Reference line track detection method and device for automatic driving

Abstract

The invention discloses a reference line track detection method and device for automatic driving. The method comprises the following steps: acquiring a first reference line track and a second reference line track of a vehicle; wherein the first reference line track is obtained based on a preset high-precision map and positioning parameters of the vehicle, and the second reference line track is obtained based on a lane line map; performing coordinate unification based on a coordinate system conversion relation between the high-precision map and the lane line map, and generating a track distance after coordinate unification; based on the first reference line track, the second reference line track and the vehicle driving parameters, a first control instruction corresponding to the first reference line track and a second control instruction corresponding to the second reference line track are generated, and an instruction difference value of the first control instruction and the second control instruction is generated; a reference line trajectory detection result is determined based on at least one of the trajectory distance and the instruction difference value. According to the invention, the reference line track detection result can be quickly and accurately obtained.

Description

Technical Field

[0001] This application relates to the field of autonomous driving technology, and in particular to a method and apparatus for detecting reference line trajectories in autonomous driving. Background Technology

[0002] In autonomous driving, decision planning is used to generate the driving trajectory of the autonomous vehicle. The decision planning module mainly includes: global navigation path, reference lines, and local trajectory. The reference lines are the transitional stage from macro-path planning to local trajectory generation. Based on the reference lines, decision planning considers factors such as traffic rules, obstacle projection, path optimization, and speed decisions to generate the final driving trajectory.

[0003] For autonomous vehicles, generating accurate reference lines helps achieve goals such as obstacle avoidance, path smoothing, and speed optimization. However, the accuracy of the reference line trajectory can be affected when the vehicle is on special road sections with large positioning errors, or when sensors such as vision cameras are interfered with in adverse weather conditions.

[0004] In related technologies, to detect reference line trajectories during autonomous driving, fault detection of upstream modules is typically relied upon. This can be achieved through fault diagnosis based on camera images, sensor sensing range, or satellite positioning parameters generated by the positioning module. When camera images are blurry, sensor sensing range is inaccurate, or vehicle positioning signals are missing or have significant errors, it is determined that the corresponding upstream module is faulty, thus indicating an anomaly in the reference line trajectory.

[0005] In the aforementioned methods for detecting anomalies in the reference trajectory, fault diagnosis of sensors and positioning parameters requires the construction of complex judgment models based on a large amount of sample data and expert experience. The detection results of various state parameters are then input into the model to achieve fault diagnosis. However, the detection information of these state parameters is only used for anomaly detection of the reference trajectory, resulting in significant information redundancy and consuming substantial detection and computational resources. Furthermore, the aforementioned fault diagnosis method also suffers from a certain time delay; by the time an anomaly or error is detected, the reference trajectory has often already deviated, making it impossible to quickly and in real-time determine the detection result. In addition, the method of indirectly generating the reference trajectory detection result by detecting whether each upstream module is abnormal may fail to detect anomalies in the reference trajectory if the reference trajectory has already deviated but the state parameters of the upstream modules have not reached the anomaly threshold due to insufficiently precise preset thresholds. Therefore, how to detect the reference trajectory in real-time and effectively during autonomous driving has become an urgent technical problem to be solved. Summary of the Invention

[0006] This application provides a method and apparatus for detecting reference line trajectories in autonomous driving, which can improve the technical problem in related technologies that the reference line trajectory cannot be detected in real time and effectively during autonomous driving.

[0007] In a first aspect, embodiments of this application provide a reference line trajectory detection method for autonomous driving, the method comprising: The vehicle's first reference line trajectory and second reference line trajectory are obtained; wherein, the first reference line trajectory is obtained based on a preset high-precision map and the vehicle's positioning parameters, and the second reference line trajectory is obtained based on a lane line map; the lane line map is generated based on road image information around the vehicle. Based on the coordinate system transformation relationship between the high-precision map and the lane line map, the coordinates of the first reference line trajectory and the second reference line trajectory are unified, and the trajectory distance of the first reference line trajectory and the second reference line trajectory after the coordinate unification is generated. Based on the first reference line trajectory, the second reference line trajectory, and vehicle driving parameters, a first control command corresponding to the first reference line trajectory and a second control command corresponding to the second reference line trajectory are generated, and the command difference value between the first control command and the second control command is generated. The reference line trajectory detection result is determined based on at least one of the trajectory distance and instruction difference value.

[0008] Secondly, embodiments of this application provide a reference line trajectory detection device for autonomous driving, comprising: The acquisition module is used to acquire a first reference line trajectory and a second reference line trajectory; wherein, the first reference line trajectory is obtained based on a high-precision map and vehicle satellite positioning parameters, and the second reference line trajectory is obtained based on a lane line map; the lane line map is generated based on captured road image information. The first generation module is used to unify the coordinates of the first reference line trajectory and the second reference line trajectory based on the coordinate system transformation relationship between the high-precision map and the lane line map, and generate the trajectory distance of the first reference line trajectory and the second reference line trajectory after the coordinate unification. The second generation module is used to generate a first control command corresponding to the first reference line trajectory and a second control command corresponding to the second reference line trajectory based on the first reference line trajectory, the second reference line trajectory and vehicle driving parameters, and to generate the command difference value between the first control command and the second control command. The judgment module is used to determine the reference line trajectory detection result based on at least one of the trajectory distance and instruction difference value.

[0009] Thirdly, embodiments of this application provide an electronic device, including: a processor and a memory storing computer program instructions; The steps of the reference line trajectory detection method for autonomous driving in the first aspect are implemented when the processor executes computer program instructions.

[0010] Fourthly, embodiments of this application provide a computer-readable storage medium storing computer program instructions, which, when executed by a processor, implement the steps of the reference line trajectory detection method for autonomous driving described in the first aspect.

[0011] Fifthly, embodiments of this application provide a computer program product, which includes a computer program. When the instructions in the computer program product are executed by a processor, they implement the steps of the reference line trajectory detection method for autonomous driving described in the first aspect.

[0012] The autonomous driving reference line trajectory detection method and apparatus of this application, after determining a first reference line trajectory under a high-precision map and a second reference line trajectory under a lane line map in two ways, can unify the coordinates of the two reference line trajectories based on the coordinate system transformation relationship between the two maps, and calculate the trajectory distance between the two reference line trajectories after coordinate unification. Based on the two reference line trajectories, corresponding control commands during autonomous driving can be generated, and the command difference value between the two control commands can be determined. Based on at least one of the trajectory distance and the command difference value, the detection result of the reference line trajectory can be determined. When a fault is diagnosed in either method, a reference line trajectory fault can be determined, improving the sensitivity, accuracy, and reliability of reference line trajectory detection. Attached Figure Description

[0013] To more clearly illustrate the technical solutions of the embodiments of this application, the accompanying drawings used in the embodiments of this application will be briefly introduced below. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0014] Figure 1 This is a schematic flowchart of a reference line trajectory detection method for autonomous driving provided in an embodiment of this application; Figure 2 This is a schematic diagram illustrating the generation of two reference line trajectories provided in an embodiment of this application; Figure 3 This is a partial flowchart of a reference line trajectory detection method for autonomous driving provided in another embodiment of this application; Figure 4 This is a partial flowchart of a reference line trajectory detection method for autonomous driving provided in another embodiment of this application; Figure 5 This is a comparative schematic diagram of a reference line trajectory and control command provided in an embodiment of this application; Figure 6This is a schematic diagram of a fault flag bit based on a reference line sampling point provided in an embodiment of this application; Figure 7 This is a schematic diagram of a fault flag bit based on control commands provided in an embodiment of this application; Figure 8 This is a schematic diagram of fault diagnosis for a vehicle control module provided in an embodiment of this application; Figure 9 This is a schematic diagram of the structure of an autonomous driving reference line trajectory detection device provided in an embodiment of this application; Figure 10 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. Detailed Implementation

[0015] The features and exemplary embodiments of various aspects of this application will be described in detail below. To make the objectives, technical solutions, and advantages of this application clearer, the application will be further described in detail below with reference to the accompanying drawings and specific embodiments. It should be understood that the specific embodiments described herein are only intended to explain this application and not to limit it. For those skilled in the art, this application can be implemented without some of these specific details. The following description of the embodiments is merely to provide a better understanding of this application by illustrating examples.

[0016] It should be noted that, in this document, relational terms such as "first" and "second" are used merely to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising..." does not exclude the presence of additional identical elements in the process, method, article, or apparatus that includes the element.

[0017] To address at least one of the aforementioned technical problems, embodiments of this application provide a method and apparatus for detecting reference line trajectories in autonomous driving. The method for detecting reference line trajectories in autonomous driving provided by embodiments of this application will be described below.

[0018] Figure 1 A flowchart illustrating a reference line trajectory detection method for autonomous driving according to an embodiment of this application is shown. The method may include the following steps: S110, acquire the first reference line trajectory and the second reference line trajectory of the vehicle; wherein, the first reference line trajectory is obtained based on a preset high-precision map and the vehicle's positioning parameters, and the second reference line trajectory is obtained based on a lane line map; the lane line map is generated based on road image information around the vehicle; S120, based on the coordinate system transformation relationship between the high-precision map and the lane line map, the coordinates of the first reference line trajectory and the second reference line trajectory are unified, and the trajectory distance between the first reference line trajectory and the second reference line trajectory after the coordinate unification is calculated; S130, based on the first reference line trajectory, the second reference line trajectory and the vehicle driving parameters, generate the first control command corresponding to the first reference line trajectory and the second control command corresponding to the second reference line trajectory, and generate the command difference value between the first control command and the second control command; S140, determine the reference line trajectory detection result based on at least one of the trajectory distance and the instruction difference value.

[0019] In related technologies, the generation of reference lines for autonomous vehicles mainly relies on perception sensors such as radar and cameras, or on positioning modules such as GPS combined with high-precision maps. Current diagnostic methods are primarily based on fault detection of perception sensors and positioning systems. For example, diagnosis based on camera imaging results indicates a perception sensor malfunction if the camera image is blurry or the perception range is inaccurate, further suggesting a significant error in the generated reference line trajectory. Similarly, faulty positioning information also affects reference line generation; therefore, fault diagnosis based on the positioning module is also used to assess the accuracy of the reference line trajectory. For instance, detecting a lack of GPS positioning signal or significant errors in specific road sections suggests a fault in reference line trajectory generation.

[0020] The aforementioned fault detection methods for reference line trajectories in autonomous vehicles have certain limitations. First, fault diagnosis methods for perception sensors and positioning systems require complex expert experience and a large amount of measurable data. They necessitate the establishment of complex models and the design of state observation methods to diagnose faults, and this information is redundant for reference line faults. Second, these diagnostic methods have a certain time delay. By the time a significant error in the perception sensor or positioning information is detected, the generated reference line has often already deviated, making it impossible to make a timely judgment based on the actually generated reference line. Finally, since the detection of perception sensors and positioning information is an indirect method for diagnosing reference line faults, the fault thresholds set by upstream modules may not meet the accuracy requirements of the planning layer, easily leading to situations where reference line deviations go undetected.

[0021] In this embodiment, after determining the first reference line trajectory under the high-precision map and the second reference line trajectory under the lane line map using two methods respectively, the coordinates of the two reference line trajectories can be unified based on the coordinate system transformation relationship between the two maps, and the trajectory distance between the two reference line trajectories after coordinate unification can be calculated. Based on the two reference line trajectories, corresponding control commands for the autonomous driving process can be generated respectively, and the command difference value between the two control commands can be determined. Based on at least one of the trajectory distance and command difference value, the detection result of the reference line trajectory can be determined. When a fault is diagnosed in either method, a reference line trajectory fault can be determined, improving the sensitivity, accuracy, and reliability of reference line trajectory detection.

[0022] The specific implementation methods for each of the above steps are described below.

[0023] In S110, the vehicle, in autonomous driving mode, can acquire a first reference line trajectory and a second reference line trajectory. The first reference line trajectory can be obtained based on a preset high-precision map and vehicle positioning parameters, while the second reference line trajectory can be obtained based on a lane line map, which can be generated based on road image information surrounding the vehicle.

[0024] Understandably, taking a vehicle equipped with a monocular camera as an example, a lane line map can be generated by combining road images captured by the monocular camera with lane line detection in the road images using a neural network model. However, lane line maps can also be acquired and generated using other methods, which are not limited here.

[0025] In some embodiments, obtaining the first reference line trajectory includes: S210 determines the corresponding global navigation path based on the starting and ending points of autonomous driving; the global navigation path includes road segment information of multiple road segments. S220 determines the target road segment from multiple road segments based on the vehicle's satellite positioning parameters; S230, based on the road segment information of the target road segment, obtains the corresponding lane centerline from the high-precision map to form the first reference line trajectory.

[0026] In this embodiment, during autonomous driving, a global navigation path can be determined based on the starting and ending points, and a target road segment can be determined from the global navigation path based on the vehicle's positioning information. Within the target road segment, the lane centerline can be obtained from a high-precision map to form a first reference line trajectory.

[0027] In S210, under autonomous driving mode, the first reference line trajectory is obtained by using a global path planning algorithm to determine the global navigation path from the starting point to the ending point in a high-precision map, based on the pre-determined starting and ending points in autonomous driving mode. The aforementioned high-precision map (topological map) refers to an abstract map in cartography that maintains the correct relative positional relationships between points and lines but does not necessarily maintain the correct shape, area, distance, or direction of the graphic.

[0028] The global navigation path includes information on multiple road segments, each of which may include information such as the starting point, ending point, and length of the road segment.

[0029] In one example, the global navigation path can be represented as: ; in, This indicates a global navigation path that includes multiple road segments. This indicates the road segment information in the navigation path, including the start point, end point, length, and other information for each road segment; It is a road segment that includes the starting point of the vehicle. It is a road segment that includes the location of the vehicle's termination point.

[0030] In S220, based on the vehicle's current satellite positioning parameters, the target road segment corresponding to the vehicle's current location can be determined from multiple road segments of the global navigation path.

[0031] In S230, after determining the target road segment, the corresponding lane centerline can be obtained from the high-precision map based on the road segment information of the target road segment to form the first reference line trajectory.

[0032] The aforementioned high-precision map (HD map) is a type of high-precision map used for autonomous driving, which can include map elements such as roads, lanes, roadside traffic signs, and ground markings.

[0033] As an optional implementation, based on the road segment information of the target road segment, the lane centerline can be extracted from a high-precision map as an initial reference line. This lane centerline can be represented as a set of multiple waypoints. ; in This represents the lane centerline in a high-precision map, containing the horizontal and vertical coordinates of multiple path points. ; After obtaining the initial reference line, it can be smoothed based on the x and y coordinates of multiple path points, taking into account the geometric information of the initial reference line and the uniformity among the path points. The final smoothed reference line can then be obtained. ; in It is a smoothed reference line, containing the x and y coordinates of the smoothed path points. .

[0034] The above smoothed reference line This is the trajectory of the first reference line.

[0035] In some embodiments, obtaining the second reference line trajectory includes: S310 determines multiple lane sampling points based on captured road image information, forming a visually detected lane line composed of multiple lane sampling points; S320 generates lane line maps based on visually detected lane lines when no lane line map is generated for the vehicle. S330, when the vehicle has already generated a lane line map, associates the original lane lines in the lane line map with the visually detected lane lines, and optimizes the lane line map based on the association results; S340, based on the lane line map, obtain the second reference line trajectory.

[0036] In this embodiment, during autonomous driving, multiple lane sampling points are determined based on captured road image information, and a visually detected lane line composed of these lane sampling points is generated. If the vehicle has not generated a lane line map, it can generate a lane line map based on the visually detected lane lines. If the vehicle has already generated a lane line map, the original lane lines in the lane line map can be associated with the visually detected lane lines to optimize the lane line map and determine the second reference line trajectory.

[0037] In S310, during autonomous driving mode, the second reference line trajectory is obtained by using sensors such as cameras and radar mounted on the vehicle to capture or sense the surrounding environment, thus acquiring corresponding visual information. For example, when the vehicle is equipped with a monocular or tricular camera, road image information can be obtained by capturing images of the road.

[0038] After acquiring road image information as visual information, a lane detection model based on a neural network can be used to process the visual information to identify visually detected lane lines formed by multiple lane sampling points. This visually detected lane line can be represented as a set of multiple lane sampling points: ; in, This indicates visually detected lane lines, including sampling points for multiple lane lines. This represents the sampling points of the detected lane lines, including the horizontal and vertical coordinate values ​​of the sampling points.

[0039] In S320, if the vehicle has not yet generated a lane map, for example, when the vehicle has just completed the power-on process or the user has reset the original lane map, the vehicle can generate a lane map based on the lane lines detected by vision.

[0040] In S330, if a lane map has already been generated during vehicle operation, the vehicle can retrieve the existing lane lines from the lane map. The existing lane lines can be composed of multiple lane line control points. The existing lane lines can be represented as a set of multiple lane line control points: ; This indicates the existing lane markings. This indicates the control points of the lane line spline.

[0041] After determining the original lane lines and visual detection lane lines Then, the two can be associated, and the association result can be used to determine whether they are the same reference line.

[0042] After determining the original lane lines and visual detection lane lines After using the same reference line, lane lines can be detected visually. Multiple sampling points on the map are used to optimize the control points of each lane spline of the original lane line L in the lane line map to obtain the updated spline curve.

[0043] In S340, after optimizing the control points in the lane line map, an updated lane line spline curve can be determined based on the optimized control points in the lane line map. This lane line spline curve can then be used as the second reference line trajectory. The second reference line trajectory can be represented as: ; in, This represents the second reference line trajectory calculated and optimized based on the detection results in the lane map, including the horizontal and vertical coordinates of each optimized control point. .

[0044] Please refer to Figure 2 , Figure 2 The reference line 2 shown is the first reference line trajectory obtained through the above implementation method, and the reference line 1 is the second reference line trajectory obtained through the above implementation method.

[0045] In S120, after determining the first reference line trajectory and the second reference line trajectory, the coordinates of the first reference line trajectory and the second reference line trajectory can be unified based on the coordinate system transformation relationship between the high-precision map and the lane line map. Based on the first reference line trajectory and the second reference line trajectory after coordinate unification, the trajectory distance between the first reference line trajectory and the second reference line trajectory can be calculated.

[0046] It should be noted that the above trajectory distance can be the distance between a certain feature point in the first reference line trajectory and a certain feature point in the second reference line trajectory.

[0047] Please refer to Figure 3 In some embodiments, the coordinate unification of the first reference line trajectory and the second reference line trajectory based on the coordinate system transformation relationship between the high-precision map and the lane line map includes: S410, acquire the coordinate parameters of the vehicle and the feature points around the vehicle in the lane line map and the high-precision map respectively; S420, based on coordinate parameters, determines the coordinate system transformation relationship between the high-precision map and the lane line map; S430, based on coordinate system transformation relationship, performs coordinate transformation on multiple first sampling points of the first reference line trajectory or multiple second sampling points of the second reference line trajectory.

[0048] In this embodiment, when performing coordinate system transformation between the two maps, the coordinate parameters of the vehicle in each map and the coordinate parameters of the feature points around the vehicle in each map can be obtained in advance. Based on the coordinate parameters of multiple positioning points in different maps, the coordinate system transformation relationship between the two maps can be obtained. Using this coordinate system transformation relationship, the coordinates of the first sampling point and the second sampling point can be unified.

[0049] In S410, when performing coordinate system transformation between the high-precision map and the lane line map, the coordinate parameters of the vehicle and the feature points around the vehicle in the lane line map and the high-precision map respectively can be obtained, so as to achieve coordinate transformation between the two coordinate systems based on the coordinate parameters of the same positioning point in different maps.

[0050] In some embodiments, the above-described S410 may include: Obtain the vehicle's location point information in the lane line map and the vehicle's satellite positioning parameters in the high-precision map at the same time; Based on the collected road image information, determine the coordinate parameters of the corresponding feature points in the lane line map, and obtain the coordinate parameters of the feature points matched in the high-precision map.

[0051] In this embodiment, when performing coordinate system conversion between the high-precision map and the lane line map, it is necessary to collect the vehicle's location point information in the lane line map and the vehicle's satellite positioning parameters in the high-precision map at the same time, so as to serve as the vehicle's coordinate parameters in the lane line map and the high-precision map respectively.

[0052] After a vehicle acquires road image information around it using a monocular camera or other acquisition modules, some feature points can be selected from the road image information, and the coordinate parameters of these feature points in the lane line map and the coordinate parameters of these feature points matched in the high-precision map can be determined.

[0053] In S420, the coordinate system transformation relationship between the two maps can be determined based on the coordinate parameters of the vehicle in the two maps and the coordinate parameters of the feature points around the vehicle in the two maps.

[0054] Specifically, based on the vehicle's satellite positioning parameters on the lane map at the same time, the vehicle's satellite positioning parameters on the high-precision map using BeiDou GPS or Galileo, and the location information of feature points in the visual images obtained by the vehicle's camera on the high-precision map, the transformation matrix between the lane map coordinate system and the high-precision map coordinate system is calculated. This transformation matrix can be expressed as: ; in Characterizing the rotation matrix, Characterizes the translation vector; After obtaining the transformation matrix, the coordinates of the first and second sampling points can be unified using the transformation matrix.

[0055] In step S430, after obtaining the coordinate system transformation relationship, i.e., the aforementioned transformation matrix, coordinate transformation can be performed on multiple first sampling points of the first reference line trajectory or multiple second sampling points of the second reference line trajectory. For example, if the lane line map is used as the unified coordinate system, the aforementioned transformation matrix can be used to transform the coordinates of multiple first sampling points of the first reference line trajectory, unifying them to the position coordinates of the lane line map. Conversely, if a high-precision map is used as the unified coordinate system, the aforementioned transformation matrix can be used to transform the coordinates of multiple second sampling points of the second reference line trajectory, unifying them to the position coordinates of the high-precision map.

[0056] Please refer to Figure 4 In some embodiments, the trajectory distance between the first reference line trajectory and the second reference line trajectory after coordinate unification includes: S510, determine the first sampling point and the second sampling point corresponding to the same moment from the first reference line trajectory and the second reference line trajectory; the first sampling point and the second sampling point contain coordinate parameters in the same coordinate system; S520 generates the distance between the first sampling point and the second sampling point.

[0057] In this embodiment, a first sampling point and a second sampling point can be determined from the first reference line trajectory and the second reference line trajectory, respectively. The distance between the two sampling points can be calculated as the trajectory distance between the two reference line trajectories.

[0058] In S510, after determining the first reference line trajectory and the second reference line trajectory, a first sampling point can be determined from the first reference line trajectory and a second sampling point can be determined from the second reference line trajectory, using the same time point as a reference. Based on the above transformation matrix, the final first sampling point and the second sampling point can contain coordinate parameters in the same coordinate system, that is, the first sampling point and the second sampling point have completed the unification of the coordinate system.

[0059] In S520, after determining the first sampling point and the second sampling point, the distance between the first sampling point and the second sampling point can be calculated as the trajectory distance between the first reference line trajectory and the second reference line trajectory.

[0060] As an optional implementation, based on the first and second sampling points after coordinate unification, the Euclidean distance between the first and second sampling points can be calculated through coordinate operations, and used as the trajectory distance between the first and second reference line trajectories.

[0061] The Euclidean distance between the first and second sampling points can be expressed as: ; express At time t, the Euclidean distance between the first and second sampling points is , They represent The x-coordinates of the first and second sampling points at time 1. , They represent The ordinates of the first and second sampling points at time t.

[0062] In S130, based on the first reference trajectory, the second reference trajectory, and the vehicle driving parameters in autonomous driving mode, a first control command corresponding to the first reference trajectory and a second control command corresponding to the second reference trajectory can be generated respectively. That is, the first control command is used to control the vehicle's autonomous driving under the first reference trajectory, and the second control command is used to control the vehicle's autonomous driving under the second reference trajectory.

[0063] After obtaining the first control instruction and the second control instruction, the first control instruction and the second control instruction can be numerically quantized, and the instruction difference value between the first control instruction and the second control instruction can be calculated.

[0064] In some embodiments, the above-mentioned generation of a first control command corresponding to the first reference line trajectory and a second control command corresponding to the second reference line trajectory based on the first reference line trajectory, the second reference line trajectory, and vehicle driving parameters includes: S610, based on the vehicle's position information and speed information, respectively determine the first longitudinal control quantity corresponding to the first reference line trajectory and the second longitudinal control quantity corresponding to the second reference line trajectory; S620, based on the vehicle's position information and heading angle information, determines the first lateral control quantity corresponding to the first reference line trajectory and the second lateral control quantity corresponding to the second reference line trajectory, respectively.

[0065] In this embodiment, for the first reference line trajectory, a first lateral control quantity and a first longitudinal control quantity corresponding to the first reference line trajectory can be generated; for the second reference line trajectory, a second lateral control quantity and a second longitudinal control quantity corresponding to the second reference line trajectory can be generated.

[0066] In S610, under autonomous driving mode, the vehicle's longitudinal controller can generate a longitudinal control quantity corresponding to a reference trajectory based on the vehicle's position and speed information. Therefore, after obtaining the first and second reference trajectories, a first longitudinal control quantity corresponding to the first reference trajectory and a second longitudinal control quantity corresponding to the second reference trajectory can be generated, respectively.

[0067] As an optional implementation, the above-mentioned longitudinal control quantity can be expressed as: ; in, Indicates the trajectory based on the first reference line. Indicates the trajectory based on the second reference line. express The coordinates of the reference point on the first or second reference line trajectory at any given time. This indicates the vehicle's current position coordinates. This indicates the vehicle's current speed.

[0068] Based on the above parameters, the longitudinal controller can calculate the vehicle's position deviation, speed deviation, acceleration compensation, etc. at the current moment, and generate the corresponding speed control quantity: ; in This indicates that the trajectory is generated based on the first reference line. This indicates that the trajectory is generated based on the second reference line. This represents the longitudinal control quantity output by the longitudinal controller, including... Speed ​​control quantity at any moment .

[0069] In S620, under autonomous driving mode, the vehicle's lateral controller can generate lateral control quantities corresponding to the reference trajectory based on the vehicle's position and heading angle information. Therefore, after obtaining the first and second reference trajectories, a first lateral control quantity corresponding to the first reference trajectory and a second lateral control quantity corresponding to the second reference trajectory can be generated, respectively.

[0070] As an optional implementation, the above-mentioned lateral control quantity can be expressed as: ; in, This indicates the generation of the first reference line trajectory. This indicates the generation of the second reference line trajectory. Inputs representing the inputs to the lateral controller, including The coordinates of the reference point on the first or second reference line trajectory at time t. The current orientation of the vehicle .

[0071] Based on the above parameters, the lateral controller can calculate the vehicle's lateral error, lateral error rate, heading error, and heading error rate at the current moment, and generate the corresponding front wheel steering angle control values. ; in This indicates that the trajectory is generated based on the first reference line. This indicates that the trajectory is generated based on the second reference line. This represents the longitudinal control quantity output by the lateral controller, including Front wheel steering angle control amount at any time .

[0072] As an optional implementation, after obtaining the speed control amount and front wheel steering angle control amount under the first reference line trajectory, a first control command can be generated; after obtaining the speed control amount and front wheel steering angle control amount under the second reference line trajectory, a second control command can be generated.

[0073] The first and second control commands mentioned above can be represented as: ; in This indicates that the trajectory is generated based on the first reference line. This indicates that the trajectory is generated based on the first reference line. Indicates a first control command or a second control command, including Speed ​​and front wheel steering angle control at any moment .

[0074] The instruction difference between the first control instruction and the second control instruction can be expressed as: ; In S140, after obtaining the trajectory distance between the first reference line trajectory and the second reference line trajectory, as well as the instruction difference value between the first control instruction and the second control instruction, the trajectory distance can be matched with the first fault diagnosis condition, and the instruction difference value can be matched with the second fault diagnosis condition.

[0075] A reference line trajectory fault can be identified when at least one of the trajectory distance and command difference values ​​does not match the corresponding fault diagnosis condition.

[0076] In some embodiments, S140 includes: S710, if the trajectory distance is greater than the first association threshold, determine that the autonomous driving is abnormal or faulty; S720, if the trajectory distance is less than or equal to a first association threshold and greater than a distance threshold, determine that the reference line trajectory is faulty; wherein, the first association threshold is greater than the distance threshold; or, S730 determines an abnormal fault in autonomous driving if the instruction difference value is greater than the second association threshold; S740, if the instruction difference value is less than or equal to the second association threshold and greater than the instruction error threshold, a reference line trajectory fault is determined; wherein, the second association threshold is greater than the instruction error threshold.

[0077] In this embodiment, before fault diagnosis, the trajectory distance and command difference value can be correlated and judged. If either the trajectory distance or the command difference value exceeds its corresponding correlation threshold, it indicates that there is a large deviation between the two reference line trajectories, and an autonomous driving abnormality fault can be identified. After determining that no abnormality fault has occurred in autonomous driving, the trajectory distance and command difference value can be used to judge the reference line trajectory fault. When the trajectory distance is greater than the distance threshold, it can be determined that the trajectory distance does not meet the corresponding fault diagnosis condition, and a reference line trajectory fault can be identified. Similarly, when the command difference value is greater than the command error threshold, it can be determined that the command difference value does not meet the corresponding fault diagnosis condition, and a reference line trajectory fault can be identified.

[0078] In S710, the trajectory distance With distance threshold Before making the comparison, the trajectory distance can be compared with a first association threshold. This first association threshold is greater than the distance threshold.

[0079] If the trajectory distance is greater than the first association threshold When the first reference line trajectory differs significantly from the second reference line trajectory, an abnormal fault in autonomous driving can be identified.

[0080] In S720, if the trajectory distance is less than or equal to the first association threshold, the trajectory distance can be compared with a preset distance threshold after the trajectory distance is obtained. If the trajectory distance is less than the distance threshold, it can be determined that the reference line trajectory has not malfunctioned; if the trajectory distance is greater than the distance threshold, it can be determined that the reference line trajectory has malfunctioned.

[0081] In one example, after obtaining the trajectory distance... Then, the trajectory distance can be... With distance threshold The comparison can be expressed as follows: ; in, The reference line generates a fault flag bit. A value of 1 indicates a fault in the reference line trajectory. A value of 0 indicates that the reference line trajectory is normal.

[0082] In S730, the instruction difference value is compared with the instruction error threshold. Before comparison, the instruction difference value can be compared with the second correlation threshold. A comparison is made. The second correlation threshold is greater than the instruction difference value.

[0083] If the instruction difference value is greater than the second association threshold This also indicates a significant difference between the first reference line trajectory and the second reference line trajectory, at which point an abnormal fault in autonomous driving can be identified.

[0084] In S740, if the instruction difference value is less than or equal to the second association threshold, after obtaining the instruction difference value, the instruction difference value can be compared with the preset instruction error threshold. When the instruction difference value is less than the instruction error threshold, it can be determined that the reference line trajectory has not failed; while when the instruction difference value is greater than the instruction error threshold, it can be determined that the reference line trajectory has failed.

[0085] In one example, after obtaining the instruction difference value, the instruction difference value can be compared with the instruction error threshold. The comparison can be expressed as follows: ; in, The reference line generates a fault flag bit. A value of 1 indicates a fault in the reference line trajectory. A value of 0 indicates that the reference line trajectory is normal; This represents the upper bound of the instruction error threshold.

[0086] In some embodiments, S740 includes: S810, if the instruction difference value is less than or equal to the second association threshold and greater than the instruction error threshold, determine the duration for which the instruction difference value meets the threshold condition; S820 determines that the reference line trajectory is faulty when the duration reaches a duration threshold.

[0087] In this embodiment, In S810, when the instruction difference value is less than or equal to the second association threshold and greater than the instruction error threshold, in order to avoid the vehicle misjudgment caused by the fluctuation of the control quantity, the duration for which the instruction difference value meets the threshold condition can be accumulated.

[0088] In S820, when the cumulative duration reaches the duration threshold, it can be determined that the instruction difference value is not due to fluctuations that cause the threshold condition to be met, and at this time, the reference line trajectory fault can be determined.

[0089] As an optional implementation, the method of accumulating the duration for which the command difference value meets the threshold condition can be to set a sway counter. The initial value of the sway counter is 0. Each time the command difference value is detected to meet the threshold condition, the sway counter value can be incremented by 1. When the command difference value is detected to not meet the threshold condition, the sway counter value is reset to zero. By determining whether the sway counter value has reached the preset counting threshold, the fault of the reference line trajectory can be determined.

[0090] Following the aforementioned autonomous driving malfunctions, the following also includes: Send a prompt to the driver to discontinue autonomous driving, allowing the driver to take over the vehicle; or, Based on the lane line map, control the vehicle to park at the side of the road.

[0091] In this embodiment, upon determining an abnormal malfunction in autonomous driving, the detection and diagnosis process of the reference line trajectory can be terminated, and corresponding measures can be taken to protect the vehicle and its occupants. For example, the vehicle can send a prompt to the driver to disengage from autonomous driving, reminding the driver to take over the vehicle in a timely manner. The vehicle can also control itself to pull over to the side of the road based on visual information captured by the vehicle's cameras and visual odometer readings, thereby protecting the safety of the occupants.

[0092] As an optional implementation method, the following is used... Figure 5Taking the two reference line trajectories shown as examples, reference line 1 is the correct reference line, and reference line 2 is the reference line with errors.

[0093] Reference line 1 is generated based on high-precision maps and GNSS positioning, while reference line 2 is generated based on cameras and vision algorithms. Figure 5 The reference line paths for two different reference lines are shown over a time period of approximately 70 seconds. Figure 5 It also includes curves showing the changes in control parameters such as the front wheel steering angle and speed of the vehicle over time under two reference lines.

[0094] Figure 6 The diagram illustrates fault diagnosis based on trajectory distance, where the distance threshold... Set as . Figure 7 A schematic diagram illustrating fault diagnosis based on instruction difference values ​​is shown. Figure 7 The instruction difference value is the difference between the control instructions generated by the MPC trajectory control module when the trajectory information of the two reference lines is input.

[0095] like Figures 5 to 7 As shown, within a time frame of 0 to 10 seconds, the two reference lines coincide, the trajectory distance between the two reference lines is 0, and the difference in the instructions generated based on the two reference lines is 0.

[0096] Depend on Figure 5 and Figure 6 It can be seen that within the time interval of 0~10s, the trajectory distance between the two reference lines is 0, which means that the fault flag bit output by the fault diagnosis method based on the trajectory distance between the sampling points of the reference lines is 0, that is, the reference line trajectory is fault-free.

[0097] Similarly, combining Figure 5 and Figure 7 Within the time interval of 0 to 10 seconds, the instruction difference value of the control commands generated based on the two reference lines is 0, which means that the fault flag bit output by the fault diagnosis method based on the control commands is 0, that is, there is no fault in the reference line trajectory.

[0098] It should be noted that, because the MPC method predicts future trajectories, Figure 5 The data showed that the front wheel steering angle and speed control values ​​differed to varying degrees before 10 seconds.

[0099] like Figure 6 As shown, the two reference lines deviate within the time interval of 10-40 seconds. At 15 seconds, the distance between the trajectories of reference lines 1 and 2 exceeds the distance threshold. At this point, the reference line fault diagnosis flag is set to 1, which confirms that the reference line trajectory is faulty.

[0100] At 35 seconds, the distance between the trajectories of reference lines 1 and 2 returned to being less than the distance threshold. Within this range, the reference line fault diagnosis flag is at position 0, indicating that the reference line trajectory has become fault-free.

[0101] like Figure 7 As shown, at 14s, the instruction difference value exceeds the instruction error threshold. However, based on the stabilization count setting, the stabilization count has not reached the threshold at this time, so the control command fault flag is still 0, indicating that a reference line trajectory fault has not been detected. At 16s, the stabilization count reaches the threshold, the control command fault flag is set to 1, and a reference line trajectory fault is identified.

[0102] At 32 seconds, the instruction difference value decreased to less than the instruction error threshold. At this point, the anti-shake count had not been reached. Until the 34th second, the control command fault flag was set to 0, indicating that the reference line trajectory had returned to normal.

[0103] Within the time range of 40-50 seconds, such as Figure 5 and Figure 6 As shown, reference line 1 and reference line 2 coincide, and at this time the distance between the two reference lines is 0, which is lower than the distance threshold. The result of fault diagnosis based on trajectory distance is that the reference trajectory is fault-free.

[0104] However, due to differences in the historical condition of the vehicles, the control actions are constantly adjusted, and the control commands still vary considerably. For example... Figure 7 As shown, the instruction difference value is not zero within the time interval of 40-50 seconds. During this period, the instruction difference value hovers around the instruction error threshold. Even if fluctuations occur, since they do not exceed the anti-shake count, the fault diagnosis result based on the command difference value is also that the reference line trajectory has no fault.

[0105] Within the time interval of 50-70 seconds, reference line 1 and reference line 2 deviated again. Furthermore, due to the predictive nature of the MPC control method, the front wheel steering angle control values ​​under the two reference lines showed a significant difference before 50 seconds.

[0106] At 55s, the distance between the trajectory of reference line 1 and reference line 2 exceeds the distance threshold. At this time, the reference line fault diagnosis flag is set to 1, indicating that the reference line trajectory is faulty.

[0107] like Figure 7 As shown, starting from the 48th second, the command difference value exceeds the command error threshold. After the anti-shake counting, the command flag is set to 1 at the 50th second, which means that the reference line trajectory is faulty.

[0108] Please refer to Figure 8In the above embodiments, the fault diagnosis method based on reference line sampling points can sensitively diagnose whether reference lines generated by different methods have faults based on the deviation of path points under the two reference lines. Its reaction speed is affected by the distance threshold setting. Furthermore, this method requires the path points of the two reference lines to be unified in the same coordinate system. The fault diagnosis method based on control commands does not require the unification of reference line paths and focuses more on the feedback information of the vehicle's own state. However, due to the influence of the vehicle's historical state, the detection method for control commands is not sensitive enough and is affected by the autonomous driving control method and the command error threshold setting. Therefore, when the fault flag position is 1 in at least one of the above two fault diagnosis methods, a reference line trajectory fault can be determined.

[0109] Based on the same inventive concept, this application also provides a reference line trajectory detection device for autonomous driving. Specifically, in conjunction with... Figure 9 Please provide a detailed explanation.

[0110] Figure 9 This is a schematic diagram of the structure of a reference line trajectory detection device 900 for autonomous driving provided in an embodiment of this application.

[0111] like Figure 9 As shown, the reference line trajectory detection device 900 for autonomous driving may include: The acquisition module 901 is used to acquire a first reference line trajectory and a second reference line trajectory; wherein, the first reference line trajectory is obtained based on a high-precision map and vehicle satellite positioning parameters, and the second reference line trajectory is obtained based on a lane line map; the lane line map is generated based on captured road image information. The first generation module 902 is used to unify the coordinates of the first reference line trajectory and the second reference line trajectory based on the coordinate system transformation relationship between the high-precision map and the lane line map, and generate the trajectory distance of the first reference line trajectory and the second reference line trajectory after the coordinate unification. The second generation module 903 is used to generate a first control command corresponding to the first reference line trajectory and a second control command corresponding to the second reference line trajectory based on the first reference line trajectory, the second reference line trajectory and vehicle driving parameters, and to generate the command difference value between the first control command and the second control command. The judgment module 904 is used to determine the reference line trajectory detection result based on at least one of the trajectory distance and the instruction difference value.

[0112] Figure 10 A schematic diagram of the hardware structure of an electronic device provided in an embodiment of this application is shown. The electronic device can be at least one of a computer, a server, and a dedicated document generation device. The electronic device includes a processor 1001 and a memory 1002 storing computer program instructions.

[0113] Specifically, the processor 1001 may include a central processing unit (CPU), an application-specific integrated circuit (ASIC), or one or more integrated circuits that can be configured to implement the embodiments of this application.

[0114] Memory 1002 may include mass storage for data or instructions. For example, and not limitingly, memory 1002 may include a hard disk drive (HDD), floppy disk drive, flash memory, optical disk, magneto-optical disk, magnetic tape, or Universal Serial Bus (USB) drive, or a combination of two or more of these. Where appropriate, memory 1002 may include removable or non-removable (or fixed) media. Where appropriate, memory 1002 may be internal or external to an electronic device. In a particular embodiment, memory 1002 is a non-volatile solid-state memory.

[0115] Memory 1002 may include read-only memory (ROM), flash memory device, random access memory (RAM), disk storage medium device, optical storage medium device, electrical, optical or other physical / tangible memory storage device. Therefore, typically, memory 1002 includes one or more tangible (non-transitory) computer-readable storage media (e.g., memory device) of software that may be encoded with computer-executable instructions and, when executed (e.g., by one or more processors), is operable to perform the operations described with reference to the methods of the foregoing aspects of this disclosure.

[0116] The processor 1001 reads and executes computer program instructions stored in the memory 1002 to implement any of the reference line trajectory detection methods for autonomous driving in the above embodiments.

[0117] In one example, the electronic device may also include a communication interface 1003 and a bus 1010. For example, Figure 10 As shown, the processor 1001, memory 1002, and communication interface 1003 are connected through bus 1010 and complete communication with each other.

[0118] The communication interface 1003 is mainly used to realize communication between various modules, devices, units and / or equipment in the embodiments of this application.

[0119] Bus 1010 includes hardware, software, or both, that couples components of an electronic device together. For example, and not limitingly, the bus may include an Accelerated Graphics Port (AGP) or other graphics bus, an Enhanced Industry Standard Architecture (EISA) bus, a Front Side Bus (FSB), HyperTransport (HT) interconnect, an Industry Standard Architecture (ISA) bus, an Infinite Bandwidth Interconnect, a Low Pin Count (LPC) bus, a memory bus, a Microchannel Architecture (MCA) bus, a Peripheral Component Interconnect (PCI) bus, a PCI-Express (PCI-X) bus, a Serial Advanced Technology Attachment (SATA) bus, a Video Electronics Standards Association Local (VLB) bus, or other suitable buses, or combinations of two or more of these. Where appropriate, bus 1010 may include one or more buses. Although specific buses are described and illustrated in embodiments of this application, any suitable bus or interconnect is contemplated herein.

[0120] This electronic device can achieve a combination of reference line trajectory detection methods for autonomous driving. Figures 1 to 8 The method and apparatus for detecting reference line trajectories in autonomous driving are described.

[0121] Furthermore, in conjunction with the reference line trajectory detection method for autonomous driving in the above embodiments, this application embodiment can provide a computer storage medium for implementation. The computer storage medium stores computer program instructions; when these computer program instructions are executed by a processor, they implement any of the reference line trajectory detection methods for autonomous driving in the above embodiments.

[0122] In addition, this application also provides a computer program product, including a computer program, which, when executed by a processor, can implement the steps and corresponding content of the aforementioned method embodiments.

[0123] Furthermore, the term "and / or" in this article is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, or B existing alone. Additionally, the character " / " in this article generally indicates that the preceding and following related objects have an "or" relationship.

[0124] It should be understood that in the embodiments of this application, "B corresponding to A" means that B is associated with A, and B can be determined based on A. However, it should also be understood that determining B based on A does not mean that B is determined solely based on A; B can also be determined based on A and / or other information.

[0125] The above are merely specific embodiments of this application, but the scope of protection of this application is not limited thereto. Any person skilled in the art can easily conceive of various equivalent modifications or substitutions within the technical scope disclosed in this application, and these modifications or substitutions should all be covered within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A method for detecting reference line trajectories in autonomous driving, characterized in that, include: The vehicle's first reference line trajectory and second reference line trajectory are obtained; wherein, the first reference line trajectory is obtained based on a preset high-precision map and the vehicle's positioning parameters, and the second reference line trajectory is obtained based on a lane line map; the lane line map is generated based on road image information around the vehicle. Based on the coordinate system transformation relationship between the high-precision map and the lane line map, the coordinates of the first reference line trajectory and the second reference line trajectory are unified, and the trajectory distance between the first reference line trajectory and the second reference line trajectory after coordinate unification is generated. Based on the first reference line trajectory, the second reference line trajectory, and vehicle driving parameters, a first control command corresponding to the first reference line trajectory and a second control command corresponding to the second reference line trajectory are generated respectively, and the command difference value between the first control command and the second control command is generated. The reference line trajectory detection result is determined based on at least one of the trajectory distance and the instruction difference value.

2. The reference line trajectory detection method for autonomous driving according to claim 1, characterized in that, The step of unifying the coordinates of the first reference line trajectory and the second reference line trajectory based on the coordinate system transformation relationship between the high-precision map and the lane line map includes: Obtain the coordinate parameters of the vehicle and its surrounding feature points in the lane map and the high-precision map, respectively. Based on the coordinate parameters, determine the coordinate system transformation relationship between the high-precision map and the lane line map; Based on the coordinate system transformation relationship, coordinate transformation is performed on multiple first sampling points of the first reference line trajectory or multiple second sampling points of the second reference line trajectory.

3. The reference line trajectory detection method for autonomous driving according to claim 2, characterized in that, The acquisition of the coordinate parameters of the vehicle and its surrounding feature points in the lane map and the high-precision map, respectively, includes: Acquire the vehicle's location point information in the lane map and the vehicle's satellite positioning parameters in the high-precision map at the same time. Based on the collected road image information, determine the coordinate parameters of the corresponding feature points in the lane line map, and obtain the coordinate parameters of the feature points matched in the high-precision map.

4. The reference line trajectory detection method for autonomous driving according to claim 2, characterized in that, The distance between the first reference line trajectory and the second reference line trajectory after coordinate unification includes: Determine the first sampling point and the second sampling point corresponding to the same moment from the first reference line trajectory and the second reference line trajectory; the first sampling point and the second sampling point contain coordinate parameters in the same coordinate system; Generate the distance between the first sampling point and the second sampling point.

5. The reference line trajectory detection method for autonomous driving according to claim 1, characterized in that, Obtain the trajectory of the first reference line, including: Based on the starting and ending points of autonomous driving, a corresponding global navigation path is determined; wherein, the global navigation path includes road segment information of multiple road segments; The target road segment is determined from multiple road segments based on the vehicle's satellite positioning parameters. Based on the road segment information of the target road segment, the corresponding lane centerline is obtained from the high-precision map to form a first reference line trajectory.

6. The reference line trajectory detection method for autonomous driving according to claim 1, characterized in that, Obtain the trajectory of the second reference line, including: Based on the captured road image information, multiple lane sampling points are determined to form a visually detected lane line composed of multiple lane sampling points; In the absence of a lane line map generated for the vehicle, the lane line map is generated based on the visually detected lane lines. If a lane map has already been generated for the vehicle, the original lane lines in the lane map are associated with the visually detected lane lines, and the lane map is optimized based on the association results. The second reference line trajectory is obtained based on the lane line map.

7. The reference line trajectory detection method for autonomous driving according to claim 1, characterized in that, The step of generating a first control command corresponding to the first reference line trajectory and a second control command corresponding to the second reference line trajectory based on the first reference line trajectory, the second reference line trajectory, and vehicle driving parameters includes: Based on the vehicle's position and speed information, the first longitudinal control quantity corresponding to the first reference line trajectory and the second longitudinal control quantity corresponding to the second reference line trajectory are determined respectively. Based on the vehicle's position information and heading angle information, the first lateral control quantity corresponding to the first reference line trajectory and the second lateral control quantity corresponding to the second reference line trajectory are determined respectively.

8. The reference line trajectory detection method for autonomous driving according to claim 7, characterized in that, Determining the reference line trajectory detection result based on at least one of the trajectory distance and the instruction difference value includes: If the trajectory distance is greater than a first association threshold, an autonomous driving malfunction is determined. If the trajectory distance is less than or equal to the first association threshold and greater than the distance threshold, a reference line trajectory fault is determined; wherein, the first association threshold is greater than the distance threshold; or, If the instruction difference value is greater than the second association threshold, an autonomous driving abnormality fault is determined; If the instruction difference value is less than or equal to the second association threshold and greater than the instruction error threshold, a reference line trajectory fault is determined; wherein the second association threshold is greater than the instruction error threshold.

9. The reference line trajectory detection method for autonomous driving according to claim 8, characterized in that, The step of determining a reference line trajectory fault when the instruction difference value is less than or equal to the second association threshold and greater than the instruction error threshold includes: If the instruction difference value is less than or equal to the second association threshold and greater than the instruction error threshold, determine the duration for which the instruction difference value satisfies the threshold condition; If the duration reaches a duration threshold, a fault is determined in the reference line trajectory.

10. A reference line trajectory detection device for autonomous driving, characterized in that, include: The acquisition module is used to acquire a first reference line trajectory and a second reference line trajectory; wherein, the first reference line trajectory is obtained based on a high-precision map and vehicle satellite positioning parameters, and the second reference line trajectory is obtained based on a lane line map; the lane line map is generated based on captured road image information. The first generation module is used to unify the coordinates of the first reference line trajectory and the second reference line trajectory based on the coordinate system transformation relationship between the high-precision map and the lane line map, and generate the trajectory distance between the first reference line trajectory and the second reference line trajectory after coordinate unification. The second generation module is used to generate a first control command corresponding to the first reference line trajectory and a second control command corresponding to the second reference line trajectory based on the first reference line trajectory, the second reference line trajectory and vehicle driving parameters, and to generate a command difference value between the first control command and the second control command. The judgment module is used to determine the reference line trajectory detection result based on at least one of the trajectory distance and the instruction difference value.

Images