixed pedestrian and vehicle situations are likely to occur, which can easily lead to sudden events such as lane changes and braking.

[0064] In this embodiment, the road range to be data mined can be determined in at least one of the following road scenarios: road entrance / exit scenario, road marking scenario, speed measurement scenario at non-intersections, map point of interest (POI) scenario, and speed-limited road segment scenario. Based on historical driving data within the road range, driving feature data corresponding to the road range is analyzed to improve the safety of autonomous vehicles in at least one of the following road scenarios: road entrance / exit scenario, road marking scenario, speed measurement scenario at non-intersections, map point of interest (POI) scenario, and speed-limited road segment scenario.

[0065] Furthermore, road access scenarios can include main and auxiliary road access scenarios and / or ramp access scenarios. In main and auxiliary road access scenarios, vehicles can exit from the main road and enter the auxiliary road; in ramp access scenarios, vehicles can exit from the main road and enter the ramp. When preparing to enter the auxiliary road or ramp, vehicles need to change lanes and / or decelerate in advance. Therefore, sudden lane changes and sudden decelerations are prone to occur in main and auxiliary road access scenarios and / or ramp access scenarios. Setting road access scenarios including main and auxiliary road access scenarios and / or ramp access scenarios improves the safety of autonomous vehicles in these scenarios.

[0066] Furthermore, road marking scenarios can include long solid line scenarios. In long solid line scenarios, vehicles cannot change lanes at the long solid line; however, vehicles may change lanes at the beginning or end of the long solid line, and may accelerate or decelerate simultaneously during the lane change. Therefore, sudden lane changes and sudden speed changes are also prone to occur in long solid line scenarios. Including long solid line scenarios in road entrance and exit scenarios improves the safety of autonomous vehicles in these areas. In addition, road marking scenarios may also include stop line scenarios, double yellow line scenarios, and deceleration marking scenarios, etc., which will not be described in detail here.

[0067] In some embodiments, the driving characteristic data corresponding to the road range includes at least one of the following: the location area where deceleration and / or lane change occurs corresponding to the road range, the time period of pedestrian-vehicle mixing corresponding to the road range, and the recommended driving speed corresponding to the road range.

[0068] Specifically, in the areas where deceleration and / or lane changes occur within the road area, vehicles are more likely to engage in deceleration and / or lane changes; during the mixed pedestrian and vehicle time periods within the road area, pedestrians and vehicles are mixed within the road area; and the recommended driving speed within the road area is more consistent with the actual driving speed of vehicles within the road area.

[0069] In this embodiment, when the location area corresponding to the road range where deceleration and / or lane change occur is used to assist the autonomous driving of the vehicle, the autonomous vehicle can decelerate and / or change lanes within this location area. It can also make driving decisions to avoid deceleration and / or lane changes of other vehicles within this location area, reducing sudden braking and lane changes, and improving the safety and intelligence of autonomous driving. When the pedestrian-vehicle mixed time period corresponding to the road range is used to assist the autonomous driving of the vehicle, the autonomous vehicle can avoid this road range during the pedestrian-vehicle mixed time period. It can also make driving decisions during the pedestrian-vehicle mixed time period to avoid pedestrians and other vehicles within the road range, reducing sudden braking and lane changes, and improving the safety and intelligence of autonomous driving. When the recommended driving speed corresponding to the road range is used to assist the autonomous driving of the vehicle, the autonomous vehicle can drive at the recommended driving speed to avoid safety issues caused by inappropriate autonomous driving speed.

[0070] Below, we provide corresponding implementation examples for data mining in different scenarios.

[0071] Figure 3 This is a flowchart illustrating the data mining method provided according to embodiments of the present disclosure. Figure 2 .like Figure 3 As shown, when the target scenario includes at least one of the following: road entrance / exit scenario, road marking scenario, and speed measurement scenario at non-intersections, the data mining methods include:

[0072] S301, extract the locations of key points in the target scene from electronic map data.

[0073] Among them, the key point location is related to the vehicle's deceleration and lane change in the target scenario. Typically, the vehicle decelerates and changes lanes when it is about to reach the key point location.

[0074] The key locations include at least one of the following: entrance / exit locations, road marking locations, and the projected location of speed measuring devices at non-intersection locations on the map road. The projected locations of entrance / exit locations, road marking locations, and speed measuring devices at non-intersection locations on the map road are all closely related to vehicle deceleration and lane changes. Vehicles typically decelerate and change lanes when they are about to reach the location of the entrance / exit location, road marking location, or the location of the speed measuring device at a non-intersection location.

[0075] In this embodiment, when the target scenario includes a road entrance / exit scenario, the entrance / exit location can be extracted from the electronic map data; when the target scenario includes a road marking scenario, the road marking location can be extracted from the electronic map data; when the target scenario includes a speed measurement scenario at a non-intersection location, the projection location of the speed measurement device at the non-intersection location on the map road can be extracted from the electronic map data.

[0076] In this embodiment, during the extraction of key point locations, the key point locations in the target scene can be extracted from the electronic map data based on the road characteristics of the target scene. The road characteristic of a road entrance / exit scene is that there is an entrance / exit location between two roads; this characteristic can be used to extract the entrance / exit location from the electronic map data. The road characteristic of a road marking scene is that it contains road markings; this characteristic can be used to extract the road marking location from the electronic map data. The road characteristic of a speed measurement scene at a non-intersection is that a speed measuring device is installed on the road; this characteristic can be used to extract the projection location of the speed measuring device at a non-intersection location on the map road from the electronic map data.

[0077] In one possible implementation, when the target scenario includes a road entrance / exit scenario, step S301 may include: determining a target lane in the electronic map data, wherein the target lane is a lane that indicates a vehicle's entry from a main road onto an auxiliary road or ramp; and determining and extracting the entrance / exit location based on the target lane. Thus, by utilizing the characteristic that the road entrance / exit scenario includes a lane indicating a vehicle's entry from a main road onto an auxiliary road or ramp, and leveraging the specific characteristics of this lane, the entrance / exit location is determined, improving the accuracy of determining the entrance / exit location in the road entrance / exit scenario.

[0078] In this implementation, lanes containing entrance / exit locations can be pre-marked in the electronic map data. This allows the target lane to be determined by identifying lanes with corresponding markings within the electronic map data. In road entrance / exit scenarios, vehicles in the target lane can enter auxiliary roads or ramps from the main road. Therefore, the target lane may be interrupted, meaning it may be interrupted at the entrance / exit location. Before reaching the entrance / exit location, the vehicle can travel along the target lane to the auxiliary road or ramp. If the vehicle passes the entrance / exit location but does not enter, it can continue along the target lane. Therefore, the target lane can also be identified in the electronic map based on the characteristic of being interrupted by entrance / exit locations. After determining the target lane, the entrance / exit locations can be determined on the target lane based on the characteristic of being interrupted by entrance / exit locations, and then extracted from the electronic map data.

[0079] Furthermore, based on the changes in the attribute values ​​of the target lane, the locations of entrances and exits can be determined on the target lane and extracted from the electronic map data. This improves the accuracy of entrance and exit location extraction by utilizing the lane attribute values ​​in the electronic map data.

[0080] In this implementation, the electronic map data includes the attribute values ​​corresponding to the lanes, so as to identify the lanes by the attribute values. Because the target lane is interrupted, the attribute values ​​of the target lane change. The location where the attribute values ​​of the target lane change is the entrance / exit location.

[0081] As an example, Figure 4 Examples of road entrance / exit scenarios provided in embodiments of this disclosure Figure 1 ,like Figure 4 As shown, the main road has 3 lanes and the auxiliary road (or ramp) has 2 lanes. There is an entrance / exit position N1 on the lane of the main road closer to the auxiliary road. This lane is interrupted by the entrance / exit position N1. There are two attribute values ​​before and after N1: L1 and L2. Therefore, N1 can be determined on this lane based on the changes of L1 and L2, and N1 can be extracted from the electronic map data.

[0082] In another possible implementation, where the target scenario includes road marking scenarios, specifically long solid line scenarios, and the road marking locations include the endpoints of the long solid lines, step S301 includes: determining the long solid lines on the map roads from the electronic map data; and extracting the endpoints of the long solid lines from the electronic map data. Thus, for long solid line scenarios, considering the characteristic that vehicles are prone to speed changes and / or lane changes before or after long solid lines, the endpoints of the long solid lines are extracted as key point locations in the long solid line scenario, improving the accuracy of key point locations.

[0083] In this implementation, the attribute values ​​of road markings in the electronic map data include the type of road marking and the endpoint positions of the road markings. Therefore, road markings of the long solid line type can be identified in the electronic map data to obtain the long solid lines on the map roads, and then the endpoint positions of the long solid lines can be extracted from the electronic map data. The endpoint positions of the long solid lines can include the starting position and / or the ending position of the long solid lines.

[0084] As an example, Figure 5 This is an example diagram of a long solid line scenario provided in an embodiment of this disclosure. For example... Figure 5 As shown, lanes L2 and L3 exist between lanes L1 and L4. Part of the lane divider between lanes L1 and L2 is a long solid line. If a vehicle needs to move from lane L1 to lane L4, it must decelerate and change lanes before the long solid line between lanes L1 and L2. In this scenario, the starting position N1 of the long solid line between lanes L1 and L2 can be extracted.

[0085] In another possible implementation, when the target scenario includes speed measurement scenarios at non-intersection locations, S301 includes: extracting the actual location of the speed measuring device at the non-intersection location from the electronic map data; projecting the actual location onto the road line corresponding to the map road to obtain the projected location. Thus, by projecting the actual location of the speed measuring device, the accuracy of the projected location is improved.

[0086] In this implementation, considering that in speed measurement scenarios at non-intersection locations, vehicles may suddenly decelerate when they are about to reach the speed measuring device, the actual location of the speed measuring device at non-intersection locations is extracted from the electronic map data; the actual location is projected onto the road lines (such as road dividing lines and lane dividing lines) corresponding to the map road to obtain the projected location, and this projected location is used as the key point location in the speed measurement scenario at non-intersection locations.

[0087] The projection process involves making a perpendicular line from a point to a line, which will not be described in detail here.

[0088] As an example, Figure 6 This is an example diagram of a speed measurement scenario at a non-intersection location provided in an embodiment of this disclosure. Figure 6 As shown, a line (i.e., a perpendicular line) can be drawn from the actual position of the speed camera (i.e., the speed measuring device at a non-intersection location) to the adjacent lane dividing line. The intersection of the line and the adjacent lane dividing line is the projected position.

[0089] As an example, Figure 7 An example diagram of a testing device at a non-intersection location provided in an embodiment of this disclosure. (See diagram for reference.) Figure 7 As shown, a speed measuring device is installed above the non-intersection section of the road. Trucks must slow down to the speed limit specified by the road when passing the speed measuring device.

[0090] S302, Determine the road area to be data mined based on the location of key points.

[0091] In this embodiment, after obtaining the location of the key point, since the location of the key point is related to the sudden event, the road range to be mined can be determined on the road where the key point is located, wherein the location of the key point is included in the road valve.

[0092] In one possible implementation, such as Figure 3As shown, S302 includes: S3021, determining the starting point of the road range based on the key point location and the driving direction of the lane where the key point is located, wherein the direction from the starting point to the key point location is the driving direction; S3022, determining the ending point of the road range as the key point location. Thus, based on the characteristic that sudden events are prone to occur when vehicles travel towards the key point location, the road range to be data mined is determined using the driving direction and the key point location, ensuring that the road range covers the area where sudden events are likely to occur, effectively improving the accuracy of the road range determination.

[0093] In this implementation, key points are identified as key locations within the road area. A predetermined distance is traveled from the key point in the opposite direction of traffic flow in the lane containing the key point to obtain the starting point of the road area. This ensures the road area covers the area extending a predetermined distance backward from the key point, i.e., the oncoming traffic zone around the key point. For example, after determining the entrance / exit location, a 2-kilometer radius in the opposite direction of traffic flow can be selected as the road area for data mining.

[0094] S303, obtain historical driving data within the road area.

[0095] The implementation principle and technical effect of S303 can be referred to the aforementioned embodiments, and will not be repeated here.

[0096] S304, perform data mining on changes in driving speed and / or driving direction in historical driving data to obtain the location areas where deceleration and / or lane change occur within the road range. The location areas where deceleration and / or lane change occur are used to assist the vehicle's autonomous driving.

[0097] In this implementation, when the target scenario includes at least one of the following: road entrance / exit scenario, road marking scenario, and speed measurement scenario at non-intersections, vehicles are highly likely to decelerate and / or change lanes in the target scenario. Therefore, data mining can be performed on the changes in driving speed and / or driving direction in the historical driving data within the road area to analyze the location area where vehicles decelerate and / or change lanes within the road area. That is, the location area where deceleration and / or lane change occurs within the road area. Most vehicles that need to decelerate and / or change lanes will decelerate and / or change lanes within this location area when driving within the road area. Using this location area, autonomous vehicles can learn the deceleration and lane-changing habits of human-driven vehicles in the target scenario, and can also more accurately avoid other vehicles in the target scenario.

[0098] Among these methods, data mining is performed on changes in driving speed and / or driving direction in historical driving data within the road area to analyze the process of vehicle deceleration and / or lane change within the road area. Deep learning models can be used for analysis and mining, and there are no restrictions on this method.

[0099] As an example, Figure 8 Examples of road entrance / exit scenarios provided in embodiments of this disclosure Figure 2 .like Figure 8 As shown, in a road scenario where a straight-ahead ring road exit ramp leads to another direction, after determining the location of the exit / exit, a 2-kilometer range can be taken in the opposite direction of traffic. This 2-kilometer range is the road area. By performing data mining on vehicle driving data within this road area, a precise 100-meter range is found where vehicles are likely to slow down and change lanes during their journey. Figure 8 The precise range is indicated by a rectangle. Next to the 100-meter precise range, the historical driving trajectory within that range is shown, reflecting changes in the vehicle's speed.

[0100] In this embodiment of the disclosure, for target scenarios including at least one of road entrance / exit scenarios, road marking scenarios, and speed measurement scenarios at non-intersections, key point locations related to vehicle deceleration and lane changes in the target scenario are extracted from electronic map data. Based on the key point locations, the road range to be data mined is determined, improving the accuracy of the road range. Data mining is performed on historical driving data within the road range to obtain the location areas where deceleration and / or lane changes occur. These location areas are then used to assist in the autonomous driving of the vehicle, improving the safety and passenger experience of the autonomous vehicle in at least one of the road entrance / exit scenarios, road marking scenarios, and speed measurement scenarios at non-intersections, thereby enhancing the intelligence level of the autonomous vehicle.

[0101] In some embodiments, after extracting key point locations from electronic map data, the key point locations can be verified using actual road images at those locations. This verification of key point locations using actual road images improves their accuracy.

[0102] In this embodiment, considering that actual road conditions may change, actual road images at key point locations can be acquired. For example, actual road images at key point locations can be acquired via satellite or by roadside equipment. The key point locations are then matched with the actual road images. Based on the matching results, it is determined whether the key point locations are correct. If incorrect, the key point locations are modified.

[0103] Specifically, if the matching result between the key point location and the actual road image indicates that the key point location does not exist on the actual road image (e.g., there is no entrance / exit location, no speed measuring device, etc.), the key point location can be deleted. If the matching result between the key point location and the actual road image indicates that the actual location of the key point on the actual road image is inconsistent with the key point location extracted from the electronic map data (e.g., the actual entrance / exit location is inconsistent with the entrance / exit location extracted from the electronic map, or the actual endpoint location of the long solid line is inconsistent with the endpoint location of the long solid line extracted from the electronic map), the key point location can be modified to the actual location of the key point.

[0104] In some embodiments, after extracting key point locations from electronic map data, these locations can be stored for future data mining. During storage, key point location identifiers can be associated with the key point locations for easy retrieval. The key point location identifier may include the key point location's sequence number and / or name, and the name can be determined based on the target scene in which the key point location is located.

[0105] As an example, the locations of key points can be stored in the following table format:

[0106] Serial Number Name of key point location Key point location 1 Long solid line starting point coordinates XY 2 Entrance and exit locations coordinates XY 3 Projected position of speed measuring device coordinates XY

[0107] Figure 9 This is a flowchart illustrating the data mining method provided according to embodiments of the present disclosure. Figure 3 .like Figure 9 As shown, the target scenarios include POI scenarios, and the data mining methods include:

[0108] S901, extract the start and end points of adjacent road segments of the POI location from the electronic map data.

[0109] Points of Interest (POIs) include, for example, schools, shopping malls, and tourist attractions.

[0110] One POI location can correspond to one or more adjacent road segments.

[0111] In this embodiment, the location of a Point of Interest (POI) in a POI scene can be determined from the electronic map data, and adjacent road segments can be found based on the POI location. The start and end points of the adjacent road segments are extracted from the electronic map data.

[0112] As an example, Figure 10 Examples of POI scenarios provided in embodiments of this disclosure Figure 1 .like Figure 10As shown, the POI is XX School. Extract the entire road segment adjacent to XX School, including the starting and ending points of the road segment.

[0113] S902, determine the road range to be mined based on the starting and ending points.

[0114] In this embodiment, for each adjacent road segment, the road range to be data mined can be determined within the range formed by the starting position and the ending position of the adjacent road segment.

[0115] In one possible implementation, the starting point of the road range is determined as the starting point of the adjacent road segment; the ending point of the road range is determined as the ending point of the adjacent road segment. Thus, by defining the entire adjacent road segment as the road range to be data mined, data mining is performed on the entire adjacent road segment to analyze the vehicle driving characteristics of the entire adjacent road segment.

[0116] S903, acquires historical driving data within the road area.

[0117] The implementation principle and technical effects of S903 can be referred to in the aforementioned embodiments, and will not be repeated here.

[0118] S904 performs data mining on historical driving data to obtain the mixed pedestrian and vehicle time period corresponding to the road area. The mixed pedestrian and vehicle time period is used to assist the autonomous driving of vehicles.

[0119] In this embodiment, within the road area, due to the influence of POIs (Points of Interest), pedestrian and vehicle mixing occurs during certain time periods, causing inconvenience to driving. For example, when the POI is a school, pedestrian and vehicle mixing occurs during morning school arrival and afternoon school dismissal times. Under such conditions, vehicle speed is significantly affected. Therefore, data mining can be performed on historical driving speed data to obtain the corresponding pedestrian and vehicle mixing situation within the road area.

[0120] In one possible implementation, driving trajectories with driving speeds below a speed threshold and / or parking durations within a certain range are identified in historical driving data; the mixed human-vehicle time period is determined based on the driving time corresponding to the driving trajectory in the historical driving data.

[0121] Among them, there are one or more time periods for mixed pedestrian and vehicle traffic.

[0122] In this implementation, the driving speeds on multiple driving trajectories in historical driving data are compared with speed thresholds. If the driving speed on a trajectory is less than the speed threshold, the driving time corresponding to multiple driving speeds on the trajectory can be obtained. The driving times corresponding to multiple speeds are combined to obtain at least one human-vehicle mixing time period. And / or, the duration of continuous zero speed on multiple driving trajectories in historical driving data is detected to obtain the vehicle's stopping duration on multiple driving trajectories. If the vehicle's stopping duration is within a certain range, it indicates that the vehicle made a short stop. The probability of human-vehicle mixing is high during short vehicle stops. Therefore, the time period during which the vehicle's stopping duration occurred can be determined from the driving time corresponding to the driving trajectory with a stopping duration within the range, and the human-vehicle mixing period includes this time period. Thus, by detecting low-speed driving periods and / or short-term stopping periods of vehicles in the historical driving data of the road area in the POI scenario, the accuracy of the human-vehicle mixing time period is improved.

[0123] As an example, Figure 11 Examples of POI scenarios provided in embodiments of this disclosure Figure 2 .like Figure 11 As shown, data analysis reveals that on the same road segment, the vehicle speed (i.e., travel speed) was low between 07:20 and 08:30, while the vehicle speed between 08:40 and 09:30 was significantly faster than that between 07:20 and 08:30. This suggests that the period between 07:20 and 08:30 likely involved a mix of pedestrians and vehicles. Therefore, the period between 07:20 and 08:30 can be identified as the time period of mixed pedestrian and vehicle traffic.

[0124] In one possible implementation, after obtaining the pedestrian-vehicle mixing time period corresponding to the road area, this time period can be stored so that it can be provided to autonomous driving. When storing the pedestrian-vehicle mixing time period corresponding to the road area, the identifiers corresponding to the road area, the locations corresponding to the road area, and the pedestrian-vehicle mixing time period corresponding to the road area can be stored in a one-to-one correspondence.

[0125] The identifier corresponding to the road range may include the sequence number of the road range and / or the road name of the adjacent road segment corresponding to the road range.

[0126] As an example, the sequence number corresponding to the road range, the road name of the adjacent road segment corresponding to the road range, the location corresponding to the road range, and the time period of mixed pedestrian and vehicle traffic corresponding to the road range can be stored in the following table format:

[0127] Serial Number road name Location Time range 1 XX Road coordinate range 07：20-08：30 2 XX Road coordinate range 15：00-15：50 3 XX Road coordinate range 16：50-17：30

[0128] In this embodiment of the disclosure, for POI scenarios, the road range to be mined is determined by the POI location, improving the accuracy of the road range and thus the accuracy of data mining. Data mining is performed on historical driving data within the road range to obtain the pedestrian-vehicle mixing cycle corresponding to that road range. This provides autonomous driving with the pedestrian-vehicle mixing cycle for specific road segments within POI scenarios, enabling autonomous driving to avoid pedestrian-vehicle mixing cycles or effectively avoid pedestrians and vehicles within those cycles, thereby improving the safety, ride experience, and intelligence of autonomous driving.

[0129] Figure 12 This is a flowchart illustrating the data mining method provided according to embodiments of the present disclosure. Figure 4 .like Figure 12 As shown, when the target scenario includes speed-limited road sections, the data mining methods include:

[0130] S1201, Extract the starting and ending points of the speed-limited road sections from the electronic map data.

[0131] Speed-limited road sections may include sections with restrictions on maximum and / or minimum speeds.

[0132] In this embodiment, speed-limited road sections with speed limit signs can be identified from the electronic map data, and the starting and ending positions of the speed-limited road sections can be extracted.

[0133] S1202, Determine the road range to be mined based on the starting and ending points.

[0134] In this embodiment, the range of roads to be mined can be determined within the area formed by the starting point of the speed-limited road segment and the ending point of the speed-limited road segment.

[0135] In one possible implementation, the starting point of the road area is determined as the starting point of the speed-limited section; the ending point of the road area is determined as the ending point of the speed-limited section. Thus, by defining the entire speed-limited section as the road area to be data-mined, data mining is performed on the entire speed-limited section to analyze the vehicle driving characteristics of the entire speed-limited section.

[0136] S1203, Obtain historical driving data within the road area.

[0137] The implementation principle and technical effect of S1203 can be referred to the aforementioned embodiments, and will not be repeated here.

[0138] S1204 performs data mining on historical driving data to obtain recommended driving speeds for the corresponding road ranges. These recommended driving speeds are used to assist in the autonomous driving of vehicles.

[0139] In this embodiment, a recommended driving speed corresponding to the road range can be determined based on the driving speed in historical driving data, so that the recommended driving speed can be close to the actual driving speed of human-driven vehicles within the road range. This allows the autonomous driving system to learn the driving speed of human-driven vehicles within the road range, thereby improving the rationality of the driving speed of autonomous vehicles in speed-limited road scenarios.

[0140] In one possible implementation, the recommended driving speed can be represented as a recommended speed range. S1204 may include: mining driving speeds on multiple driving trajectories in historical driving data to obtain the recommended speed range. For example, the highest and lowest driving speeds on the driving trajectories can be obtained, and the recommended speed range can be determined based on the highest and lowest driving speeds. Alternatively, the driving speeds on the driving trajectories can be clustered, and the recommended speed range can be determined based on the distance results. This provides a speed range for autonomous vehicles, facilitating flexible decision-making within the speed range and improving the safety, passenger experience, and intelligence of autonomous driving.

[0141] As an example, Figure 13 An example diagram illustrating a speed-limited road section scenario provided in an embodiment of this disclosure. For example... Figure 13 As shown, speed limit signs are set up on speed-limited road sections. The speed limit sign corresponding to the speed limit sign can be identified in the electronic map data, and then the speed limit road section corresponding to the speed limit sign can be determined.

[0142] In this embodiment of the disclosure, for speed-limited road sections, historical driving data of the road area within the speed-limited road section is mined to obtain the recommended driving speed corresponding to the road area. The recommended driving speed assists the autonomous driving of the vehicle, enabling the autonomous driving to drive at a reasonable speed on the speed-limited road section, reducing sudden braking and sudden acceleration of the autonomous driving behavior, and also enabling the autonomous driving vehicle to grasp the driving speed of other vehicles and effectively avoid other vehicles.

[0143] In some embodiments, the speed limit conditions of a road segment can be verified based on actual road images to improve the accuracy of the speed limit.

[0144] In this embodiment, considering that the electronic map data was collected in the past and that real-world road information may have changed, road segments marked with speed limits in the electronic map data may no longer be speed-limited in reality. Therefore, based on the road location information of the speed-limited road segments, actual road images within those segments can be obtained. If speed limit markings (such as speed limit signs or speed limit symbols printed on the road) are identified from the actual road images within the speed-limited road segments, then the speed-limited road segments are determined to meet the speed limit conditions; otherwise, they are determined not to meet the speed limit conditions. If a speed-limited road segment meets the speed limit conditions, data mining can continue relative to that speed-limited road segment.

[0145] In some embodiments, driving characteristic data corresponding to a road range can be stored. Specifically, the identifier corresponding to the road range, the location corresponding to the road range, and the driving characteristic data corresponding to the road range can be stored in a one-to-one correspondence to facilitate the retrieval of the driving characteristic data corresponding to the road range.

[0146] The identifier corresponding to the road range may include the sequence number of the road range and / or the road sign of the road within the road range.

[0147] As an example, the driving characteristic data corresponding to the road area can be stored in the following table:

[0148] Serial Number road name Location Problem Description 1 XX Road coordinate range Easy braking range for human-driven vehicles 2 XX Road coordinate range Areas where drivers and vehicles are prone to lane changes 3 XX Road coordinate range Pedestrian and vehicle mixed traffic cycle 4 XX Road coordinate range Ramp user passage speed range

[0149] The problem description can provide driving characteristic data corresponding to the road range. Among them, the braking range of pedestrians and vehicles and the range where pedestrians and vehicles are likely to change lanes are equivalent to the location range where deceleration and / or lane change occur. The time period of mixed pedestrian and vehicle passage is equivalent to the mixed pedestrian and vehicle time period. The speed range of ramp users is equivalent to the recommended driving speed.

[0150] Figure 14 This is a schematic flowchart of a vehicle control method provided according to an embodiment of the present disclosure. Figure 14 As shown, the vehicle control method includes:

[0151] S1401, Obtain driving characteristic data.

[0152] The driving characteristic data is obtained according to the data mining method provided in any of the foregoing embodiments.

[0153] In this embodiment, driving feature data corresponding to the road range in the target scene can be obtained from the database, or driving feature data corresponding to the road range in the target scene can be received from the data mining device.

[0154] S1402 generates driving decisions based on driving characteristic data and perception data of autonomous vehicles.

[0155] In this embodiment, when it is detected that the vehicle is traveling within the road area of ​​the target scenario, or when it is detected that the vehicle is about to travel within the road area of ​​the target scenario, a driving decision is generated based on driving characteristic data and perception data of the autonomous vehicle. No restrictions are placed on how the driving decision is generated based on the driving characteristic data and perception data of the autonomous vehicle.

[0156] S1403 controls the autonomous vehicle based on driving decisions.

[0157] In this embodiment, driving decisions are sent to the on-board unit of the autonomous vehicle to control the autonomous vehicle to perform autonomous driving behaviors, such as controlling the autonomous vehicle to decelerate or change lanes.

[0158] In this embodiment of the disclosure, the driving feature data mined in the foregoing embodiments is used to assist the autonomous driving of autonomous vehicles, improve the safety of autonomous vehicles and the passenger riding experience, and enhance the intelligence level of autonomous vehicles.

[0159] Figure 15 Schematic diagram of the data mining apparatus provided in the embodiments of this disclosure Figure 1 .like Figure 15 As shown, the data mining apparatus 1500 includes:

[0160] The location extraction unit 1501 is used to extract road location data in a target scene from electronic map data. The target scene is a scene where the probability of an emergency occurring is greater than a set probability.

[0161] The scope determination unit 1502 is used to determine the scope of the road to be data mined based on the road location data.

[0162] Data acquisition unit 1503 is used to acquire historical driving data within the road area;

[0163] The data mining unit 1504 is used to perform data mining on historical driving data to obtain driving feature data corresponding to the road range. The driving feature data is used to assist the autonomous driving of the vehicle.

[0164] Figure 16 Schematic diagram of the data mining apparatus provided in the embodiments of this disclosure Figure 2 .like Figure 16 As shown, the data mining apparatus 1600 includes:

[0165] The location extraction unit 1601 is used to extract road location data in a target scene from electronic map data. The target scene is a scene where the probability of an emergency occurring is greater than a set probability.

[0166] The scope determination unit 1602 is used to determine the scope of the road to be data mined based on the road location data.

[0167] Data acquisition unit 1603 is used to acquire historical driving data within the road area;

[0168] The data mining unit 1604 is used to perform data mining on historical driving data to obtain driving feature data corresponding to the road range. The driving feature data is used to assist the autonomous driving of the vehicle.

[0169] In some embodiments, the target scene includes at least one of road entrance / exit scene, road marking scene, and speed measurement scene at non-intersection. The location extraction unit 1601 includes a key point location extraction module 16011, which is used to extract key point locations in the target scene from electronic map data. The key point locations include at least one of the following: entrance / exit location, road marking location, and the projection location of the speed measurement device at non-intersection on the map road.

[0170] In some embodiments, the range determination unit 1602 includes: a first starting point determination module 16021, configured to determine the starting point position of the road range based on the key point position and the driving direction of the lane where the key point position is located, wherein the direction from the starting point position to the key point position is the driving direction; and a first ending point determination module 16022, configured to determine the ending point position of the road range as the key point position.

[0171] In some embodiments, when the target scenario includes a road entrance / exit scenario, the key point location extraction module 16011 includes: a lane determination submodule (not shown in the figure), used to determine the target lane in the electronic map data, wherein the target lane is a lane that indicates a vehicle to enter an auxiliary road or a ramp from the main road; and an entrance / exit extraction submodule (not shown in the figure), used to determine and extract the entrance / exit location based on the target lane.

[0172] In some embodiments, the entrance / exit extraction submodule is specifically used to: determine the entrance / exit location on the target lane based on the change in the attribute value of the target lane, and extract the entrance / exit location from the electronic map data.

[0173] In some embodiments, when the target scenario includes a road marking scenario, the road marking scenario includes a long solid line scenario, and the road marking location includes the endpoint locations of the long solid line, the key point location extraction module 16011 includes: a long solid line determination submodule, used to determine the long solid line on the map road in the electronic map data; and an endpoint extraction submodule, used to extract the endpoint locations of the long solid line from the electronic map data.

[0174] In some embodiments, when the target scenario includes a speed measurement scenario at a non-intersection location, the key point location extraction module 16011 includes: a speed measurement location extraction submodule (not shown in the figure), used to extract the actual location of the speed measurement device at the non-intersection location from the electronic map data; and a projection location determination submodule, used to project the actual location onto the road line corresponding to the map road to obtain the projection location.

[0175] In some embodiments, the driving feature data includes the location area where deceleration and / or lane change occurs corresponding to the road range. The data mining unit 1604 includes a deceleration and lane change determination module 16041, which is used to perform data mining on the changes in driving speed and / or driving direction in historical driving data to obtain the location area where deceleration and / or lane change occurs.

[0176] In some embodiments, the data mining apparatus further includes a first verification unit 1605, configured to verify the location of the key point based on the actual road image at the key point location.

[0177] In some embodiments, the target scene includes a POI scene, and the road location data includes the start position and end position of the adjacent road segments of the POI location in the POI scene. The range determination unit 1602 includes: a second start position determination module 16023, used to determine the start position of the road range as the start position of the adjacent road segments; and a second end position determination module 16024, used to determine the end position of the road range as the end position of the adjacent road segments.

[0178] In some embodiments, the driving feature data includes the mixed time period of pedestrians and vehicles corresponding to the road range, and the data mining unit 1604 includes: a trajectory recognition module 16042, used to identify driving trajectories in historical driving data where the driving speed is less than the speed threshold and / or the stopping time is within the time range; and a time determination module 16043, used to determine the mixed time period of pedestrians and vehicles based on the driving time corresponding to the driving trajectory in the historical driving data.

[0179] In some embodiments, the target scenario includes a speed-limited road segment scenario, and the road location data includes the starting position and the ending position of the speed-limited road segment. The range determination unit 1602 includes: a third starting point determination module 16025, used to determine the starting position of the road range as the starting position of the speed-limited road segment; and a third ending point determination module 16026, used to determine the ending position of the road range as the ending position of the speed-limited road segment.

[0180] In some embodiments, the driving feature data includes a recommended driving speed corresponding to the road range, and the data mining unit 1604 includes a speed determination module 16044, which is used to determine the recommended driving speed based on the driving speed in historical driving data.

[0181] In some embodiments, the data mining apparatus further includes a second verification unit 1606, used to verify whether the speed limit section meets the speed limit conditions based on the actual road image of the speed limit section.

[0182] Figures 15-16 The provided data mining apparatus can execute the corresponding implementation methods of the above-mentioned data mining methods. Its implementation principle and technical effect are similar, and will not be described again here.

[0183] Figure 17 This is a schematic diagram of the structure of a vehicle control device provided in an embodiment of this disclosure. Figure 17 As shown, the vehicle control device 1700 includes:

[0184] The acquisition unit 1701 is used to acquire driving feature data, wherein the driving feature data is obtained according to the data mining method provided in any of the foregoing embodiments;

[0185] The decision unit 1702 is used to generate driving decisions based on driving characteristic data and perception data of the autonomous vehicle;

[0186] Control unit 1703 is used to control the autonomous vehicle based on driving decisions.

[0187] Figure 17 The provided vehicle control device can execute the corresponding method embodiments of the above-mentioned vehicle control methods. Its implementation principle and technical effect are similar, and will not be described again here.

[0188] 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 disclosure 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 portals are provided for users to choose to authorize or refuse.

[0189] According to embodiments of this disclosure, this disclosure also provides an electronic device, including: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, the instructions being executed by the at least one processor to enable the at least one processor to perform the scheme provided in any of the above embodiments.

[0190] According to embodiments of this disclosure, this disclosure also provides a non-transitory computer-readable storage medium storing computer instructions, wherein the computer instructions are used to cause a computer to execute the scheme provided in any of the above embodiments.

[0191] According to embodiments of this disclosure, this disclosure also provides a computer program product comprising: a computer program stored in a readable storage medium, at least one processor of an electronic device being able to read the computer program from the readable storage medium, and the at least one processor executing the computer program causing the electronic device to perform the scheme provided in any of the above embodiments.

[0192] Figure 18 This is a schematic block diagram of an example electronic device 1800 that can be used to implement embodiments of the present disclosure. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device may also represent various forms of mobile devices, such as personal digital processors, cellular phones, smartphones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely illustrative and are not intended to limit the implementation of the present disclosure described and / or claimed herein.

[0193] like Figure 18 As shown, the electronic device 1800 includes a computing unit 1801, which can perform various appropriate actions and processes based on a computer program stored in a read-only memory (ROM) (e.g., ROM 1802) or a computer program loaded from a storage unit 1808 into a random access memory (RAM) (e.g., RAM 1803). The RAM 1803 may also store various programs and data required for the operation of the electronic device 1800. The computing unit 1801, ROM 1802, and RAM 1803 are interconnected via a bus 1804. An input / output (I / O) interface (e.g., I / O interface 1805) is also connected to the bus 1804.

[0194] Multiple components in electronic device 1800 are connected to I / O interface 1805, including: input unit 1806, such as keyboard, mouse, etc.; output unit 1807, such as various types of displays, speakers, etc.; storage unit 1808, such as disk, optical disk, etc.; and communication unit 1809, such as network card, modem, wireless transceiver, etc. Communication unit 1809 allows electronic device 1800 to exchange information / data with other devices through computer networks such as the Internet and / or various telecommunications networks.

[0195] The computing unit 1801 can be a variety of general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of the computing unit 1801 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special-purpose artificial intelligence (AI) computing chips, various computing units running machine learning model algorithms, digital signal processors (DSPs), and any suitable processor, controller, microcontroller, etc. The computing unit 1801 performs the various methods and processes described above, such as data mining methods and / or vehicle control methods. For example, in some embodiments, the data mining methods and / or vehicle control methods may be implemented as computer software programs tangibly contained in a machine-readable medium, such as storage unit 1808. In some embodiments, part or all of the computer program may be loaded and / or installed on the electronic device 1800 via ROM 1802 and / or communication unit 1809. When the computer program is loaded into RAM 1803 and executed by computing unit 1801, one or more steps of the data mining method and / or vehicle control method described above can be performed. Alternatively, in other embodiments, computing unit 1801 can be configured to perform the data mining method and / or vehicle control method by any other suitable means (e.g., by means of firmware).

[0196] Various embodiments of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field programmable gate arrays (FPGAs), application-specific integrated circuits (ASICs), application-specific standard parts (ASSPs), systems on a chip (SOCs), complex programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments may include implementations in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which may be a dedicated or general-purpose programmable processor, capable of receiving data and instructions from a storage system, at least one input device, and at least one output device, and transmitting data and instructions to the storage system, the at least one input device, and the at least one output device.

[0197] The program code used to implement the methods of this disclosure may be written in any combination of one or more programming languages. This program code may be provided to a processor or controller of a general-purpose computer, special-purpose computer, or other programmable data processing apparatus, such that when executed by the processor or controller, the program code causes the functions / operations specified in the flowcharts and / or block diagrams to be implemented. The program code may be executed entirely on a machine, partially on a machine, as a standalone software package partially on a machine and partially on a remote machine, or entirely on a remote machine or server.

[0198] In the context of this disclosure, a machine-readable medium can be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, apparatus, or device. A machine-readable medium can be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium can be, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination of the foregoing. More specific examples of machine-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, compact disc read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination of the foregoing.

[0199] To provide interaction with a user, the systems and techniques described herein can be implemented on a computer having: a display device for displaying information to the user (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor); and a keyboard and pointing device (e.g., a mouse or trackball) through which the user provides input to the computer. Other types of devices can also be used to provide interaction with the user; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including sound input, voice input, or tactile input).

[0200] The systems and technologies described herein can be implemented in computing systems that include backend components (e.g., as a data server), or computing systems that include middleware components (e.g., an application server), or computing systems that include frontend components (e.g., a user computer with a graphical user interface or web browser through which a user can interact with implementations of the systems and technologies described herein), or any combination of such backend, middleware, or frontend components. The components of the system can be interconnected via digital data communication of any form or medium (e.g., a communication network). Examples of communication networks include local area networks (LANs), wide area networks (WANs), and the Internet.

[0201] Computer systems can include clients and servers. Clients and servers are generally geographically separated and typically interact via communication networks. The client-server relationship is created by computer programs running on the respective computers and having a client-server relationship with each other. A server can be a cloud server, also known as a cloud computing server or cloud host, a hosting product within the cloud computing service system that addresses the management difficulties and weak business scalability inherent in traditional physical hosts and VPS (Virtual Private Server) services. Servers can also be servers for distributed systems or servers integrated with blockchain technology.

[0202] It should be understood that the various forms of processes shown above can be used to rearrange, add, or delete steps. For example, the steps described in this disclosure can be executed in parallel, sequentially, or in different orders, as long as the desired result of the technical solution of this disclosure can be achieved, and this is not limited herein.

[0203] The specific embodiments described above do not constitute a limitation on the scope of protection of this disclosure. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this disclosure should be included within the scope of protection of this disclosure.

Claims

1. A data mining method, comprising: Based on the road characteristics and unexpected events in the target scene, road location data is extracted from electronic map data. The target scene is a scene where the probability of the unexpected event occurring is greater than a set probability. The unexpected events include sudden driving behavior events. Based on the road location data, the scope of roads to be data mined is determined; Obtain historical driving data within the aforementioned road area; Data mining is performed on the historical driving data to obtain driving feature data corresponding to the road range, wherein the driving feature data is used to assist the autonomous driving of the vehicle; The target scene includes a map point of interest (POI) scene, and the road location data includes the start and end points of adjacent road segments of the POI location in the POI scene. Determining the road range to be data mined based on the road location data includes: The starting point of the road range is determined as the starting point of the adjacent road segment; The endpoint of the road range is determined as the endpoint of the adjacent road segment; The scope of the road to be mined is determined based on the starting and ending points. The driving characteristic data includes the time period of mixed pedestrian and vehicle traffic corresponding to the road range. The step of data mining the historical driving data to obtain the driving characteristic data corresponding to the road range includes: In the historical driving data, driving trajectories with driving speeds less than a speed threshold and / or parking durations within a certain range are identified; Based on the driving time corresponding to the driving trajectory in the historical driving data, a pedestrian-vehicle mixing time period corresponding to the road range is determined; the pedestrian-vehicle mixing time period corresponding to the road range is used to enable the autonomous vehicle to avoid the road range during the pedestrian-vehicle mixing time period, or to avoid pedestrians and vehicles within the road range during the pedestrian-vehicle mixing time period; the method further includes: The identifier corresponding to the road range, the location corresponding to the road range, and the driving characteristic data corresponding to the road range are stored accordingly; wherein, the identifier corresponding to the road range includes the sequence number of the road range and / or the road name of the road in which the road range is located.

2. The data mining method according to claim 1, wherein, The target scenario also includes at least one of the following: road entrance / exit scenario, road marking scenario, and speed measurement scenario at non-intersections. The step of extracting road location data from electronic map data based on the road characteristics and unexpected events of the target scenario includes: Extract the locations of key points in the target scene from the electronic map data; The key point locations include at least one of the following: entrance / exit locations, road marking locations, and the projection location of speed measuring devices on the map road at non-intersection locations.

3. The data mining method according to claim 2, wherein, The step of determining the road range to be data mined based on the road location data includes: Based on the location of the key point and the driving direction of the lane where the key point is located, the starting point of the road range is determined, wherein the direction from the starting point to the key point is the driving direction; The endpoint of the road range is determined as the key point location.

4. The data mining method according to claim 2, wherein, When the target scene includes a road entrance / exit scene, extracting the key point locations in the target scene from the electronic map data includes: In the electronic map data, a target lane is determined, wherein the target lane is a lane that indicates a vehicle's entry from the main road into an auxiliary road or a ramp; Based on the target lane, determine and extract the entrance / exit location.

5. The data mining method according to claim 4, wherein, The step of determining and extracting the entrance / exit location based on the target lane includes: Based on the changes in the attribute values ​​of the target lane, the location of the entrance / exit is determined on the target lane, and the location of the entrance / exit is extracted from the electronic map data.

6. The data mining method according to claim 2, wherein, When the target scene includes a road marking scene, and the road marking scene includes a long solid line scene, and the road marking location includes the endpoint location of the long solid line, the step of extracting the key point location in the target scene from the electronic map data includes: In the electronic map data, determine the long solid lines on the map roads; Extract the endpoint positions of the long solid line from the electronic map data.

7. The data mining method according to claim 2, wherein, When the target scenario includes speed measurement scenarios outside intersections, the step of extracting key point locations in the target scenario from the electronic map data includes: Extract the actual locations of speed measuring devices at non-intersection locations from the electronic map data; The actual location is projected onto the road line corresponding to the road on the map to obtain the projected location.

8. The data mining method according to any one of claims 2 to 7, wherein, The driving characteristic data includes the location areas where deceleration and / or lane changes occur corresponding to the road range. The step of data mining the historical driving data to obtain the driving characteristic data corresponding to the road range includes: Data mining is performed on the changes in driving speed and / or driving direction in the historical driving data to obtain the location areas where deceleration and / or lane change occurred.

9. The data mining method according to any one of claims 2 to 7, further comprising, after extracting the key point locations of the target scene from the electronic map data: The location of the key point is verified based on the actual road image at the key point location.

10. The data mining method according to any one of claims 1 to 7, wherein, The target scenario also includes speed-limited road segment scenarios. The road location data includes the starting location and ending location of the speed-limited road segment. Determining the road range to be data mined based on the road location data includes: The starting point of the road range is determined as the starting point of the speed-limited road section; The endpoint of the road range is determined as the endpoint of the speed-limited road section.

11. The data mining method according to claim 10, wherein, The driving characteristic data includes the recommended driving speed corresponding to the road range. The step of data mining the historical driving data to obtain the driving characteristic data corresponding to the road range includes: The recommended driving speed is determined based on the driving speed in the historical driving data.

12. The data mining method according to claim 10, further comprising: Based on the actual road images of the speed-limited road sections, it is verified whether the speed-limited road sections meet the speed limit conditions.

13. A vehicle control method, comprising: Acquire driving feature data, wherein the driving feature data is obtained according to the data mining method as described in any one of claims 1 to 12; Based on the driving characteristic data and the perception data of the autonomous vehicle, a driving decision is generated; Based on the driving decision, control the autonomous vehicle.

14. A data mining apparatus, comprising: The location extraction unit is used to extract road location data from electronic map data based on the road characteristics of the target scene and unexpected events. The target scene is a scene where the probability of the unexpected event occurring is greater than a set probability. The unexpected events include sudden driving behavior events. The scope determination unit is used to determine the scope of the road to be data mined based on the road location data. The data acquisition unit is used to acquire historical driving data within the road area; A data mining unit is used to perform data mining on the historical driving data to obtain driving feature data corresponding to the road range, wherein the driving feature data is used to assist the autonomous driving of the vehicle. The target scene includes a Point of Interest (POI) scene, the road location data includes the start position and end position of adjacent road segments of the POI location in the POI scene, and the range determination unit includes: The second starting point determination module is used to determine the starting point position of the road range as the starting point position of the adjacent road segment; The second endpoint determination module is used to determine the endpoint position of the road range as the endpoint position of the adjacent road segment; The driving characteristic data includes the time period of pedestrian-vehicle mixing corresponding to the road range, and the data mining unit includes: The trajectory recognition module is used to identify driving trajectories in the historical driving data where the driving speed is less than a speed threshold and / or the dwell time is within a certain range. The time determination module is used to determine the pedestrian and vehicle mixing time period corresponding to the road range based on the driving time corresponding to the driving trajectory in the historical driving data; the pedestrian and vehicle mixing time period corresponding to the road range is used to enable the autonomous vehicle to avoid the road range within the pedestrian and vehicle mixing time period, or to avoid pedestrians and vehicles within the road range within the pedestrian and vehicle mixing time period. The device further includes: The storage module is used to store the identifier corresponding to the road range, the location corresponding to the road range, and the driving characteristic data corresponding to the road range; wherein, the identifier corresponding to the road range includes the sequence number of the road range and / or the road name of the road in which the road range is located.

15. The data mining apparatus according to claim 14, wherein, The target scenario also includes at least one of the following: road entrance / exit scenario, road marking scenario, and speed measurement scenario at non-intersection locations. The location extraction unit includes: The key point location extraction module is used to extract the locations of key points in the target scene from the electronic map data; The key point locations include at least one of the following: entrance / exit locations, road marking locations, and the projection location of speed measuring devices on the map road at non-intersection locations.

16. The data mining apparatus according to claim 15, wherein, The range determination unit includes: The first starting point determination module is used to determine the starting point of the road range based on the key point location and the driving direction of the lane where the key point location is located, wherein the direction from the starting point location to the key point location is the driving direction; The first endpoint determination module is used to determine the endpoint location of the road range as the key point location.

17. The data mining apparatus according to claim 15, wherein, When the target scene includes a road entrance / exit scene, the key point location extraction module includes: The lane determination submodule is used to determine the target lane in the electronic map data, wherein the target lane is a lane that indicates a vehicle to enter an auxiliary road or ramp from the main road; The entrance / exit extraction submodule is used to determine and extract the location of the entrance / exit based on the target lane.

18. The data mining apparatus according to claim 17, wherein, The inlet / outlet extraction submodule is specifically used for: Based on the changes in the attribute values ​​of the target lane, the location of the entrance / exit is determined on the target lane, and the location of the entrance / exit is extracted from the electronic map data.

19. The data mining apparatus according to claim 15, wherein, In the case where the target scene includes a road marking scene, the road marking scene includes a long solid line scene, the road marking position includes the endpoint position of the long solid line, and the key point position extraction module includes: The long solid line determination submodule is used to determine long solid lines on the map roads in the electronic map data; The endpoint extraction submodule is used to extract the endpoint positions of the long solid line from the electronic map data.

20. The data mining apparatus according to claim 15, wherein, In the case where the target scenario includes speed measurement scenarios outside of intersections, the key point location extraction module includes: The speed measurement location extraction submodule is used to extract the actual location of the speed measurement device at non-intersection locations from the electronic map data; The projection position determination submodule is used to project the actual position onto the road line corresponding to the map road to obtain the projection position.

21. The data mining apparatus according to any one of claims 15 to 20, wherein, The driving characteristic data includes the location areas where deceleration and / or lane changes occur corresponding to the road range, and the data mining unit includes: The deceleration and lane change determination module is used to perform data mining on the changes in driving speed and / or driving direction in the historical driving data to obtain the location area where deceleration and / or lane change occur.

22. The data mining apparatus according to any one of claims 15 to 20, further comprising: The first verification unit is used to verify the location of the key point based on the actual road image at the location of the key point.

23. The data mining apparatus according to any one of claims 14 to 20, wherein, The target scenario includes a speed-limited road section scenario, the road location data includes the starting position and the ending position of the speed-limited road section, and the range determination unit includes: The third starting point determination module is used to determine the starting point of the road range as the starting point of the speed-limited road section; The third endpoint determination module is used to determine the endpoint location of the road range as the endpoint location of the speed-limited road section.

24. The data mining apparatus according to claim 23, wherein, The driving feature data includes the recommended driving speed corresponding to the road range, and the data mining unit includes: The speed determination module is used to determine the recommended driving speed based on the driving speed in the historical driving data.

25. The data mining apparatus according to claim 23, further comprising: The second verification unit is used to verify whether the speed-limited road section meets the speed limit conditions based on the actual road image of the speed-limited road section.

26. A vehicle control device, comprising: An acquisition unit is used to acquire driving feature data, wherein the driving feature data is obtained according to the data mining method as described in any one of claims 1 to 12; The decision-making unit is used to generate driving decisions based on the driving characteristic data and the perception data of the autonomous vehicle; A control unit for controlling the autonomous vehicle based on the driving decision.

27. An electronic device comprising: At least one processor; as well as A memory communicatively connected to the at least one processor; wherein, The memory stores instructions executable by the at least one processor, which, when executed by the at least one processor, enables the at least one processor to perform the data mining method according to any one of claims 1 to 12 and / or the vehicle control method according to claim 13.

28. A non-transitory computer-readable storage medium storing computer instructions, wherein, The computer instructions are used to cause the computer to execute the data mining method according to any one of claims 1 to 12 and / or execute the vehicle control method according to claim 13.

29. A computer program product comprising a computer program that, when executed by a processor, implements the steps of the data mining method according to any one of claims 1 to 12 and / or the steps of the vehicle control method according to claim 13.

Images