Driving method and device of autonomous vehicle, electronic equipment and computer readable storage medium
By integrating the target detection results of the bicycle with memory map information, the target detection of autonomous driving vehicles is optimized, and the problem of difficult target recognition is solved, and driving safety and comfort are improved.
Patent Information
- Application Number
- CN202411333578.9
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-09-24
- Publication Date
- 2024-12-06
AI Technical Summary
In autonomous driving systems, difficult-to-detect targets such as specific traffic signs, rare road obstacles, special-shaped vehicles and certain types of pedestrians are not obvious, making it difficult for traditional detection algorithms to accurately identify and classify, affecting the autonomous driving system. The decision-making process may lead to safety hazards and passenger comfort issues.
By obtaining the target detection results of the bicycle and the memory map information for fusion, the target detection is optimized using the historical target reference information in the memory map to improve the detection accuracy, and the driving strategy is determined based on the fused target information to control the vehicle's driving.
It improves the detection accuracy of autonomous driving vehicles for difficult-to-detection targets, optimizes driving strategies, and enhances driving safety and comfort in scenarios facing difficult-to-detection targets.
Smart Images

Figure CN119078882A_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the field of autonomous driving technology, and in particular to a driving method, device, electronic device, and computer-readable storage medium for an autonomous driving vehicle. Background Art
[0002] In autonomous driving systems, it is difficult to detect some targets due to their low frequency of appearance in real-world scenarios and sparse LiDAR and visual features.
[0003] These rare targets include, for example, certain types of traffic signs, rare road obstacles, special-shaped vehicles, and certain types of pedestrians. Since the features of these targets are not obvious, traditional detection algorithms are difficult to accurately identify and classify them, which in turn affects the decision-making process of the autonomous driving system. This detection difficulty may not only lead to safety hazards, such as failure to avoid obstacles in time, but also affect the comfort of passengers, such as sudden braking or oversteering in unnecessary places. Therefore, improving the detection ability of these rare targets is an important research direction to ensure the safety and comfort of the self-driving car.
[0004] Currently, there are no good detection methods for these targets in the existing technology, and they usually need to be solved by continuously collecting data and strengthening models. However, due to the difficulty in obtaining data for these targets, the model optimization and solution cycle is relatively long. Summary of the invention
[0005] The embodiments of the present application provide a driving method, device, electronic device, and computer-readable storage medium for an autonomous driving vehicle to improve target detection effects and the safety of the autonomous driving vehicle.
[0006] The present application embodiment adopts the following technical solutions:
[0007] In a first aspect, an embodiment of the present application provides a driving method of an autonomous driving vehicle, wherein the driving method of the autonomous driving vehicle is executed by a vehicle side, and the driving method of the autonomous driving vehicle includes:
[0008] Obtaining a target detection result of the vehicle and determining target information output by the vehicle according to the target detection result of the vehicle;
[0009] Acquire memory map information, wherein the memory map information includes target reference information provided by other vehicles or the vehicle in history;
[0010] Fusing the memory map information with the target information output by the vehicle to obtain fused target information;
[0011] The driving strategy of the vehicle is determined according to the fused target information, and the driving of the vehicle is controlled according to the driving strategy of the vehicle.
[0012] Optionally, the acquiring the target detection result of the vehicle and determining the target information output by the vehicle according to the target detection result of the vehicle includes:
[0013] Create and maintain raster maps in a global coordinate system;
[0014] Acquire the sensor perception data of the vehicle and perform target detection to obtain the target detection result of the vehicle;
[0015] updating the grid map according to the target detection result to obtain updated grid map information;
[0016] The target detection result is updated according to the relative distance and relative viewing angle between the vehicle and the target detected in the target detection result, an updated target detection result is obtained and updated target features are recorded.
[0017] Optionally, after obtaining the target detection result of the self-vehicle and determining the target information output by the self-vehicle according to the target detection result of the self-vehicle, the driving method of the automatic driving vehicle further includes:
[0018] Determine whether the vehicle currently meets the preset data uploading conditions;
[0019] When the preset data uploading conditions are met, the data to be uploaded is determined according to the target information output by the vehicle, and the data to be uploaded is uploaded to the cloud or the road side, so that the cloud or the road side updates the memory map information according to the data to be uploaded.
[0020] Optionally, determining whether the vehicle currently meets a preset data uploading condition includes:
[0021] Determine whether the ego vehicle has detected a difficult-to-detect target based on the target detection result of the ego vehicle;
[0022] If so, it is determined that the vehicle currently meets the preset data uploading condition;
[0023] Otherwise, it is determined that the vehicle currently does not meet the preset data uploading condition;
[0024] When the preset data uploading condition is met, determining the data to be uploaded according to the target information output by the vehicle, and uploading the data to be uploaded to the cloud or the road side includes:
[0025] When the preset data upload conditions are met, the target detection results of the vehicle are tracked and compared to obtain target difference information;
[0026] The target difference information is uploaded to the cloud or the road side.
[0027] Optionally, acquiring memory map information includes:
[0028] Obtain the current location information of the vehicle and upload it to the cloud or road side;
[0029] Receive the memory map information corresponding to the current positioning information sent by the cloud or road end.
[0030] Optionally, the target information output by the vehicle includes a target position detected by the vehicle, and fusing the memory map information with the target information output by the vehicle to obtain the fused target information includes:
[0031] Determining whether there is corresponding target information output by the ego vehicle at the target position in the target reference information according to the target position in the target reference information and the target position detected by the ego vehicle;
[0032] When there is corresponding target information output by the ego vehicle at the target position in the target reference information, the target reference information is fused with the corresponding target information output by the ego vehicle to obtain fused target information;
[0033] The driving method of the autonomous driving vehicle further includes:
[0034] When there is no corresponding target information output by the vehicle at the target position in the target reference information, the target position is marked as the position of a potential target, and the target information of the potential target is determined according to the target reference information.
[0035] Optionally, the fused target information includes fused target information of difficult-to-detect targets, and determining the driving strategy of the vehicle according to the fused target information, and controlling the driving of the vehicle according to the driving strategy of the vehicle includes:
[0036] Determining, according to the fused target information of the difficult-to-detect target, that the self-vehicle driving strategy is a driving strategy set for the difficult-to-detect target, and controlling the driving of the self-vehicle according to the driving strategy set for the difficult-to-detect target;
[0037] The driving method of the autonomous driving vehicle further includes:
[0038] The vehicle driving strategy is determined to be a driving strategy set for the potential target according to the target information of the potential target, and the vehicle driving is controlled according to the driving strategy set for the potential target.
[0039] In a second aspect, an embodiment of the present application further provides a driving device of an autonomous driving vehicle, wherein the driving device of the autonomous driving vehicle is applied to a vehicle side, and the driving device of the autonomous driving vehicle includes:
[0040] A first acquisition unit, used to acquire a target detection result of the vehicle and determine target information output by the vehicle according to the target detection result of the vehicle;
[0041] A second acquisition unit is used to acquire memory map information, wherein the memory map information includes target reference information provided by other vehicles or the vehicle in history;
[0042] a fusion unit, used for fusing the memory map information with the target information output by the vehicle to obtain fused target information;
[0043] The control unit is used to determine the driving strategy of the vehicle according to the fused target information, and control the driving of the vehicle according to the driving strategy of the vehicle.
[0044] In a third aspect, an embodiment of the present application also provides a driving system for an autonomous driving vehicle, wherein the driving system for the autonomous driving vehicle includes a vehicle side and a cloud side and / or a road side, and the vehicle side is used to execute any of the aforementioned driving methods for the autonomous driving vehicle.
[0045] In a fourth aspect, an embodiment of the present application further provides an electronic device, including:
[0046] Processor; and
[0047] A memory arranged to store computer executable instructions, which, when executed, cause the processor to perform any of the aforementioned methods for driving an autonomous vehicle.
[0048] In a fifth aspect, an embodiment of the present application also provides a computer-readable storage medium, which stores one or more programs. When the one or more programs are executed by an electronic device including multiple application programs, the electronic device executes any of the aforementioned driving methods for the autonomous driving vehicle.
[0049] At least one of the above technical solutions adopted in the embodiments of the present application can achieve the following beneficial effects: the driving method of the autonomous driving vehicle in the embodiments of the present application is executed by the vehicle side, first obtaining the target detection result of the self-vehicle and determining the target information output by the self-vehicle according to the target detection result of the self-vehicle; then obtaining the memory map information, which contains the target reference information provided by other vehicles or the history of the self-vehicle; then fusing the memory map information with the target information output by the self-vehicle to obtain the fused target information; finally determining the driving strategy of the self-vehicle according to the fused target information, and controlling the driving of the self-vehicle according to the driving strategy of the self-vehicle. The driving method of the autonomous driving vehicle in the embodiments of the present application aims at the problem of difficult-to-detect targets in the autonomous driving scenario, and adopts a method of fusing the target information currently perceived by the self-vehicle with the target reference information provided by other vehicles or the history of the self-vehicle, optimizing the target information currently detected by the self-vehicle by sharing the target information, improving the target detection accuracy of the self-vehicle, and facilitating the self-vehicle to plan the corresponding driving strategy in advance, so that the current vehicle can drive more smoothly, and improving the driving safety of the autonomous driving vehicle in such scenarios. BRIEF DESCRIPTION OF THE DRAWINGS
[0050] The drawings described herein are used to provide a further understanding of the present application and constitute a part of the present application. The illustrative embodiments of the present application and their descriptions are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings:
[0051] Figure 1 A schematic diagram of a process flow of a driving method for an autonomous driving vehicle in an embodiment of the present application;
[0052] Figure 2 This is a schematic diagram of the structure of a driving device of an autonomous driving vehicle in an embodiment of the present application;
[0053] Figure 3 This is a schematic diagram of the structure of an electronic device in an embodiment of the present application. DETAILED DESCRIPTION
[0054] In order to make the purpose, technical solution and advantages of the present application clearer, the technical solution of the present application will be clearly and completely described below in combination with the specific embodiments of the present application and the corresponding drawings. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present application.
[0055] The technical solutions provided by various embodiments of the present application are described in detail below in conjunction with the accompanying drawings.
[0056] The present application embodiment provides a driving method for an autonomous driving vehicle, such as Figure 1As shown, a flow chart of a driving method of an autonomous driving vehicle in an embodiment of the present application is provided. The driving method of the autonomous driving vehicle is executed by the vehicle side, and the driving method of the autonomous driving vehicle at least includes the following steps S110 to S140:
[0057] Step S110, obtaining a target detection result of the own vehicle and determining target information output by the own vehicle according to the target detection result of the own vehicle.
[0058] The driving method of the autonomous driving vehicle in the embodiment of the present application is executed by the vehicle side. During the driving process of the vehicle, the target detection results output by the vehicle are obtained in real time. The target detection results here can be the results obtained by multiple sensors on the vehicle side using existing target detection models. For example, they can include point cloud target detection results output by a lidar perception module, image target detection results output by a camera perception module, etc., and can also be fusion detection results of multiple sensors.
[0059] The target detection result of the vehicle may specifically include information such as the detected target location and category. Based on the target detection result, other dimensions of target information may be further obtained, such as the relative distance and relative angle between the target and the vehicle. Therefore, this information may be integrated as the target information output by the vehicle. Of course, it should be noted that the specific target information included may be flexibly set according to actual needs and is not specifically limited here.
[0060] Step S120, obtaining memory map information, wherein the memory map information includes target reference information provided by other vehicles or the vehicle in history.
[0061] Considering that there may be some typical and common difficult-to-detect targets in actual road scenes, such as irregular tree crowns, construction barriers, overlong obstacles, haystacks, and cement piles, the reason why these targets are difficult to detect may be due to the occlusion of the vehicle itself, the occlusion of other vehicles passing by, the sparseness of the targets, etc., resulting in unclear features of such targets, and it is difficult to accurately identify and classify targets simply by relying on traditional detection algorithms. At the same time, considering that the external environment or the state of the autonomous driving vehicle may change in real time, different vehicles or the same vehicle passing through the same area or location at different times may have different detection effects for the same target.
[0062] Therefore, in order to make up for the problem that it is difficult to accurately identify the above-mentioned difficult-to-detect targets by relying solely on the perception information of the vehicle model, the embodiment of the present application further obtains the memory map information corresponding to the current area or position of the vehicle based on the target information output by the vehicle. The memory map information can be understood as being generated based on the grid map constructed in the global coordinate system, and is mainly used to record the target information provided by other vehicles in a certain area in front of the vehicle or the target information recorded when the vehicle repeatedly drove in the area. It is called "reference" information because this information mainly provides the vehicle with richer and more comprehensive target information based on the current perception information of the vehicle, and assists the vehicle in more accurate target detection.
[0063] Step S130: fusing the memory map information with the target information output by the vehicle to obtain fused target information.
[0064] The target information output by the ego vehicle obtained in the above steps is integrated with the target reference information in the memory map information, thereby enriching the target information so that the ego vehicle can obtain more accurate and stable target information.
[0065] Step S140, determining a driving strategy of the vehicle according to the fused target information, and controlling the driving of the vehicle according to the driving strategy of the vehicle.
[0066] After obtaining the fused target information, the planning and control module of the vehicle can plan the current driving strategy of the vehicle based on the more accurate and stable fused target information. For example, it may include driving strategies designed for difficult-to-detect targets encountered by the vehicle, or it may include driving strategies designed for some uncertain targets. The vehicle's driving is controlled according to the currently determined driving strategy.
[0067] The driving method of the autonomous driving vehicle in the embodiment of the present application addresses the problem of difficult-to-detect targets in autonomous driving scenarios by integrating target information currently perceived by the vehicle with target reference information provided by other vehicles or the vehicle's history. By sharing target information, the target information currently detected by the vehicle is optimized, the target detection accuracy of the vehicle is improved, and the vehicle is facilitated to plan corresponding driving strategies in advance, so that the current vehicle can travel more smoothly, thereby improving the driving safety of autonomous driving vehicles in such scenarios.
[0068] In some embodiments of the present application, obtaining the target detection result of the own vehicle and determining the target information output by the own vehicle based on the target detection result of the own vehicle includes: creating and maintaining a grid map in a global coordinate system; obtaining sensor perception data of the own vehicle and performing target detection to obtain the target detection result of the own vehicle; updating the grid map based on the target detection result to obtain updated grid map information; updating the target detection result based on the relative distance and relative viewing angle between the own vehicle and the target detected in the target detection result to obtain an updated target detection result and record updated target features.
[0069] When the vehicle in the embodiment of the present application performs target detection, it can first maintain a grid map under the global coordinate system. For example, a high-precision map can be used as a basis, combined with the vehicle's own positioning information such as GPS, IMU, wheel speed meter, etc., to map the vehicle's target detection results such as obstacles, road signs, etc. into the global coordinate system.
[0070] According to the target detection results detected by the vehicle in real time, the grid map is dynamically updated and the target features of each grid are recorded. Here, the probability occupancy grid map can be used for updating. The probability occupancy grid map divides the environment into a series of grids (Cells) and assigns a value representing the probability of occupancy to each grid. This value is usually between 0 and 1, where 0 indicates that the probability of the grid being idle (i.e., unoccupied) is the highest, and 1 indicates that the probability of the grid being occupied is the highest. In this way, the map not only represents the spatial structure of the environment, but also contains the uncertainty information of the occupation of each position.
[0071] The target detection results output by the vehicle, the updated grid map, the target features in the grid, etc. can all be used as an aspect of the target information output by the vehicle. In addition, the embodiment of the present application further defines the confidence of the target information. The confidence of the target information here does not refer to the confidence of the original output of the target detection model, but is a value calculated by combining the confidence of the original output of the target detection model and other judgment conditions such as the relative distance between the vehicle and the target, the relative viewing angle, etc., which can measure the credibility of the target information more comprehensively and accurately.
[0072] Specifically, since the ego vehicle is in a real-time driving state, on the one hand, as the distance between the ego vehicle and the target shortens, the confidence of the target information can be improved through more precise scanning and richer feature extraction. On the other hand, recording the relative viewing angle from the ego vehicle to the target can gradually improve the confidence as the cumulative viewing angle range increases.
[0073] In some embodiments of the present application, after obtaining the target detection result of the self-vehicle and determining the target information output by the self-vehicle based on the target detection result of the self-vehicle, the driving method of the autonomous driving vehicle also includes: determining whether the self-vehicle currently meets the preset data upload conditions; when the preset data upload conditions are met, determining the data to be uploaded based on the target information output by the self-vehicle, and uploading the data to be uploaded to the cloud or the road side, so that the cloud or the road side updates the memory map information according to the data to be uploaded.
[0074] The memory map information of the embodiment of the present application can be maintained and updated by the cloud or the road side. For the vehicle itself, the data currently collected and the information perceived by the vehicle itself may also provide a certain reference for the subsequent driving of other vehicles. Therefore, the vehicle itself can also upload the data collected and the information perceived by itself to the cloud or the road side to facilitate the cloud or the road side to update the memory map information.
[0075] Considering that the amount of data generated by the vehicle may be large, if all data is uploaded in real time, it may affect the performance and transmission efficiency of the vehicle. Therefore, the embodiment of the present application can first determine whether the vehicle currently meets the conditions for data upload. The determination of the data upload conditions will constrain the amount of data uploaded by the vehicle to a certain extent, thereby avoiding the vehicle end from affecting performance and transmission efficiency due to excessive uploaded data. If the vehicle currently meets the data upload conditions, the data to be uploaded can be determined based on the target information output by the vehicle. Of course, the specific target information to be uploaded can be flexibly set by those skilled in the art according to actual needs, and no specific limitation is made here.
[0076] In some embodiments of the present application, determining whether the vehicle currently meets the preset data upload conditions includes: determining whether the vehicle detects a difficult-to-detect target based on the target detection results of the vehicle; if so, determining that the vehicle currently meets the preset data upload conditions; otherwise, determining that the vehicle currently does not meet the preset data upload conditions; when the preset data upload conditions are met, determining the data to be uploaded based on the target information output by the vehicle, and uploading the data to be uploaded to the cloud or the road side includes: when the preset data upload conditions are met, tracking and comparing the target detection results of the vehicle to obtain target difference information; uploading the target difference information to the cloud or the road side.
[0077] Based on the above embodiments, since the traditional target detection model is not accurate enough in identifying the difficult-to-detect targets in the scene, it is difficult for the autonomous driving vehicle to fully consider the existence of the difficult-to-detect targets during driving. Therefore, the memory map information stored in the cloud or road side in the embodiment of the present application mainly maintains the information of the difficult-to-detect targets encountered by each vehicle during driving. This information comes from the results reported by each vehicle, and as the vehicle travels, the distance between the vehicle and the difficult-to-detect target gradually approaches and the cumulative viewing angle gradually increases, etc., so that the vehicle can obtain better target detection effects, so that more complete and accurate information such as the location, shape, and characteristics of the difficult-to-detect targets can be uploaded to the cloud or road side, thereby providing a more accurate reference for other subsequent vehicles.
[0078] Based on this, the vehicle can first determine whether a difficult-to-detect target is currently detected, which can be determined by combining the confidence of the original output of the target detection model, the distance and relative viewing angle between the vehicle and the target, the size of the target's bounding box, and other factors. When a difficult-to-detect target is detected, the corresponding difficult-to-detect target information is uploaded, and as the vehicle continues to drive, the updated difficult-to-detect target information is continuously uploaded. The specific updated information can be determined by continuously tracking and comparing the difficult-to-detect target information output by the vehicle, and the difference is reported, thereby improving and updating the information stored in the cloud or on the road.
[0079] In some embodiments of the present application, the obtaining of memory map information includes: obtaining the current positioning information of the vehicle and uploading it to the cloud or the road side; and receiving the memory map information corresponding to the current positioning information sent by the cloud or the road side.
[0080] Since the memory map information maintained by the cloud or road side is complete map information under the global coordinate system, when requesting the memory map information from the cloud or road side, the current positioning information of the vehicle can be obtained first and reported to the cloud or road side. In this way, the cloud or road side can obtain the memory map information within the local area according to the current position of the vehicle and return it to the vehicle side, avoiding the problem of inefficient processing of global map information by the vehicle side to meet the real-time requirements of autonomous driving.
[0081] In some embodiments of the present application, the target information output by the self-vehicle includes a target position detected by the self-vehicle, and the memory map information is fused with the target information output by the self-vehicle to obtain the fused target information, including: determining whether there is corresponding target information output by the self-vehicle at the target position in the target reference information according to the target position in the target reference information and the target position detected by the self-vehicle; in the case where there is corresponding target information output by the self-vehicle at the target position in the target reference information, fusing the target reference information with the corresponding target information output by the self-vehicle to obtain the fused target information; the driving method of the autonomous driving vehicle also includes: in the case where there is no corresponding target information output by the self-vehicle at the target position in the target reference information, marking the target position as the position of a potential target, and determining the target information of the potential target according to the target reference information.
[0082] When fusing the memory map information with the target information output by the ego vehicle, the target position in the target reference information can be compared with the target position in the target information output by the ego vehicle. If the ego vehicle also outputs the corresponding target detection result at the target position in the target reference information, then the target reference information can be directly fused with the corresponding target information output by the ego vehicle. Of course, two situations can be further distinguished here. One situation is that the other vehicle believes that the target detected at this position is a difficult target, but the ego vehicle believes that the target at this position is not a difficult target. The other situation is that both the other vehicle and the ego vehicle believe that the target at this position is a difficult target. For the first situation, when fusing, the target detection result output by the ego vehicle can be trusted more, for example, a higher weight can be given to the target detection result of the ego vehicle. For the second situation, the fusion weight of the target reference information can be appropriately increased relative to the first situation.
[0083] In addition, when fusing, factors such as the update time of the target reference information, the degree of match between the target in the target reference information and the target detected by the own vehicle, etc. can be considered to determine the fusion weight. If the target reference information has not been updated for a long time, and the degree of match between the target in the target reference information and the target detected by the own vehicle is low, the detection result of the own vehicle is updated, that is, the fusion weight of the target reference information is correspondingly reduced.
[0084] If the vehicle does not output a corresponding target detection result at the target position in the target reference information, it means that other vehicles have detected a difficult-to-detect target at this position, but the vehicle has not detected it at present. In order to ensure driving safety, the position can be marked to indicate the suspected presence of a difficult-to-detect target, i.e., a potential target, and the target reference information at this position can be used as the target information of the potential target.
[0085] In some embodiments of the present application, the fused target information includes the target information of the fused difficult-to-detect target, and determining the driving strategy of the self-vehicle according to the fused target information, and controlling the driving of the self-vehicle according to the driving strategy of the self-vehicle includes: determining the driving strategy of the self-vehicle as a driving strategy set for the difficult-to-detect target according to the target information of the fused difficult-to-detect target, and controlling the driving of the self-vehicle according to the driving strategy set for the difficult-to-detect target; the driving method of the autonomous driving vehicle also includes: determining the driving strategy of the self-vehicle as a driving strategy set for the potential target according to the target information of the potential target, and controlling the driving of the self-vehicle according to the driving strategy set for the potential target.
[0086] During the planning and control stage, the planning and control module of the vehicle can make differentiated responses to different targets based on the optimized target information. For example, for fused, more stable and difficult-to-detect targets, avoidance planning can be performed at a longer distance, making the driving trajectory of the vehicle smoother. For potential targets, it is possible to slow down in advance or consider the possibility of the potential target when other decisions and plans are needed, thereby reducing safety and somatosensory problems caused by sudden detection of the target. Of course, the specific driving strategy can be flexibly designed by technicians in this field in combination with the actual scenario, and no specific limitation is made here.
[0087] The present application also provides a driving device 200 for an autonomous driving vehicle, such as Figure 2 As shown, a schematic diagram of the structure of a driving device of an autonomous driving vehicle in an embodiment of the present application is provided. The driving device of the autonomous driving vehicle is applied to the vehicle side. The driving device 200 of the autonomous driving vehicle includes: a first acquisition unit 210, a second acquisition unit 220, a fusion unit 230 and a control unit 240, wherein:
[0088] A first acquisition unit 210, configured to acquire a target detection result of the vehicle and determine target information output by the vehicle according to the target detection result of the vehicle;
[0089] A second acquisition unit 220 is used to acquire memory map information, wherein the memory map information includes target reference information provided by other vehicles or the vehicle in history;
[0090] A fusion unit 230, configured to fuse the memory map information with the target information output by the vehicle to obtain fused target information;
[0091] The control unit 240 is used to determine the driving strategy of the vehicle according to the fused target information, and control the driving of the vehicle according to the driving strategy of the vehicle.
[0092] In some embodiments of the present application, the first acquisition unit 210 is specifically used to: create and maintain a grid map in a global coordinate system; acquire sensor perception data of the vehicle and perform target detection to obtain a target detection result of the vehicle; update the grid map according to the target detection result to obtain updated grid map information; update the target detection result according to the relative distance and relative viewing angle between the vehicle and the target detected in the target detection result, obtain an updated target detection result and record the updated target features.
[0093] In some embodiments of the present application, the driving device of the autonomous driving vehicle also includes: a determination unit, which is used to determine whether the vehicle currently meets the preset data upload conditions after obtaining the target detection result of the vehicle and determining the target information output by the vehicle based on the target detection result of the vehicle; an uploading unit, which is used to determine the data to be uploaded based on the target information output by the vehicle when the preset data upload conditions are met, and upload the data to be uploaded to the cloud or the road side, so that the cloud or the road side updates the memory map information according to the data to be uploaded.
[0094] In some embodiments of the present application, the determination unit is specifically used to: determine whether the vehicle detects a difficult-to-detect target based on the target detection result of the vehicle; if so, determine that the vehicle currently meets the preset data upload condition; otherwise, determine that the vehicle currently does not meet the preset data upload condition; when the preset data upload condition is met, determine the data to be uploaded based on the target information output by the vehicle, and uploading the data to be uploaded to the cloud or road side, including: when the preset data upload condition is met, track and compare the target detection result of the vehicle to obtain target difference information; and upload the target difference information to the cloud or road side.
[0095] In some embodiments of the present application, the second acquisition unit 220 is specifically used to: obtain the current positioning information of the vehicle and upload it to the cloud or the road side; receive the memory map information corresponding to the current positioning information sent by the cloud or the road side.
[0096] In some embodiments of the present application, the target information output by the self-vehicle includes a target position detected by the self-vehicle, and the fusion unit 230 is specifically used to: determine whether there is corresponding target information output by the self-vehicle at the target position in the target reference information according to the target position in the target reference information and the target position detected by the self-vehicle; when there is corresponding target information output by the self-vehicle at the target position in the target reference information, fuse the target reference information with the corresponding target information output by the self-vehicle to obtain fused target information; the driving device of the autonomous driving vehicle also includes: a marking unit, which is used to mark the target position as a potential target position when there is no corresponding target information output by the self-vehicle at the target position in the target reference information, and determine the target information of the potential target according to the target reference information.
[0097] In some embodiments of the present application, the fused target information includes the fused target information of the difficult-to-detect target, and the control unit 240 is specifically used to: determine that the self-vehicle driving strategy is a driving strategy set for the difficult-to-detect target according to the fused target information of the difficult-to-detect target, and control the driving of the self-vehicle according to the driving strategy set for the difficult-to-detect target; the control unit is also used to: determine that the self-vehicle driving strategy is a driving strategy set for the potential target according to the target information of the potential target, and control the driving of the self-vehicle according to the driving strategy set for the potential target.
[0098] It can be understood that the driving device of the above-mentioned autonomous driving vehicle can implement the various steps of the driving method of the autonomous driving vehicle provided in the aforementioned embodiments. The relevant explanations about the driving method of the autonomous driving vehicle are applicable to the driving device of the autonomous driving vehicle and will not be repeated here.
[0099] An embodiment of the present application also provides a driving system for an autonomous driving vehicle, wherein the driving system for the autonomous driving vehicle includes a vehicle side and a cloud side and / or a road side, and the vehicle side is used to execute any of the aforementioned driving methods for the autonomous driving vehicle.
[0100] The cloud or road side of the embodiment of the present application is also used to execute:
[0101] Receive target information sent by the vehicle, including target location, shape, features, confidence, and grid map information, and continuously update the map information in the global coordinate system based on the information uploaded by each vehicle. In addition, some difficult-to-detect targets can be re-identified offline. For example, large models such as GPT models can be called to re-identify difficult-to-detect targets, and the memory map information maintained in the cloud or on the road side can be further updated based on the recognition results of the large model.
[0102] Figure 3This is a schematic diagram of the structure of an electronic device according to an embodiment of the present application. Figure 3 At the hardware level, the electronic device includes a processor, and optionally also includes an internal bus, a network interface, and a memory. The memory may include a memory, such as a high-speed random access memory (RAM), and may also include a non-volatile memory (non-volatile memory), such as at least one disk storage. Of course, the electronic device may also include hardware required for other services.
[0103] The processor, network interface and memory can be interconnected through an internal bus, which can be an ISA (Industry Standard Architecture) bus, a PCI (Peripheral Component Interconnect) bus or an EISA (Extended Industry Standard Architecture) bus. The bus can be divided into an address bus, a data bus, a control bus, etc. For ease of representation, Figure 3 Only one bidirectional arrow is used in the diagram, but this does not mean that there is only one bus or only one type of bus.
[0104] The memory is used to store the program. Specifically, the program may include a program code, and the program code includes a computer operation instruction. The memory may include a memory and a non-volatile memory, and provides instructions and data to the processor.
[0105] The processor reads the corresponding computer program from the non-volatile memory into the memory and then runs it, forming the driving device of the autonomous driving vehicle at the logical level. The processor executes the program stored in the memory and is specifically used to perform the following operations:
[0106] Obtaining a target detection result of the vehicle and determining target information output by the vehicle according to the target detection result of the vehicle;
[0107] Acquire memory map information, wherein the memory map information includes target reference information provided by other vehicles or the vehicle in history;
[0108] Fusing the memory map information with the target information output by the vehicle to obtain fused target information;
[0109] The driving strategy of the vehicle is determined according to the fused target information, and the driving of the vehicle is controlled according to the driving strategy of the vehicle.
[0110] The above application Figure 1The method performed by the driving device of the autonomous driving vehicle disclosed in the illustrated embodiment can be applied to a processor or implemented by a processor. The processor may be an integrated circuit chip with signal processing capabilities. In the implementation process, each step of the above method can be completed by an integrated logic circuit of hardware in the processor or instructions in software form. The above processor can be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), etc.; it can also be a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA) or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components. The methods, steps and logic block diagrams disclosed in the embodiments of the present application can be implemented or executed. The general-purpose processor can be a microprocessor or the processor can also be any conventional processor, etc. The steps of the method disclosed in the embodiments of the present application can be directly embodied as a hardware decoding processor for execution, or a combination of hardware and software modules in the decoding processor for execution. The software module can be located in a storage medium mature in the art such as a random access memory, a flash memory, a read-only memory, a programmable read-only memory, or an electrically erasable programmable memory, a register, etc. The storage medium is located in the memory, and the processor reads the information in the memory and completes the steps of the above method in combination with its hardware.
[0111] The present application also provides a computer-readable storage medium, which stores one or more programs, wherein the one or more programs include instructions, which, when executed by an electronic device including multiple application programs, enable the electronic device to execute Figure 1 The method performed by the driving device of the autonomous driving vehicle in the illustrated embodiment is specifically used to perform:
[0112] Obtaining a target detection result of the vehicle and determining target information output by the vehicle according to the target detection result of the vehicle;
[0113] Acquire memory map information, wherein the memory map information includes target reference information provided by other vehicles or the vehicle in history;
[0114] Fusing the memory map information with the target information output by the vehicle to obtain fused target information;
[0115] The driving strategy of the vehicle is determined according to the fused target information, and the driving of the vehicle is controlled according to the driving strategy of the vehicle.
[0116] Those skilled in the art will appreciate that embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.
[0117] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.
[0118] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.
[0119] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.
[0120] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.
[0121] The memory may include non-permanent storage in a computer-readable medium, random access memory (RAM) and / or non-volatile memory in the form of read-only memory (ROM) or flash RAM. The memory is an example of a computer-readable medium.
[0122] Computer readable media include permanent and non-permanent, removable and non-removable media that can be implemented by any method or technology to store information. Information can be computer readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, compact disk read-only memory (CD-ROM), digital versatile disk (DVD) or other optical storage, magnetic cassettes, magnetic tape magnetic disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device. As defined herein, computer readable media does not include temporary computer readable media (transitory media), such as modulated data signals and carrier waves.
[0123] It should also be noted that the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, commodity or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, commodity or device. In the absence of more restrictions, the elements defined by the sentence "comprises a ..." do not exclude the existence of other identical elements in the process, method, commodity or device including the elements.
[0124] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment or an embodiment in combination with software and hardware. Moreover, the present application may adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that contain computer-usable program code.
[0125] The above is only an embodiment of the present application and is not intended to limit the present application. For those skilled in the art, the present application may have various changes and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application should be included in the scope of the claims of the present application.
Claims
1. A driving method for an autonomous driving vehicle, in, The driving method of the autonomous driving vehicle is executed by the vehicle side, and the driving method of the autonomous driving vehicle includes: Obtaining a target detection result of the vehicle and determining target information output by the vehicle according to the target detection result of the vehicle; Acquire memory map information, wherein the memory map information includes target reference information provided by other vehicles or the vehicle in history; Fusing the memory map information with the target information output by the vehicle to obtain fused target information; The driving strategy of the vehicle is determined according to the fused target information, and the driving of the vehicle is controlled according to the driving strategy of the vehicle.
2. The method for driving an autonomous vehicle according to claim 1, in, The step of obtaining the target detection result of the vehicle and determining the target information output by the vehicle according to the target detection result of the vehicle comprises: Create and maintain raster maps in a global coordinate system; Acquire the sensor perception data of the vehicle and perform target detection to obtain the target detection result of the vehicle; updating the grid map according to the target detection result to obtain updated grid map information; The target detection result is updated according to the relative distance and relative viewing angle between the vehicle and the target detected in the target detection result, an updated target detection result is obtained and updated target features are recorded.
3. The method for driving an autonomous vehicle according to claim 1, in, After obtaining the target detection result of the self-vehicle and determining the target information output by the self-vehicle according to the target detection result of the self-vehicle, the driving method of the automatic driving vehicle further includes: Determine whether the vehicle currently meets the preset data uploading conditions; When the preset data uploading conditions are met, the data to be uploaded is determined according to the target information output by the vehicle, and the data to be uploaded is uploaded to the cloud or the road side, so that the cloud or the road side updates the memory map information according to the data to be uploaded.
4. The method for driving an autonomous vehicle according to claim 3, in, Determining whether the vehicle currently meets the preset data uploading conditions includes: Determine whether the ego vehicle has detected a difficult-to-detect target based on the target detection result of the ego vehicle; If so, it is determined that the vehicle currently meets the preset data uploading condition; Otherwise, it is determined that the vehicle currently does not meet the preset data uploading condition; When the preset data uploading condition is met, determining the data to be uploaded according to the target information output by the vehicle, and uploading the data to be uploaded to the cloud or the road side includes: When the preset data upload conditions are met, the target detection results of the vehicle are tracked and compared to obtain target difference information; The target difference information is uploaded to the cloud or the road side.
5. The driving method of the autonomous driving vehicle according to claim 1, in, The obtaining of memory map information comprises: Obtain the current location information of the vehicle and upload it to the cloud or road side; Receive the memory map information corresponding to the current positioning information sent by the cloud or road end.
6. The method for driving an autonomous vehicle according to claim 1, in, The target information output by the vehicle includes the target position detected by the vehicle, and the fusion of the memory map information with the target information output by the vehicle to obtain the fused target information includes: Determining whether there is corresponding target information output by the ego vehicle at the target position in the target reference information according to the target position in the target reference information and the target position detected by the ego vehicle; When there is corresponding target information output by the ego vehicle at the target position in the target reference information, the target reference information is fused with the corresponding target information output by the ego vehicle to obtain fused target information; The driving method of the autonomous driving vehicle further includes: When there is no corresponding target information output by the vehicle at the target position in the target reference information, the target position is marked as the position of a potential target, and the target information of the potential target is determined according to the target reference information.
7. The driving method of the autonomous driving vehicle according to claim 6, in, The fused target information includes fused target information of the difficult-to-detect target, and determining the driving strategy of the vehicle according to the fused target information, and controlling the driving of the vehicle according to the driving strategy of the vehicle includes: Determining, according to the fused target information of the difficult-to-detect target, that the self-vehicle driving strategy is a driving strategy set for the difficult-to-detect target, and controlling the driving of the self-vehicle according to the driving strategy set for the difficult-to-detect target; The driving method of the autonomous driving vehicle further includes: The vehicle driving strategy is determined to be a driving strategy set for the potential target according to the target information of the potential target, and the vehicle driving is controlled according to the driving strategy set for the potential target.
8. A driving device for an autonomous vehicle, in, The driving device of the autonomous driving vehicle is applied to the vehicle side, and the driving device of the autonomous driving vehicle includes: A first acquisition unit, used to acquire a target detection result of the vehicle and determine target information output by the vehicle according to the target detection result of the vehicle; A second acquisition unit is used to acquire memory map information, wherein the memory map information includes target reference information provided by other vehicles or the vehicle in history; a fusion unit, used for fusing the memory map information with the target information output by the vehicle to obtain fused target information; The control unit is used to determine the driving strategy of the vehicle according to the fused target information, and control the driving of the vehicle according to the driving strategy of the vehicle.
9. A driving system for an autonomous vehicle, in, The driving system of the autonomous driving vehicle includes a vehicle side and a cloud side and / or a road side, and the vehicle side is used to execute the driving method of the autonomous driving vehicle according to any one of claims 1 to 7.
10. An electronic device, include: processor; as well as A memory arranged to store computer executable instructions, which, when executed, cause the processor to execute the driving method of the autonomous driving vehicle as described in any one of claims 1 to 7.
11. A computer-readable storage medium storing one or more programs, which, when executed by an electronic device including a plurality of application programs, enables the electronic device to execute the driving method of the autonomous driving vehicle of any one of claims 1 to 7.
