Control method and device of automatic driving vehicle and automatic driving system

Abstract

The present application provides a control method, device and automatic driving system for an autonomous driving vehicle, the method comprising: during the driving process of the autonomous driving vehicle, when the autonomous driving vehicle meets predetermined conditions, calibrating the predetermined parameters of the target sensor to obtain the target parameters, wherein the predetermined parameters include at least one of the yaw angle and the pitch angle, and the predetermined conditions include: the rate of change of the yaw angle of the autonomous driving vehicle is less than a first threshold, and the rate of change of the predetermined acceleration is less than a second threshold, and / or the rate of change of the slope of the road where the autonomous driving vehicle is located is less than a third threshold; when the target parameter is within a first predetermined range, correcting the position information of the obstacle measured by the target sensor according to the target parameter; when the target parameter is not within the first predetermined range, reducing the automatic driving level of the autonomous driving vehicle. The present application solves the problem of inaccurate sensor positioning of autonomous driving vehicles as the driving time increases.

Description

Technical Field

[0001] The present application relates to the field of autonomous driving, and more specifically, to a control method and device for an autonomous driving vehicle, a computer-readable storage medium, a processor, and an autonomous driving system. Background Art

[0002] Nowadays, self-driving vehicles basically use sensors such as cameras, millimeter-wave radars, lidars, and high-precision maps, as well as many other technologies to ensure that self-driving vehicles can accurately locate themselves and surrounding obstacles, thereby improving the safety of self-driving. In the multi-sensor fusion algorithm modules in the self-driving algorithm, many of them directly use the obstacle position information sent by the sensor for data processing and fusion. However, in the actual vehicle driving process, the probability of the sensor's physical installation position changing as the driving time increases, and its installation angle also greatly increases. At this time, the positioning of surrounding obstacles is no longer accurate, and in serious cases, even self-driving safety accidents may occur.

[0003] Therefore, there is an urgent need for a method to solve the problem of inaccurate positioning of surrounding obstacles by autonomous vehicles as driving time increases.

[0004] The above information disclosed in the background technology section is only used to enhance the understanding of the background technology of the technology described in this article. Therefore, the background technology may contain certain information that does not form the prior art known in this country for those skilled in the art. Summary of the invention

[0005] The main purpose of this application is to provide a control method, device, computer-readable storage medium, processor and autonomous driving system for an autonomous driving vehicle to solve the problem of inaccurate sensor positioning of the autonomous driving vehicle as the driving time increases in the prior art.

[0006] According to one aspect of an embodiment of the present application, a control method for an autonomous driving vehicle is provided, comprising: during driving of the autonomous driving vehicle, when the autonomous driving vehicle satisfies predetermined conditions, calibrating predetermined parameters of a target sensor to obtain target parameters, wherein the predetermined parameters include at least one of a yaw angle and a pitch angle, and when the predetermined parameters include the yaw angle, the predetermined conditions include: a rate of change of the yaw angle of the autonomous driving vehicle is less than a first threshold, and a rate of change of a predetermined acceleration is less than a second threshold, the predetermined acceleration is the acceleration of the autonomous driving vehicle perpendicular to the driving direction, and when the predetermined parameters include the pitch angle, the predetermined conditions include: a rate of change of the slope of a road where the autonomous driving vehicle is located is less than a third threshold; when the target parameter is within a first predetermined range, correcting the position information of the obstacle measured by the target sensor according to the target parameter; and when the target parameter is not within the first predetermined range, reducing the autonomous driving level of the autonomous driving vehicle.

[0007] Optionally, the predetermined parameters of the target sensor are calibrated to obtain the target parameters, including: calibrating the yaw angle according to a frame of data collected by the target sensor and a boundary line equation to obtain a calibrated yaw angle, the boundary line equation being an equation of the boundary line of the road where the autonomous driving vehicle is located determined according to a high-precision map; performing statistical feature analysis on the calibrated yaw angles corresponding to multiple consecutive frames of data to obtain a final yaw angle.

[0008] Optionally, the yaw angle is calibrated according to a frame of data collected by the target sensor and a boundary line equation to obtain a calibrated yaw angle, including: determining the position information and type of multiple first stationary obstacles according to the one frame of data, the first stationary obstacles are obstacles located on both sides of the road and do not move; when the boundary line equation is a straight line equation and the number of first target obstacles is greater than a fourth threshold, determining the calibrated yaw angle according to the boundary line equation and the position information of multiple first target obstacles, wherein the first target obstacle is the first stationary obstacle of the same type; when the boundary line equation is not the straight line equation and the number of the first stationary obstacles of the same type is greater than the fourth threshold, determining the calibrated yaw angle according to the position information of multiple first target obstacles.

[0009] Optionally, the calibrated yaw angle is determined according to the boundary line equation and the position information of multiple first target obstacles, including: performing least squares fitting on the position information of the first target obstacles located on the same side of the road to obtain a first fitting line equation; and determining the calibrated yaw angle according to the first fitting line equation and the boundary line equation.

[0010] Optionally, the road includes a relative first boundary line and a second boundary line, the boundary line equation includes a first boundary line sub-equation and a second boundary line sub-equation, and when a plurality of the first target obstacles are located on different sides of the road, the first fitting line equation includes a first fitting line sub-equation and a second fitting line sub-equation, the first boundary line sub-equation and the first fitting line sub-equation correspond to the first boundary line respectively, and the second boundary line sub-equation and the second fitting line sub-equation correspond to the second boundary line respectively, and the calibrated yaw angle is determined according to the first fitting line equation and the boundary line equation, including: when the first target obstacle is located on different sides of the road, determining the angle between the first boundary line sub-equation and the first fitting line sub-equation as a first angle, and the angle between the second boundary line sub-equation and the second fitting line sub-equation as a second angle; when one of the first angle and the second angle is an invalid angle, Determine that the valid angle between the first angle and the second angle is the calibrated yaw angle, the number of the first target obstacles corresponding to the invalid angle is less than or equal to the fourth threshold, and the number of the first target obstacles corresponding to the valid angle is greater than the fourth threshold; when the first angle and the second angle are both the valid angles, and the difference between the first angle and the second angle is within a second predetermined range, determine that the average value of the first angle and the second angle is the calibrated yaw angle; when the first angle and the second angle are both the valid angles, and the difference between the first angle and the second angle is not within the second predetermined range, respectively obtain the boundary line distance between the autonomous driving vehicle and the first boundary line and the boundary line distance between the autonomous driving vehicle and the second boundary line, and determine that the angle corresponding to the minimum value of the boundary line distance between the first angle and the second angle is the calibrated yaw angle.

[0011] Optionally, the road includes a relative first boundary line and a second boundary line, the boundary line equation includes a first boundary line sub-equation corresponding to the first boundary line and a second boundary line sub-equation corresponding to the second boundary line, and the calibrated yaw angle is determined according to the first fitting line equation and the boundary line equation, including: when all the first target obstacles are located on the same side of the road, obtaining a first distance and a second distance, the first distance being the distance between the first boundary line sub-equation and the first fitting line equation, and the second distance being the distance between the second boundary line sub-equation and the first fitting line equation; determining that the sub-equation corresponding to the minimum value between the first distance and the second distance in the first boundary line sub-equation and the second boundary line sub-equation is the target sub-boundary line equation; and determining that the angle between the first fitting line equation and the target sub-boundary line equation is the calibrated yaw angle.

[0012] Optionally, determining the calibrated yaw angle based on the position information of multiple first target obstacles includes: determining that the distance between the position information of the second target obstacle and the boundary line equation is a third distance, and calculating the average value of each third distance, the second target obstacle is all the first target obstacles located on the same side of the road; determining whether the average value is less than a fifth threshold; when the average value is greater than or equal to the fifth threshold, determining the calibrated yaw angle based on the yaw angle, a third predetermined range and a preset angle step; when the average value is less than the fifth threshold, determining the yaw angle to be the calibrated yaw angle.

[0013] Optionally, when the average value is greater than or equal to the fifth threshold, the calibrated yaw angle is determined according to the yaw angle, the third predetermined range and the preset angle step, including: an updating step, when the average value is greater than or equal to the fifth threshold, starting from the yaw angle, increasing or decreasing the yaw angle by the preset angle step to obtain an updated angle, and the updated angle is within the third predetermined range; a correction step, according to the updated angle, correcting the position information of each second target obstacle to obtain the corrected position information; a calculation step, calculating the average distance between the corrected position information and the boundary line equation; a first determination step, determining whether the distance average is less than the fifth threshold; a second determination step, when the distance average is less than the fifth threshold, determining that the updated angle is the calibrated yaw angle, and when the distance average is greater than or equal to the fifth threshold, sequentially executing the updating step, the correction step, the calculation step and the first determination step until the distance average is less than the fifth threshold or the update angle reaches the boundary of the third predetermined range.

[0014] Optionally, before calibrating the predetermined parameters of the target sensor to obtain the target parameters, the method also includes: obtaining a sixth threshold, the speed of the autonomous driving vehicle, and the speed of each of the obstacles on both sides of the road; determining, based on the speed and each of the speeds, that the obstacle whose speed is opposite to the direction of the vehicle speed and whose predetermined difference is within a fifth predetermined range is an initial stationary obstacle, and the predetermined difference is the absolute value of the difference between the absolute value of the speed and the absolute value of the vehicle speed; calculating, based on the position information of the initial stationary obstacle and the boundary line equation, the distance between each of the initial stationary obstacles and the boundary line equation to obtain multiple fourth distances; determining that the initial stationary obstacle corresponding to the fourth distance less than the sixth threshold is the first stationary obstacle.

[0015] Optionally, the predetermined parameters of the target sensor are calibrated to obtain the target parameters, including: extracting a second stationary obstacle from the continuous multiple frames of data collected by the target sensor, and obtaining a first height, where the first height is the height information of the second stationary obstacle measured by the target sensor, and the second stationary obstacle is an obstacle located on the road or on both sides of the road, and the distance from the autonomous driving vehicle is within a fourth predetermined range and does not move; obtaining a fitting relationship between the second height and the fifth distance to obtain a second fitting line equation, where the second height is the height information of the second stationary obstacle determined according to a high-precision map, and the fifth distance is the real-time distance between the second stationary obstacle and the autonomous driving vehicle in the driving direction; when the number of the second stationary obstacles is greater than a seventh threshold, fitting the first height and the fifth distance to obtain a third fitting line equation; and determining the angle between the second fitting line equation and the third fitting line equation as the final pitch angle.

[0016] According to another aspect of an embodiment of the present application, a control device for an autonomous driving vehicle is further provided, comprising: a calibration unit, for calibrating predetermined parameters of a target sensor to obtain target parameters during the driving of the autonomous driving vehicle, when the autonomous driving vehicle satisfies predetermined conditions, wherein the predetermined parameters include at least one of a yaw angle and a pitch angle, and when the predetermined parameters include the yaw angle, the predetermined conditions include: a rate of change of the yaw angle of the autonomous driving vehicle is less than a first threshold, and a rate of change of a predetermined acceleration is less than a second threshold, the predetermined acceleration is the acceleration of the autonomous driving vehicle perpendicular to the driving direction, and when the predetermined parameters include the pitch angle, the predetermined conditions include: a rate of change of the slope of the road where the autonomous driving vehicle is located is less than a third threshold; a correction unit, for correcting the position information of the obstacle measured by the target sensor according to the target parameters when the target parameters are within a first predetermined range; a reduction unit, for reducing the autonomous driving level of the autonomous driving vehicle when the target parameters are not within the first predetermined range.

[0017] According to another aspect of the embodiments of the present application, a computer-readable storage medium is provided, wherein the computer-readable storage medium includes a stored program, wherein the program executes any one of the methods described.

[0018] According to another aspect of the embodiments of the present application, a processor is provided, wherein the processor is used to run a program, wherein any one of the methods is executed when the program is run.

[0019] According to another aspect of an embodiment of the present application, an autonomous driving system is also provided, comprising an autonomous driving vehicle, one or more processors, a memory, and one or more programs, wherein the one or more programs are stored in the memory and are configured to be executed by the one or more processors, and the one or more programs include methods for executing any one of the methods described.

[0020] In an embodiment of the present application, a control method for an autonomous driving vehicle, first, during the driving process of the autonomous driving vehicle, when the autonomous driving vehicle satisfies that the rate of change of the yaw angle of the autonomous driving vehicle is less than a first threshold value and the rate of change of the predetermined acceleration is less than a second threshold value, the yaw angle of the sensor is calibrated, and / or when the autonomous driving vehicle satisfies that the rate of change of the slope of the road where the autonomous driving vehicle is located is less than a third threshold value, the pitch angle of the sensor is calibrated to obtain a target parameter, thereby ensuring that the attitude information of the calibrated sensor is relatively accurate; then, when the target parameter is within a first predetermined range, the position information of the obstacle measured by the sensor is corrected according to the target parameter, thereby ensuring that the sensor accurately locates the obstacles around the vehicle, thereby ensuring the driving safety of the autonomous driving vehicle; when the target parameter is not within the first predetermined range, it means that the positioning of the sensor is inaccurate and the risk of autonomous driving is high. At this time, the autonomous driving level of the autonomous driving vehicle is reduced, thereby avoiding driving accidents caused by inaccurate positioning of the sensor. The present application effectively solves the problem of inaccurate sensor positioning of autonomous driving vehicles in the prior art as the driving time increases. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] The drawings constituting part of the present application are used to provide a further understanding of the present application. The exemplary embodiments and descriptions of the present application are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings:

[0022] Figure 1 A schematic diagram showing a flow chart of a method for controlling an autonomous driving vehicle according to an embodiment of the present application is shown;

[0023] Figure 2 A schematic diagram showing the positional relationship between a road boundary line and an obstacle in a horizontal straight road scenario according to an embodiment of the present application is shown;

[0024] Figure 3 A schematic diagram showing the positional relationship between a road boundary line and an obstacle in a horizontal non-straight road scenario according to an embodiment of the present application is shown;

[0025] Figure 4 A graph showing the relationship between the height of an obstacle and the distance between the obstacle and the vehicle according to an embodiment of the present application is shown;

[0026] Figure 5 A schematic diagram showing a control device for an autonomous driving vehicle according to an embodiment of the present application is shown;

[0027] Figure 6 A schematic diagram of the structure of a multi-sensor fusion algorithm module according to an embodiment of the present application is shown;

[0028] Figure 7 A schematic diagram of a horizontal calibration process according to an embodiment of the present application is shown;

[0029] Figure 8 A schematic diagram of a pitch calibration process according to an embodiment of the present application is shown. DETAILED DESCRIPTION

[0030] It should be noted that, in the absence of conflict, the embodiments and features in the embodiments of the present application can be combined with each other. The present application will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.

[0031] In order to enable those skilled in the art to better understand the solution of the present application, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. 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 creative work should fall within the scope of protection of the present application.

[0032] It should be noted that the terms "first", "second", etc. in the specification and claims of the present application and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequential order. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments of the present application described here. In addition, the terms "including" and "having" and any of their variations are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device comprising a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0033] It should be understood that when an element (such as a layer, film, region, or substrate) is described as being "on" another element, the element may be directly on the other element, or there may be intermediate elements. Moreover, in the specification and claims, when an element is described as being "connected" to another element, the element may be "directly connected" to the other element, or "connected" to the other element through a third element.

[0034] As mentioned in the background technology, in the prior art, as the driving time increases, the sensor positioning of the autonomous driving vehicle becomes inaccurate. In order to solve the above problem, in a typical embodiment of the present application, a control method, device, computer-readable storage medium, processor and autonomous driving system for an autonomous driving vehicle are provided.

[0035] According to an embodiment of the present application, a control method for an autonomous driving vehicle is provided.

[0036] Figure 1 : is a flow chart of a control method for an autonomous driving vehicle according to an embodiment of the present application. Figure 1 As shown, the method comprises the following steps:

[0037] Step S101: During the driving of the autonomous driving vehicle, if the autonomous driving vehicle meets predetermined conditions, calibrate predetermined parameters of the target sensor to obtain target parameters, wherein the predetermined parameters include at least one of a yaw angle and a pitch angle. If the predetermined parameters include the yaw angle, the predetermined conditions include: a rate of change of the yaw angle of the autonomous driving vehicle is less than a first threshold, and a rate of change of a predetermined acceleration is less than a second threshold, and the predetermined acceleration is an acceleration of the autonomous driving vehicle perpendicular to the driving direction. If the predetermined parameters include the pitch angle, the predetermined conditions include: a rate of change of the slope of the road where the autonomous driving vehicle is located is less than a third threshold.

[0038] In the above process, when the rate of change of the yaw angle of the above-mentioned autonomous driving vehicle is less than the first threshold value, and the rate of change of the predetermined acceleration is less than the second threshold value, it means that the autonomous driving vehicle is driving smoothly in a relatively horizontal direction. At this time, calibrating the yaw angle of the target sensor can make the calibrated yaw angle more accurate; when the rate of change of the slope of the road where the autonomous driving vehicle is located is less than the third threshold value, it means that the pitch angle of the autonomous driving vehicle changes little. At this time, calibrating the pitch angle of the target sensor can make the calibrated pitch angle more accurate.

[0039] The above-mentioned yaw angle of the autonomous driving vehicle, the above-mentioned predetermined acceleration and the above-mentioned slope are all obtained after filtering processing.

[0040] Depending on the location where the target sensor is installed on the autonomous driving vehicle, in some cases it can detect both sides of the road at the same time, while in some cases it can only detect one side of the road. This application applies to both situations.

[0041] In order to further ensure a better calibration effect of the yaw angle, according to another specific embodiment of the present application, when the above-mentioned predetermined parameters include the above-mentioned yaw angle, the above-mentioned predetermined conditions include: the change rate of the yaw angle of the above-mentioned autonomous driving vehicle is less than a first threshold, the change rate of the predetermined acceleration is less than a second threshold, and the change rate of the slope of the above-mentioned road is less than a third threshold, which can further avoid the interference of the slope on the yaw angle calibration result.

[0042] In addition, when the above-mentioned predetermined parameters include at least one of the yaw angle and the pitch angle, the above-mentioned predetermined conditions also include: the vehicle speed of the above-mentioned automatic driving vehicle is greater than the eighth threshold value.

[0043] The above-mentioned target sensors may include laser radar (Lidar) and / or millimeter wave radar (Radar). Among them, laser radar is used to collect point cloud information or object information of obstacles around the vehicle; millimeter wave radar has height measurement capability, and is also used to collect point cloud information or object information of obstacles around the vehicle. The dynamic calibration principles of laser radar and laser radar are similar, and the dynamic calibration algorithm data processing of point cloud information and object information is also similar. The difference is that the amount of point cloud data is large, the cache space occupied is also larger, the algorithm running time is slightly longer, and the accuracy of the calibration result will be slightly better, but not much different. Technical personnel in this field can choose the type of sensor to be used according to their own hardware platform resources and sensor data information type.

[0044] The installation positions of LiDAR and LiDAR are different, but their obstacle position information will be unified into the vehicle coordinate system, and the final dynamic calibration principle is similar, so the installation position is not affected.

[0045] In actual application, during the control process of the autonomous vehicle, various sensors will be used to collect information about the vehicle such as vehicle speed, yaw angle, lateral acceleration and longitudinal acceleration in real time. In addition, high-precision maps will be used to collect information about road slopes at different distances ahead, road boundaries, lane lines, fixed-height obstacles such as gantries and road signs, and obstacles such as railings, flower beds and walls on both sides of the road. This information is processed by a multi-sensor fusion algorithm and used as reference data for controlling the driving of the autonomous vehicle.

[0046] The data obtained based on the high-precision map described below in this application are all data of the high-precision map in a valid and usable state.

[0047] In one embodiment of the present application, when the predetermined parameters include the yaw angle, calibrating the predetermined parameters of the target sensor to obtain the target parameters includes the following specific steps:

[0048] Step S201, calibrating the yaw angle according to a frame of data collected by the target sensor and a boundary line equation to obtain a calibrated yaw angle, wherein the boundary line equation is an equation of a boundary line of the road where the autonomous driving vehicle is located determined according to a high-precision map;

[0049] Step S202 , performing statistical characteristic analysis on the calibrated yaw angle corresponding to a plurality of consecutive frames of data to obtain a final yaw angle.

[0050] The above process determines the calibrated yaw angle corresponding to multiple consecutive frames through the data of each consecutive frame of the target sensor and the boundary line equation. After statistical analysis of the calibrated yaw angle corresponding to multiple consecutive frames, inaccurate values ​​in multiple calibrated yaw angles can be eliminated, so that the final yaw angle obtained is more accurate. In addition, the above process also realizes the dynamic calibration of the yaw angle of the target sensor during the driving process of the autonomous driving vehicle, further ensuring that the obstacle information obtained from the target sensor is relatively accurate, thereby further ensuring the driving safety of the autonomous driving vehicle.

[0051] The above step S201 specifically includes:

[0052] Step S301, determining the position information and type of a plurality of first stationary obstacles according to one frame of the data, wherein the first stationary obstacles are obstacles located on both sides of the road and do not move;

[0053] The above types may include railings, flower beds, and walls. In actual applications, the boundary line equation of the road boundary line may be as follows Figure 2 The straight line shown in the figure indicates that the autonomous vehicle is on a straight road. Figure 3 The curve shown here indicates that the autonomous vehicle is in a non-straight road scenario. The calibration process of the yaw angle is different for boundary line equations of different forms, as described below:

[0054] Step S302, when the boundary line equation is a straight line equation and the number of first target obstacles is greater than a fourth threshold, determining the calibrated yaw angle according to the boundary line equation and position information of a plurality of first target obstacles, wherein the first target obstacle is the first stationary obstacle of the same type as the first target obstacle;

[0055] In the horizontal straight road scenario, the rate of change of the yaw angle of the above-mentioned autonomous driving vehicle is less than the first threshold, which means that the yaw angle fluctuates stably in a small range near 0, and the rate of change of the predetermined acceleration is less than the second threshold, which means that the predetermined acceleration fluctuates in a small range near 0.

[0056] Step S303: when the boundary line equation is not the straight line equation and the number of the first stationary obstacles of the same type is greater than the fourth threshold, determine the calibrated yaw angle according to the position information of the plurality of first target obstacles.

[0057] In the horizontal non-straight road scenario, the rate of change of the yaw angle of the above-mentioned autonomous driving vehicle is less than the first threshold, which means that the yaw angle fluctuates stably in a small range near a non-zero value, and the rate of change of the preset acceleration is less than the second threshold, which means that the preset acceleration fluctuates in a small range near a non-zero value.

[0058] The specific principle of the above embodiment is: during the driving process of the autonomous driving vehicle, the target sensor is used to detect the position information of the surrounding absolutely stationary obstacles such as railings, flower beds, walls, etc., combined with the vehicle's motion trajectory, posture and road slope information. If the yaw angle of the target sensor is normal, then the position distribution of the absolutely stationary obstacles and the road boundary line of the high-precision map are parallel or can overlap after translation. For details, please refer to Figure 2 and Figure 3 If the yaw angle of the target sensor is abnormal, then within the allowable angle range, the position of the absolutely stationary obstacle can be rotated by a certain angle α, parallelized or translated so that it coincides with the road boundary line. The angle α is the yaw angle of the target sensor after calibration.

[0059] The size of the fourth threshold value depends on the type of the first target obstacle. In the above frame of data acquired by the target sensor, the fourth threshold value corresponding to the case where the first target obstacle is a railing, the fourth threshold value corresponding to the case where the first target obstacle is a wall, and the fourth threshold value corresponding to the case where the first target obstacle is a flower bed increase in sequence, which can further ensure that the subsequent yaw angle calibration results are more accurate and reliable.

[0060] It should be noted that, in the above step S302 and the above step S303, as long as the number of one type of first target obstacles is greater than the above fourth threshold, the calibrated yaw angle can be determined according to the position information of the first target obstacles of this type. If in a frame of data, there are multiple types of first target obstacles whose number is greater than the above fourth threshold, in order to simplify the calculation process, those skilled in the art can select one type of first target obstacles greater than the first threshold to calibrate the yaw angle, and can also calibrate the yaw angle according to each type of first target obstacle, so as to further ensure good calibration accuracy.

[0061] In addition, when the number of first target obstacles of all types does not satisfy the fourth threshold, the yaw angle calibration is terminated.

[0062] In order to further ensure the stability of the final yaw angle, the specific implementation of the above step S202 can be: calculate the average value of the above calibrated yaw angle corresponding to the above data for multiple consecutive frames to obtain the above final yaw angle. The specific implementation of the above step S202 can also be: calculate the mode of the above calibrated yaw angle corresponding to the above data for multiple consecutive frames to obtain the above final yaw angle. The specific steps of the above step S202 can also be: calculate the median of the above calibrated yaw angle corresponding to the above data for multiple consecutive frames to obtain the above final yaw angle.

[0063] In other embodiments, the types and quantities of the plurality of first stationary obstacles are not limited to being obtained from the frame data collected by the target sensor, but may also be extracted from a high-precision map. That is, calibrating the predetermined parameters of the target sensor to obtain the target parameters may also include: calibrating the yaw angle according to a frame of data collected by the target sensor and the high-precision map to obtain a calibrated yaw angle; and performing statistical feature analysis on the calibrated yaw angle corresponding to a plurality of consecutive frames of the data to obtain a final yaw angle.

[0064] Then, according to a frame of data collected by the target sensor and the high-precision map, the yaw angle is calibrated to obtain the calibrated yaw angle in the following process: according to the high-precision map, the types of multiple first stationary obstacles and the boundary line equation of the road boundary line are determined, and the first stationary obstacles are obstacles located on both sides of the road and do not move; according to one frame of the data, the position information of multiple first stationary obstacles is determined; the first stationary obstacles of the same type are determined as first target obstacles, and the number of the first target obstacles corresponding to each type is calculated; when the boundary line equation is a straight line equation and the number of the first target obstacles is greater than the fourth threshold, the calibrated yaw angle is determined according to the boundary line equation and the position information of the multiple first target obstacles, wherein the first target obstacle is the first stationary obstacle of the same type; when the boundary line equation is not the straight line equation and the number of the first stationary obstacles of the same type is greater than the fourth threshold, the calibrated yaw angle is determined according to the position information of the multiple first target obstacles.

[0065] In one embodiment of the present application, determining the calibrated yaw angle according to the boundary line equation and the position information of the plurality of first target obstacles includes:

[0066] Step S3021, performing least square fitting on the position information of the first target obstacle located on the same side of the road to obtain a first fitting line equation;

[0067] Step S3022: determining the calibrated yaw angle according to the first fitting line equation and the boundary line equation.

[0068] In a straight road scenario, the position information of the first target obstacle on the same side of the road is fitted by least squares, so that the straight line equation of the first target obstacle, that is, the first fitting line equation mentioned above, can be obtained relatively simply and quickly. Then, the angle between the straight line equation and the boundary line equation of the straight line is calculated to obtain the calibrated yaw angle.

[0069] In actual applications, during the driving process of an autonomous vehicle, due to factors such as obstacles on the road and the position of the autonomous vehicle on the road, the target sensor can only collect obstacle data on one side of the road. In this case, a fitting line equation and the boundary line equation of the corresponding side can be directly calculated. In other cases, the autonomous vehicle may be able to collect obstacle data on both sides of the road, but the obstacle data on one side is not clear enough or incomplete. It may also be possible to collect obstacle data on both sides of the road relatively clearly and completely. In this case, the specific execution method can be:

[0070] The road includes a relative first boundary line and a second boundary line, the boundary line equation includes a first boundary line sub-equation and a second boundary line sub-equation, and in the case where the plurality of first target obstacles are located on different sides of the road, the first fitting line equation includes a first fitting line sub-equation and a second fitting line sub-equation, the first boundary line sub-equation and the first fitting line sub-equation correspond to the first boundary line respectively, and the second boundary line sub-equation and the second fitting line sub-equation correspond to the second boundary line respectively. The specific implementation method of the step S3022 may include: in the case where the first target obstacle is located on different sides of the road, determining that the angle between the first boundary line sub-equation and the first fitting line sub-equation is the first angle, and the angle between the second boundary line sub-equation and the second fitting line sub-equation is the second angle; in the case where one of the first angle and the second angle is an invalid angle, determining the first angle The valid angle among the above-mentioned second angles is the above-mentioned calibrated yaw angle, the number of the above-mentioned first target obstacles corresponding to the above-mentioned invalid angle is less than or equal to the above-mentioned fourth threshold value, and the number of the above-mentioned first target obstacles corresponding to the above-mentioned valid angle is greater than the above-mentioned fourth threshold value; when the above-mentioned first angle and the above-mentioned second angle are both the above-mentioned valid angles, and the difference between the above-mentioned first angle and the above-mentioned second angle is within the second predetermined range, the average value of the above-mentioned first angle and the above-mentioned second angle is determined to be the above-mentioned calibrated yaw angle; when the above-mentioned first angle and the above-mentioned second angle are both the above-mentioned valid angles, and the difference between the above-mentioned first angle and the above-mentioned second angle is not within the above-mentioned second predetermined range, the boundary line distance between the above-mentioned automatic driving vehicle and the above-mentioned first boundary line and the boundary line distance between the above-mentioned automatic driving vehicle and the above-mentioned second boundary line are respectively obtained, and it is determined that the angle corresponding to the minimum value of the above-mentioned boundary line distances among the above-mentioned first angle and the above-mentioned second angle is the above-mentioned calibrated yaw angle.

[0071] In the above processing method, when only one of the first angle and the second angle is valid, the valid angle is used as the calibrated yaw angle; when both are valid angles and the difference is not large, the average value of the two is taken as the calibrated yaw angle; when both are valid angles and the difference is large, the angle corresponding to the road boundary line that the autonomous driving vehicle is approaching is taken as the calibrated yaw angle. The above processing method can further achieve a more accurate calibration of the yaw angle.

[0072] In addition, when the first angle and the second angle are both the invalid angles, the yaw angle calibration is completed.

[0073] In the embodiment of the present application, the specific implementation of the above step 303 may also include: determining the distance between the position information of the second target obstacle and the above boundary line equation as a third distance, and calculating the average value of each of the above third distances, the above second target obstacle is all the above first target obstacles located on the same side of the above road; determining whether the above average value is less than the fifth threshold; when the above average value is greater than or equal to the above fifth threshold, determining the above calibrated yaw angle according to the above yaw angle, the third predetermined range and the preset angle step; when the above average value is less than the above fifth threshold, determining the above yaw angle as the above calibrated yaw angle. When the above average value is less than the above fifth threshold, it means that the current yaw angle of the target sensor is normal and no further calibration is required; when the above average value is greater than or equal to the above fifth threshold, the yaw angle is calibrated according to the current yaw angle, the preset third predetermined range and the preset angle step, further realizing the accurate calibration of the yaw angle in the non-straight road scenario.

[0074] According to another specific embodiment of the present application, when the above-mentioned average value is greater than or equal to the above-mentioned fifth threshold value, the above-mentioned calibrated yaw angle is determined according to the above-mentioned yaw angle, the third predetermined range and the preset angle step, including: an updating step, when the above-mentioned average value is greater than or equal to the above-mentioned fifth threshold value, starting from the above-mentioned yaw angle, the above-mentioned yaw angle is increased or decreased by the above-mentioned preset angle step to obtain an updated angle, and the above-mentioned updated angle is located in the above-mentioned third predetermined range; a correction step, according to the above-mentioned update angle, correcting the position information of each of the above-mentioned second target obstacles to obtain the corrected position information; a calculation step, calculating the average distance between the corrected position information and the above-mentioned boundary line equation; a first determination step, determining whether the above-mentioned distance average value is less than the above-mentioned fifth threshold value; a second determination step, when the above-mentioned distance average value is less than the above-mentioned fifth threshold value, determining that the above-mentioned update angle is the above-mentioned calibrated yaw angle, and when the above-mentioned distance average value is greater than or equal to the above-mentioned fifth threshold value, sequentially executing the above-mentioned updating step, the above-mentioned correction step, the above-mentioned calculation step and the above-mentioned first determination step until the above-mentioned distance average value is less than the above-mentioned fifth threshold value or the above-mentioned update angle reaches the boundary of the above-mentioned third predetermined range.

[0075] The above method also includes: determining whether the number of the above-mentioned second target obstacles is greater than the above-mentioned fourth threshold value, and when it is determined that the number of the above-mentioned second target obstacles is greater than the above-mentioned fourth threshold value, setting the traversal polling flag Flag=0; the above-mentioned updating step includes: when the above-mentioned average value is greater than or equal to the above-mentioned fifth threshold value, starting from the above-mentioned yaw angle, increasing or decreasing the above-mentioned yaw angle in steps of the above-mentioned preset angle to obtain an updated angle, and setting the above-mentioned traversal polling flag Flag=1, and the above-mentioned updated angle is within the above-mentioned third predetermined range.

[0076] Of course, the value of the above-mentioned traversal polling flag is not limited to the above-mentioned 0 and 1. Technical personnel in this field can set it specifically according to actual conditions. For example, when it is determined that the number of the above-mentioned second target obstacles is greater than the above-mentioned fourth threshold, the traversal polling flag Flag is set to Flag = the first value to obtain the updated angle, and the above-mentioned traversal polling flag Flag is set to Flag = the second value, and the first value is different from the second value.

[0077] When the above-mentioned average value is less than the above-mentioned fifth threshold value, the above-mentioned yaw angle is determined to be the above-mentioned calibrated yaw angle, including: when the above-mentioned average value is less than the above-mentioned fifth threshold value, determining whether the above-mentioned traversal polling flag Flag is equal to 0; when it is determined that the above-mentioned traversal polling flag Flag=0, determining that the above-mentioned yaw angle is the above-mentioned calibrated yaw angle; when it is determined that the above-mentioned traversal polling flag Flag≠0, determining that the above-mentioned update angle is the above-mentioned calibrated yaw angle.

[0078] When the above-mentioned distance average value is less than the above-mentioned fifth threshold value, the above-mentioned updated angle is determined to be the above-mentioned calibrated yaw angle, including: when the above-mentioned average value is less than the above-mentioned fifth threshold value, determining whether the above-mentioned traversal polling flag Flag is equal to 0; when it is determined that the above-mentioned traversal polling flag Flag=0, determining that the above-mentioned yaw angle is the above-mentioned calibrated yaw angle; when it is determined that the above-mentioned traversal polling flag Flag≠0, determining that the above-mentioned update angle is the above-mentioned calibrated yaw angle.

[0079] The method further includes: when the update angle reaches a boundary of the third predetermined range, determining that the current yaw angle exceeds an allowable operating range.

[0080] It should be noted that the third predetermined range may be the same as the first predetermined range, and the third predetermined range may also be obtained by appropriately tightening the first predetermined range.

[0081] In the case where the above-mentioned predetermined parameters include the pitch angle, the predetermined parameters of the target sensor are calibrated to obtain the target parameters, including: extracting a second stationary obstacle from the continuous multiple frames of data collected by the above-mentioned target sensor, and obtaining a first height, the above-mentioned first height is the height information of the above-mentioned second stationary obstacle measured by the above-mentioned target sensor, and the above-mentioned second stationary obstacle is an obstacle located on the above-mentioned road or on both sides of the above-mentioned road, and the distance from the above-mentioned automatic driving vehicle is within a fourth predetermined range and does not move; obtaining a fitting relationship between the second height and the fifth distance to obtain a second fitting line equation, the above-mentioned second height is the height information of the above-mentioned second stationary obstacle determined according to the high-precision map, and the above-mentioned fifth distance is the real-time distance between the above-mentioned second stationary obstacle and the above-mentioned automatic driving vehicle in the above-mentioned driving direction; in the case where the number of the above-mentioned second stationary obstacles is greater than the seventh threshold, fitting the above-mentioned first height and the above-mentioned fifth distance to obtain a third fitting line equation; determining that the angle between the above-mentioned second fitting line equation and the above-mentioned third fitting line equation is the final pitch angle.

[0082] In the above embodiment, during vehicle driving, the pitch angle of the target sensor is calibrated by utilizing the height information of the second stationary obstacle with fixed height, such as the gantry and road signs on the road ahead, and the real-time distance between the second stationary obstacle and the autonomous driving vehicle in the driving direction, combined with the second fitting line equation fed back by the high-precision map, thereby realizing dynamic calibration of the pitch angle of the sensor and ensuring the accuracy of the calibrated sensor.

[0083] In order to ensure the validity and reliability of the height information, it is necessary to select a suitable distance interval as the fourth predetermined range.

[0084] Specifically, the specific process of fitting the first height and the fifth distance to obtain the third fitting line equation can be: taking the first height as the dependent variable and the fifth distance as the independent variable, performing least squares fitting on the first height and the fifth distance to obtain the third fitting line equation.

[0085] In actual application, Figure 4 As shown, during the driving process of the autonomous driving vehicle, the vehicle's motion trajectory, motion posture and the road slope information ahead provided by the high-precision map, if the road slope is stable (fluctuates within a very small range), then when the autonomous driving vehicle approaches obstacles with fixed heights such as gantry frames and road signs, the straight line fitted by its second height Z and the fifth distance X at ZX should be parallel to the X-axis.

[0086] Therefore, in order to simplify the calibration process of the pitch angle of the target sensor, the predetermined parameters of the target sensor are calibrated to obtain the target parameters, which may also include: extracting the second stationary obstacle from the continuous multi-frame data collected by the target sensor, and obtaining the first height, the first height is the height information of the second stationary obstacle measured by the target sensor, and the second stationary obstacle is an obstacle located on the road or on both sides of the road, and the distance from the autonomous driving vehicle is within the fourth predetermined range and does not move; according to the second height, obtaining the second fitting line equation of y=second height, wherein the second height is the height information of the second stationary obstacle determined according to the high-precision map; when the number of the second stationary obstacles is greater than the seventh threshold, fitting the first height and the fifth distance to obtain the third fitting line equation; determining the angle between the second fitting line equation and the third fitting line equation as the final pitch angle. The above process omits the process of fitting the second height and the fifth distance obtained from the high-precision map, further ensuring that the calibration process of the above pitch angle is relatively simple and easy to implement.

[0087] It should be noted that one frame of data from the target sensor can only obtain one first height, so the first fitting line equation is obtained by fitting multiple frames of data from the target sensor. In order to further ensure the calibration stability of the pitch angle, the method further includes: after obtaining the final pitch angle, dynamically updating the final pitch angle according to the frame data obtained in real time by the target sensor.

[0088] It should be noted that the above-mentioned position information of the obstacle is the position information of the obstacle in the vehicle body coordinate system.

[0089] Step S102, when the target parameter is within the first predetermined range, correcting the position information of the obstacle measured by the target sensor according to the target parameter;

[0090] Specifically, before correcting the position information of the obstacle measured by the target sensor according to the target parameter, the method further includes: step S401, converting the initial position information (x, y, z) of the obstacle detected by the target sensor into a vehicle body coordinate system;

[0091] The specific process is as follows: According to the current installation angle of the target sensor Calculate the quaternion transformation matrix Rotation_Matrix; according to the quaternion transformation matrix, the installation position Pos_Tra = [x0, y0, z0] and the position information, obtain the specific position (X, Y, Z) of the obstacle in the vehicle coordinate system:

[0092] (X,Y,Z)=(x,y,z)×Rotation_Matrix+Pos_Tra

[0093] The above-mentioned transformation and translation of the obstacle position can be implemented inside the sensor of the autonomous driving vehicle, or in a multi-sensor fusion algorithm module, both of which are feasible. This application takes the implementation in a multi-sensor fusion algorithm module as an example.

[0094] Step S402, according to the time when the target sensor obtains each frame of data, the speed and acceleration of the autonomous driving vehicle, the position adjustment amount (△x, △y, △z) is determined in real time, and the specific position (X, Y, Z) of the obstacle in the vehicle coordinate system is adjusted in real time to obtain a new position (X0, Y0, Z0) = (X, Y, Z) + (△x, △y, △z), which is the position information of the obstacle measured by the above-mentioned target sensor.

[0095] According to the above-mentioned target parameters, the specific process of correcting the position information of the obstacle measured by the above-mentioned target sensor can be: according to the target parameters, the quaternion transformation matrix Rotation_Matrix is ​​recalculated, and then the specific position of the obstacle in the vehicle coordinate system is obtained according to the recalculated quaternion transformation matrix Rotation_Matrix, the installation position Pos_Tra and the position information, and then the position adjustment amount (△x, △y, △z) is used to adjust the position in real time to obtain the corrected position information.

[0096] Step S103: When the target parameter is not within the first predetermined range, lowering the autonomous driving level of the autonomous driving vehicle.

[0097] The control method of the autonomous driving vehicle of the present application, first, during the driving process of the autonomous driving vehicle, when the autonomous driving vehicle satisfies the change rate of the yaw angle of the autonomous driving vehicle is less than the first threshold and the change rate of the predetermined acceleration is less than the second threshold, the yaw angle of the sensor is calibrated, and / or when the autonomous driving vehicle satisfies the change rate of the slope of the road where the autonomous driving vehicle is located is less than the third threshold, the pitch angle of the sensor is calibrated to obtain the target parameter, thereby ensuring that the attitude information of the calibrated sensor is relatively accurate; then, when the target parameter is within the first predetermined range, the position information of the obstacle measured by the sensor is corrected according to the target parameter, thereby ensuring that the sensor accurately locates the obstacles around the vehicle, thereby ensuring the driving safety of the autonomous driving vehicle; when the target parameter is not within the first predetermined range, it means that the positioning of the sensor is inaccurate and the risk of autonomous driving is high. At this time, the autonomous driving level of the autonomous driving vehicle is reduced to avoid driving accidents caused by inaccurate positioning of the sensor. The present application effectively solves the problem of inaccurate sensor positioning of autonomous driving vehicles in the prior art as the driving time increases.

[0098] The above method can improve the precise positioning of surrounding obstacles by calibrating the yaw angle of the sensor; by calibrating the pitch angle of the sensor, the reliability of obstacle height information obtained by the autonomous vehicle can be greatly improved, which is conducive to the autonomous vehicle to determine whether the vehicle can safely pass certain height-restricted areas to ensure the safety of autonomous driving. It can also greatly reduce the dependence of sensor calibration on high-precision, high-cost calibration equipment and reduce the cost of static calibration equipment; at the same time, the above method can replace the dynamic calibration process of the vehicle offline, saving the cost of building a dynamic calibration site.

[0099] The above-mentioned reduction in the autonomous driving level of the above-mentioned autonomous driving vehicle includes controlling the autonomous driving vehicle to exit autonomous driving and notifying relevant personnel to take over control of the vehicle.

[0100] When the target parameter is not within the first predetermined range, the method further includes: issuing an alarm message. This can promptly inform the driver and / or vehicle manager that the predetermined parameter of the target sensor needs to be physically and mechanically adjusted, which is beneficial for the autonomous driving vehicle to implement functional safety modules such as autonomous driving downgrade or even exit, and ensure the safe driving of the vehicle.

[0101] In actual application, when the target sensor collects obstacle information, in addition to obtaining obstacle information on the road and on both sides of the road, it may also collect obstacle information on one side of an adjacent road. In this case, in order to further avoid the obstacle information on the road where the vehicle is not located from interfering with the subsequent process and affecting the subsequent calibration effect, according to another specific embodiment of the present application, before calibrating the predetermined parameters of the target sensor and obtaining the target parameters, the method further includes: obtaining a sixth threshold, the speed of the autonomous driving vehicle, and the speeds of the obstacles on both sides of the road; according to the speed and the obstacles located at each of the speeds, determining that the speed is opposite to the direction of the speed and the predetermined difference is located in a fifth predetermined range as the initial stationary obstacle, and the predetermined difference is the absolute value of the difference between the absolute value of the speed and the absolute value of the vehicle speed; according to the position information of the initial stationary obstacle and the boundary line equation, calculating the distance between each of the initial stationary obstacles and the boundary line equation to obtain multiple fourth distances; determining that the initial stationary obstacle corresponding to the fourth distance less than the sixth threshold is the first stationary obstacle. In the above embodiment, by setting a suitable fifth threshold, the first target obstacle is selected.

[0102] It should be noted that the steps shown in the flowcharts of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and that, although a logical order is shown in the flowcharts, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0103] The embodiment of the present application also provides a control device for an autonomous driving vehicle. It should be noted that the control device for an autonomous driving vehicle in the embodiment of the present application can be used to execute the control method for an autonomous driving vehicle provided in the embodiment of the present application. The control device for an autonomous driving vehicle provided in the embodiment of the present application is introduced below.

[0104] Figure 5 Schematic diagram of a control device for an autonomous driving vehicle according to an embodiment of the present application. Figure 5 As shown, the device comprises:

[0105] The calibration unit 10 is used for calibrating predetermined parameters of a target sensor to obtain target parameters during the driving of the autonomous driving vehicle, when the autonomous driving vehicle satisfies predetermined conditions, wherein the predetermined parameters include at least one of a yaw angle and a pitch angle. When the predetermined parameters include the yaw angle, the predetermined conditions include: a rate of change of the yaw angle of the autonomous driving vehicle is less than a first threshold value, and a rate of change of a predetermined acceleration is less than a second threshold value, and the predetermined acceleration is an acceleration of the autonomous driving vehicle perpendicular to the driving direction. When the predetermined parameters include the pitch angle, the predetermined conditions include: a rate of change of the slope of the road where the autonomous driving vehicle is located is less than a third threshold value.

[0106] In the above process, when the rate of change of the yaw angle of the above-mentioned autonomous driving vehicle is less than the first threshold value, and the rate of change of the predetermined acceleration is less than the second threshold value, it means that the autonomous driving vehicle is driving smoothly in a relatively horizontal direction. At this time, calibrating the yaw angle of the target sensor can make the calibrated yaw angle more accurate; when the rate of change of the slope of the road where the autonomous driving vehicle is located is less than the third threshold value, it means that the pitch angle of the autonomous driving vehicle changes little. At this time, calibrating the pitch angle of the target sensor can make the calibrated pitch angle more accurate.

[0107] The above-mentioned yaw angle of the autonomous driving vehicle, the above-mentioned predetermined acceleration and the above-mentioned slope are all obtained after filtering processing.

[0108] Depending on the location where the target sensor is installed on the autonomous driving vehicle, in some cases it can detect both sides of the road at the same time, while in some cases it can only detect one side of the road. This application applies to both situations.

[0109] In order to further ensure a better calibration effect of the yaw angle, according to another specific embodiment of the present application, when the above-mentioned predetermined parameters include the above-mentioned yaw angle, the above-mentioned predetermined conditions include: the change rate of the yaw angle of the above-mentioned autonomous driving vehicle is less than a first threshold, the change rate of the predetermined acceleration is less than a second threshold, and the change rate of the slope of the above-mentioned road is less than a third threshold, which can further avoid the interference of the slope on the yaw angle calibration result.

[0110] In addition, when the above-mentioned predetermined parameters include at least one of the yaw angle and the pitch angle, the above-mentioned predetermined conditions also include: the vehicle speed of the above-mentioned automatic driving vehicle is greater than the eighth threshold value.

[0111] The above-mentioned target sensors may include laser radar (Lidar) and / or millimeter wave radar (Radar). Among them, laser radar is used to collect point cloud information or object information of obstacles around the vehicle; millimeter wave radar has height measurement capability, and is also used to collect point cloud information or object information of obstacles around the vehicle. The dynamic calibration principles of laser radar and laser radar are similar, and the dynamic calibration algorithm data processing of point cloud information and object information is also similar. The difference is that the amount of point cloud data is large, the cache space occupied is also larger, the algorithm running time is slightly longer, and the accuracy of the calibration result will be slightly better, but not much different. Technical personnel in this field can choose the type of sensor to be used according to their own hardware platform resources and sensor data information type.

[0112] The installation positions of LiDAR and LiDAR are different, but their obstacle position information will be unified into the vehicle coordinate system, and the final dynamic calibration principle is similar, so the installation position is not affected.

[0113] In actual application, during the control process of the autonomous vehicle, various sensors will be used to collect information about the vehicle such as vehicle speed, yaw angle, lateral acceleration and longitudinal acceleration in real time. In addition, high-precision maps will be used to collect information about road slopes at different distances ahead, road boundaries, lane lines, fixed-height obstacles such as gantries and road signs, and obstacles such as railings, flower beds and walls on both sides of the road. This information is processed by a multi-sensor fusion algorithm and used as reference data for controlling the driving of the autonomous vehicle.

[0114] The data obtained based on the high-precision map described below in this application are all data of the high-precision map in a valid and usable state.

[0115] In one embodiment of the present application, when the above-mentioned predetermined parameters include the above-mentioned yaw angle, the above-mentioned calibration unit includes: a first calibration module, used to calibrate the above-mentioned yaw angle according to a frame of data collected by the above-mentioned target sensor and the boundary line equation to obtain a calibrated yaw angle, and the above-mentioned boundary line equation is an equation of the boundary line of the road where the above-mentioned autonomous driving vehicle is located determined according to a high-precision map; a first analysis module, used to perform statistical feature analysis on the above-mentioned calibrated yaw angle corresponding to the above-mentioned multiple frames of data to obtain a final yaw angle.

[0116] The above process determines the calibrated yaw angle corresponding to multiple consecutive frames through the data of each consecutive frame of the target sensor and the boundary line equation. After statistical analysis of the calibrated yaw angle corresponding to multiple consecutive frames, inaccurate values ​​in multiple calibrated yaw angles can be eliminated, so that the final yaw angle obtained is more accurate. In addition, the above process also realizes the dynamic calibration of the yaw angle of the target sensor during the driving process of the autonomous driving vehicle, further ensuring that the obstacle information obtained from the target sensor is relatively accurate, thereby further ensuring the driving safety of the autonomous driving vehicle.

[0117] The above-mentioned first calibration module specifically includes: a first determination submodule, used to determine the position information and type of multiple first stationary obstacles based on one frame of the above-mentioned data, and the above-mentioned first stationary obstacles are obstacles located on both sides of the above-mentioned road and do not move; a second determination submodule, used to determine the above-mentioned calibrated yaw angle according to the boundary line equation and the position information of the multiple first target obstacles when the above-mentioned boundary line equation is a straight line equation and the number of the first target obstacles is greater than the fourth threshold, wherein the above-mentioned first target obstacle is the above-mentioned first stationary obstacle of the same type; a third determination submodule, used to determine the above-mentioned calibrated yaw angle according to the position information of the multiple first target obstacles when the above-mentioned boundary line equation is not the above-mentioned straight line equation and the number of the above-mentioned first stationary obstacles of the same type is greater than the above-mentioned fourth threshold.

[0118] The above types may include railings, flower beds, and walls. In actual applications, the boundary line equation of the road boundary line may be as follows Figure 2 The straight line shown in the figure indicates that the autonomous vehicle is on a straight road. Figure 3 The curve shown here indicates that the autonomous vehicle is in a non-straight road scenario. The calibration process of the yaw angle is different for boundary line equations of different forms, as described below:

[0119] In the horizontal straight road scenario, the rate of change of the yaw angle of the above-mentioned autonomous driving vehicle is less than the first threshold, which means that the yaw angle fluctuates stably in a small range near 0, and the rate of change of the predetermined acceleration is less than the second threshold, which means that the predetermined acceleration fluctuates in a small range near 0.

[0120] In the horizontal non-straight road scenario, the rate of change of the yaw angle of the above-mentioned autonomous driving vehicle is less than the first threshold, which means that the yaw angle fluctuates stably in a small range near a non-zero value, and the rate of change of the preset acceleration is less than the second threshold, which means that the preset acceleration fluctuates in a small range near a non-zero value.

[0121] The specific principle of the above embodiment is: during the driving process of the autonomous driving vehicle, the target sensor is used to detect the position information of the surrounding absolutely stationary obstacles such as railings, flower beds, walls, etc., combined with the vehicle's motion trajectory, posture and road slope information. If the yaw angle of the target sensor is normal, then the position distribution of the absolutely stationary obstacles and the road boundary line of the high-precision map are parallel or can overlap after translation. For details, please refer to Figure 2 and Figure 3 If the yaw angle of the target sensor is abnormal, then within the allowable angle range, the position of the absolutely stationary obstacle can be rotated by a certain angle α, parallelized or translated so that it coincides with the road boundary line. The angle α is the yaw angle of the target sensor after calibration.

[0122] The size of the fourth threshold value depends on the type of the first target obstacle. In the above frame of data acquired by the target sensor, the fourth threshold value corresponding to the case where the first target obstacle is a railing, the fourth threshold value corresponding to the case where the first target obstacle is a wall, and the fourth threshold value corresponding to the case where the first target obstacle is a flower bed increase in sequence, which can further ensure that the subsequent yaw angle calibration results are more accurate and reliable.

[0123] It should be noted that, in the above-mentioned second determination submodule and the above-mentioned third determination submodule, as long as the number of one type of first target obstacles is greater than the above-mentioned fourth threshold, the calibrated yaw angle can be determined according to the position information of the first target obstacles of this type. For example, in a frame of data, there are multiple types of first target obstacles whose number is greater than the above-mentioned fourth threshold. In order to simplify the calculation process, those skilled in the art can select one type of first target obstacles greater than the first threshold to calibrate the yaw angle, and can also calibrate the yaw angle according to each type of first target obstacle, so as to further ensure good calibration accuracy.

[0124] In addition, when the number of first target obstacles of all types does not satisfy the fourth threshold, the yaw angle calibration is terminated.

[0125] In order to further ensure the stability of the final yaw angle, the above-mentioned first analysis module may specifically include: a first calculation submodule, used to calculate the average value of the above-mentioned calibrated yaw angle corresponding to the above-mentioned data for multiple consecutive frames to obtain the above-mentioned final yaw angle. The above-mentioned first analysis module may also include: a second calculation submodule, used to calculate the mode of the above-mentioned calibrated yaw angle corresponding to the above-mentioned data for multiple consecutive frames to obtain the above-mentioned final yaw angle. The above-mentioned first analysis module may also include: a third calculation submodule, used to calculate the median of the above-mentioned calibrated yaw angle corresponding to the above-mentioned data for multiple consecutive frames to obtain the above-mentioned final yaw angle.

[0126] In other embodiments, the types and quantities of the plurality of first stationary obstacles are not limited to being obtained from the frame data collected by the target sensor, but may also be extracted from a high-precision map. That is, the calibration unit may further include: a second calibration module, for calibrating the yaw angle according to a frame of data collected by the target sensor and a high-precision map to obtain a calibrated yaw angle; and a second analysis module, for performing statistical feature analysis on the calibrated yaw angle corresponding to a plurality of consecutive frames of the data to obtain a final yaw angle.

[0127] Then, the second calibration module includes: a fourth determination submodule, which is used to determine the types of multiple first stationary obstacles and the boundary line equation of the road boundary line according to the high-precision map, wherein the first stationary obstacles are obstacles located on both sides of the road and do not move; a fifth determination submodule, which is used to determine the position information of multiple first stationary obstacles according to one frame of the data; a sixth determination submodule, which is used to determine the first stationary obstacles of the same type as the first target obstacles, and calculate the number of the first target obstacles corresponding to each of the above types; a seventh determination submodule, which is used to determine the calibrated yaw angle according to the boundary line equation and the position information of the multiple first target obstacles when the boundary line equation is a straight line equation and the number of the first target obstacles is greater than the fourth threshold, wherein the first target obstacle is the first stationary obstacle of the same type; an eighth determination submodule, which is used to determine the calibrated yaw angle according to the position information of the multiple first target obstacles when the boundary line equation is not the straight line equation and the number of the first stationary obstacles of the same type is greater than the fourth threshold.

[0128] In one embodiment of the present application, the second determination submodule or the seventh determination submodule is also used to perform least squares fitting on the position information of the first target obstacle located on the same side of the road to obtain a first fitting line equation; and determine the calibrated yaw angle based on the first fitting line equation and the boundary line equation.

[0129] In a straight road scenario, the position information of the first target obstacle on the same side of the road is fitted by least squares, so that the straight line equation of the first target obstacle, that is, the first fitting line equation mentioned above, can be obtained relatively simply and quickly. Then, the angle between the straight line equation and the boundary line equation of the straight line is calculated to obtain the calibrated yaw angle.

[0130] In actual applications, during the driving process of an autonomous vehicle, due to factors such as obstacles on the road and the position of the autonomous vehicle on the road, the target sensor can only collect obstacle data on one side of the road. In this case, a fitting line equation and the boundary line equation of the corresponding side can be directly calculated. In other cases, the autonomous vehicle may be able to collect obstacle data on both sides of the road, but the obstacle data on one side is not clear enough or incomplete. It may also be possible to collect obstacle data on both sides of the road relatively clearly and completely. In this case, the specific execution method can be:

[0131] The road includes a relative first boundary line and a second boundary line, the boundary line equation includes a first boundary line sub-equation and a second boundary line sub-equation, and in the case where the plurality of first target obstacles are located on different sides of the road, the first fitting line equation includes a first fitting line sub-equation and a second fitting line sub-equation, the first boundary line sub-equation and the first fitting line sub-equation correspond to the first boundary line respectively, the second boundary line sub-equation and the second fitting line sub-equation correspond to the second boundary line respectively, and the second determination submodule or the seventh determination submodule is further used to: in the case where the first target obstacle is located on different sides of the road, determine that the angle between the first boundary line sub-equation and the first fitting line sub-equation is the first angle, and the angle between the second boundary line sub-equation and the second fitting line sub-equation is the second angle; in the case where one of the first angle and the second angle is an invalid angle, determine the first The effective angle between the first angle and the second angle is the calibrated yaw angle, the number of the first target obstacles corresponding to the invalid angle is less than or equal to the fourth threshold, and the number of the first target obstacles corresponding to the effective angle is greater than the fourth threshold; when the first angle and the second angle are both the effective angles and the difference between the first angle and the second angle is within the second predetermined range, the average value of the first angle and the second angle is determined to be the calibrated yaw angle; when the first angle and the second angle are both the effective angles and the difference between the first angle and the second angle is not within the second predetermined range, the boundary line distance between the autonomous driving vehicle and the first boundary line and the boundary line distance between the autonomous driving vehicle and the second boundary line are respectively obtained, and the angle corresponding to the minimum value of the boundary line distance between the first angle and the second angle is determined to be the calibrated yaw angle.

[0132] In the above processing method, when only one of the first angle and the second angle is valid, the valid angle is used as the calibrated yaw angle; when both are valid angles and the difference is not large, the average value of the two is taken as the calibrated yaw angle; when both are valid angles and the difference is large, the angle corresponding to the road boundary line that the autonomous driving vehicle is approaching is taken as the calibrated yaw angle. The above processing method can further achieve a more accurate calibration of the yaw angle.

[0133] In addition, when the first angle and the second angle are both the invalid angles, the yaw angle calibration is completed.

[0134] In the embodiment of the present application, the third determination submodule and the eighth determination submodule are also used to: determine the distance between the position information of the second target obstacle and the boundary line equation as the third distance, and calculate the average value of each of the third distances, the second target obstacle is all the first target obstacles located on the same side of the road; determine whether the average value is less than the fifth threshold; when the average value is greater than or equal to the fifth threshold, determine the calibrated yaw angle according to the yaw angle, the third predetermined range and the preset angle step; when the average value is less than the fifth threshold, determine the yaw angle as the calibrated yaw angle. When the average value is less than the fifth threshold, it means that the current yaw angle of the target sensor is normal and no further calibration is required; when the average value is greater than or equal to the fifth threshold, calibrate the yaw angle according to the current yaw angle, the preset third predetermined range and the preset angle step, and further realize the accurate calibration of the yaw angle in the non-straight road scenario.

[0135] According to another specific embodiment of the present application, the third determination submodule and the eighth determination submodule are also respectively used for: an updating step, when the average value is greater than or equal to the fifth threshold value, starting from the yaw angle, increasing or decreasing the yaw angle by the preset angle step to obtain an updated angle, and the updated angle is located in the third predetermined range; a correction step, according to the updated angle, correcting the position information of each of the second target obstacles to obtain the corrected position information; a calculation step, calculating the average distance between the corrected position information and the boundary line equation; a first determination step, determining whether the average distance value is less than the fifth threshold value; a second determination step, when the average distance value is less than the fifth threshold value, determining that the updated angle is the calibrated yaw angle, and when the average distance value is greater than or equal to the fifth threshold value, sequentially executing the updating step, the correction step, the calculation step and the first determination step until the average distance value is less than the fifth threshold value or the updated angle reaches the boundary of the third predetermined range.

[0136] The above-mentioned device also includes: a first determination unit, used to determine whether the number of the above-mentioned second target obstacles is greater than the above-mentioned fourth threshold value, and when it is determined that the number of the above-mentioned second target obstacles is greater than the above-mentioned fourth threshold value, setting the traversal polling flag Flag=0; the above-mentioned updating step includes: when the above-mentioned average value is greater than or equal to the above-mentioned fifth threshold value, starting from the above-mentioned yaw angle, increasing or decreasing the above-mentioned yaw angle in steps of the above-mentioned preset angle to obtain an updated angle, and setting the above-mentioned traversal polling flag Flag=1, and the above-mentioned updated angle is within the above-mentioned third predetermined range.

[0137] Of course, the value of the above-mentioned traversal polling flag is not limited to the above-mentioned 0 and 1. Technical personnel in this field can set it specifically according to actual conditions. For example, when it is determined that the number of the above-mentioned second target obstacles is greater than the above-mentioned fourth threshold, the traversal polling flag Flag is set to Flag = the first value to obtain the updated angle, and the above-mentioned traversal polling flag Flag is set to Flag = the second value, and the first value is different from the second value.

[0138] When the above-mentioned average value is less than the above-mentioned fifth threshold value, the above-mentioned third determination submodule and the above-mentioned eighth determination submodule are also respectively used for: when the above-mentioned average value is less than the above-mentioned fifth threshold value, determining whether the above-mentioned traversal polling flag Flag is equal to 0; when it is determined that the above-mentioned traversal polling flag Flag=0, determining that the above-mentioned yaw angle is the above-mentioned calibrated yaw angle; when it is determined that the above-mentioned traversal polling flag Flag≠0, determining that the above-mentioned update angle is the above-mentioned calibrated yaw angle.

[0139] The third determination submodule and the eighth determination submodule are also respectively used for: when the average value is less than the fifth threshold value, determining whether the traversal polling flag Flag is equal to 0; when it is determined that the traversal polling flag Flag=0, determining that the yaw angle is the calibrated yaw angle; when it is determined that the traversal polling flag Flag≠0, determining that the update angle is the calibrated yaw angle.

[0140] The device further includes: a second determining unit, configured to determine that the current yaw angle exceeds an allowable operating range when the update angle reaches a boundary of the third predetermined range.

[0141] It should be noted that the third predetermined range may be the same as the first predetermined range, and the third predetermined range may also be obtained by appropriately tightening the first predetermined range.

[0142] In the case where the above-mentioned predetermined parameter includes a pitch angle, the above-mentioned calibration unit includes: a first extraction module, used to extract the second stationary obstacle from the continuous multiple frames of data collected by the above-mentioned target sensor, and obtain a first height, the above-mentioned first height is the height information of the above-mentioned second stationary obstacle measured by the above-mentioned target sensor, and the above-mentioned second stationary obstacle is an obstacle located on the above-mentioned road or on both sides of the above-mentioned road, and the distance from the above-mentioned automatic driving vehicle is within a fourth predetermined range and does not move; an acquisition module, used to obtain a fitting relationship between the second height and the fifth distance, and obtain a second fitting line equation, the above-mentioned second height is the height information of the above-mentioned second stationary obstacle determined according to the high-precision map, and the above-mentioned fifth distance is the real-time distance between the above-mentioned second stationary obstacle and the above-mentioned automatic driving vehicle in the above-mentioned driving direction; a first fitting module, used to fit the above-mentioned first height and the above-mentioned fifth distance when the number of the above-mentioned second stationary obstacles is greater than the seventh threshold value, and obtain a third fitting line equation; a first determination module, used to determine that the angle between the above-mentioned second fitting line equation and the above-mentioned third fitting line equation is the final pitch angle.

[0143] In the above embodiment, during vehicle driving, the pitch angle of the target sensor is calibrated by utilizing the height information of the second stationary obstacle with fixed height, such as the gantry and road signs on the road ahead, and the real-time distance between the second stationary obstacle and the autonomous driving vehicle in the driving direction, combined with the second fitting line equation fed back by the high-precision map, thereby realizing dynamic calibration of the pitch angle of the sensor and ensuring the accuracy of the calibrated sensor.

[0144] In order to ensure the validity and reliability of the height information, it is necessary to select a suitable distance interval as the fourth predetermined range.

[0145] Specifically, the specific process of fitting the first height and the fifth distance to obtain the third fitting line equation can be: taking the first height as the dependent variable and the fifth distance as the independent variable, performing least squares fitting on the first height and the fifth distance to obtain the third fitting line equation.

[0146] In actual application, Figure 4 As shown, during the driving process of the autonomous driving vehicle, the vehicle's motion trajectory, motion posture and the road slope information ahead provided by the high-precision map, if the road slope is stable (fluctuates within a very small range), then when the autonomous driving vehicle approaches obstacles with fixed heights such as gantry frames and road signs, the straight line fitted by its second height Z and the fifth distance X at ZX should be parallel to the X-axis.

[0147] Therefore, in order to simplify the calibration process of the pitch angle of the target sensor, the calibration unit may further include: a second extraction module, for extracting the second stationary obstacle from the continuous multi-frame data collected by the target sensor, and obtaining the first height, wherein the first height is the height information of the second stationary obstacle measured by the target sensor, and the second stationary obstacle is an obstacle located on the road or on both sides of the road, and the distance from the autonomous driving vehicle is within the fourth predetermined range and does not move; a second fitting module, for obtaining a second fitting line equation of y=second height according to the second height, wherein the second height is the height information of the second stationary obstacle determined according to the high-precision map; a third fitting module, for fitting the first height and the fifth distance to obtain a third fitting line equation when the number of the second stationary obstacles is greater than the seventh threshold; and a second determination module, for determining the angle between the second fitting line equation and the third fitting line equation as the final pitch angle. The above process omits the process of fitting the second height and the fifth distance obtained from the high-precision map, further ensuring that the calibration process of the above pitch angle is relatively simple and easy to implement.

[0148] It should be noted that one frame of data from the target sensor can only obtain one first height, so the first fitting line equation is obtained by fitting multiple frames of data from the target sensor. In order to further ensure the calibration stability of the pitch angle, the device further includes: after obtaining the final pitch angle, dynamically updating the final pitch angle according to the frame data obtained in real time by the target sensor.

[0149] It should be noted that the above-mentioned position information of the obstacle is the position information of the obstacle in the vehicle body coordinate system.

[0150] A correction unit 20, configured to correct the position information of the obstacle measured by the target sensor according to the target parameter when the target parameter is within the first predetermined range;

[0151] Specifically, the above-mentioned device also includes: a conversion unit, which is used to convert the initial position information (x, y, z) of the obstacle detected by the target sensor into the vehicle body coordinate system before correcting the position information of the obstacle measured by the above-mentioned target sensor according to the above-mentioned target parameters; a third determination unit, which is used to determine the position adjustment amount (△x, △y, △z) in real time according to the time when the target sensor obtains each frame of data, the speed and acceleration of the automatic driving vehicle, and adjust the specific position (X, Y, Z) of the obstacle in the vehicle body coordinate system in real time to obtain a new position (X0, Y0, Z0) = (X, Y, Z) + (△x, △y, △z), which is the position information of the obstacle measured by the above-mentioned target sensor.

[0152] The conversion unit specifically includes: a first calculation module for calculating the current installation angle of the target sensor according to the first calculation module. Calculate the quaternion transformation matrix Rotation_Matrix; get the module, which is used to get the specific position (X, Y, Z) of the obstacle in the vehicle coordinate system according to the quaternion transformation matrix, the installation position Pos_Tra = [x0, y0, z0] and the position information:

[0153] (X,Y,Z)=(x,y,z)×Rotation_Matrix+Pos_Tra

[0154] The above-mentioned transformation and translation of the obstacle position can be implemented inside the sensor of the autonomous driving vehicle, or in a multi-sensor fusion algorithm module, both of which are feasible. This application takes the implementation in a multi-sensor fusion algorithm module as an example.

[0155] The above-mentioned correction unit includes: a second calculation module, which is used to recalculate the quaternion transformation matrix Rotation_Matrix according to the target parameters, and then obtain the specific position of the obstacle in the vehicle coordinate system according to the recalculated quaternion transformation matrix Rotation_Matrix, the installation position Pos_Tra and the position information, and then use the position adjustment amount (△x, △y, △z) to adjust the position in real time to obtain the corrected position information.

[0156] The reducing unit 30 is used to reduce the autonomous driving level of the autonomous driving vehicle when the target parameter is not within the first predetermined range.

[0157] The control device of the autonomous driving vehicle of the present application, through the calibration unit, during the driving process of the autonomous driving vehicle, when the autonomous driving vehicle satisfies the change rate of the yaw angle of the autonomous driving vehicle is less than the first threshold and the change rate of the predetermined acceleration is less than the second threshold, the yaw angle of the sensor is calibrated, and / or when the autonomous driving vehicle satisfies the change rate of the slope of the road where the autonomous driving vehicle is located is less than the third threshold, the pitch angle of the sensor is calibrated to obtain the target parameter, thereby ensuring that the attitude information of the calibrated sensor is relatively accurate; through the correction unit, when the target parameter is within the first predetermined range, according to the target parameter, the position information of the obstacle measured by the sensor is corrected, thereby ensuring that the sensor accurately locates the obstacles around the vehicle, thereby ensuring the driving safety of the autonomous driving vehicle; when the target parameter is not within the first predetermined range, it means that the positioning of the sensor is inaccurate and the risk of autonomous driving is high. The autonomous driving level of the autonomous driving vehicle is reduced by the reduction unit, thereby avoiding driving accidents caused by inaccurate positioning of the sensor. The present application effectively solves the problem of inaccurate sensor positioning of autonomous driving vehicles in the prior art as the driving time increases.

[0158] The above device can improve the precise positioning of surrounding obstacles by calibrating the yaw angle of the sensor; by calibrating the pitch angle of the sensor, the reliability of obstacle height information obtained by the autonomous vehicle can be greatly improved, which is conducive to the autonomous vehicle to determine whether the vehicle can safely pass certain height-restricted areas to ensure the safety of autonomous driving. It can also greatly reduce the dependence of sensor calibration on high-precision, high-cost calibration equipment and reduce the cost of static calibration equipment; at the same time, the above device can replace the dynamic calibration process of the vehicle offline, saving the cost of building a dynamic calibration site.

[0159] The above-mentioned reduction in the autonomous driving level of the above-mentioned autonomous driving vehicle includes controlling the autonomous driving vehicle to exit autonomous driving and notifying relevant personnel to take over control of the vehicle.

[0160] The device further includes: an issuing unit, which is used to issue an alarm message when the target parameter is not within the first predetermined range. This can promptly inform the driver and / or vehicle manager that the predetermined parameter of the target sensor needs to be physically and mechanically adjusted, which is conducive to the automatic driving vehicle to achieve automatic driving downgrade or even exit and other functional safety modules, and ensure the safe driving of the vehicle.

[0161] In actual application, when the target sensor collects obstacle information, in addition to obtaining obstacle information on the road and on both sides of the road, it may also collect obstacle information on one side of the adjacent road. In this case, in order to further avoid the obstacle information on the road where the vehicle is not located from interfering with the subsequent process and affecting the subsequent calibration effect, according to another specific embodiment of the present application, the above-mentioned device also includes: an acquisition unit, which is used to calibrate the predetermined parameters of the target sensor and obtain the sixth threshold value, the speed of the above-mentioned autonomous driving vehicle, and the speed of each of the above-mentioned obstacles on both sides of the above-mentioned road before obtaining the target parameters; A fourth determination unit is used to determine, based on the vehicle speed and each of the above speeds, that the above speed is opposite to the vehicle speed and the above obstacle whose predetermined difference is within a fifth predetermined range is the initial stationary obstacle, wherein the above predetermined difference is the absolute value of the difference between the absolute value of the above speed and the absolute value of the above vehicle speed; a calculation unit is used to calculate the distance between each of the above initial stationary obstacles and the above boundary line equation according to the position information of the above initial stationary obstacle and the above boundary line equation to obtain a plurality of fourth distances; a fifth determination unit is used to determine that the above initial stationary obstacle corresponding to the above fourth distance less than the above sixth threshold is the above first stationary obstacle. In the above embodiment, by setting an appropriate fifth threshold, the first target obstacle is selected from the first target obstacle.

[0162] The control device of the above-mentioned autonomous driving vehicle includes a processor and a memory. The above-mentioned calibration unit, the above-mentioned correction unit, the above-mentioned reduction unit, etc. are all stored in the memory as program units, and the processor executes the above-mentioned program units stored in the memory to realize corresponding functions.

[0163] The processor includes a kernel, which retrieves the corresponding program unit from the memory. One or more kernels can be set, and the problem of inaccurate sensor positioning of autonomous vehicles as the driving time increases in the prior art can be solved by adjusting kernel parameters.

[0164] The memory may include non-permanent memory in a computer-readable medium, random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM, and the memory includes at least one memory chip.

[0165] An embodiment of the present application provides a computer-readable storage medium having a program stored thereon, which, when executed by a processor, implements the above-mentioned control method of the autonomous driving vehicle.

[0166] An embodiment of the present application provides a processor, which is used to run a program, wherein the control method of the autonomous driving vehicle is executed when the program is run.

[0167] An embodiment of the present application provides a device, the device including a processor, a memory, and a program stored in the memory and executable on the processor, and when the processor executes the program, at least the following steps are implemented:

[0168] Step S101: During the driving of the autonomous driving vehicle, if the autonomous driving vehicle meets predetermined conditions, calibrate predetermined parameters of the target sensor to obtain target parameters, wherein the predetermined parameters include at least one of a yaw angle and a pitch angle. If the predetermined parameters include the yaw angle, the predetermined conditions include: a rate of change of the yaw angle of the autonomous driving vehicle is less than a first threshold, and a rate of change of a predetermined acceleration is less than a second threshold, and the predetermined acceleration is an acceleration of the autonomous driving vehicle perpendicular to the driving direction. If the predetermined parameters include the pitch angle, the predetermined conditions include: a rate of change of the slope of the road where the autonomous driving vehicle is located is less than a third threshold.

[0169] Step S102, when the target parameter is within the first predetermined range, correcting the position information of the obstacle measured by the target sensor according to the target parameter;

[0170] Step S103: When the target parameter is not within the first predetermined range, lowering the autonomous driving level of the autonomous driving vehicle.

[0171] The devices in this article can be servers, PCs, PADs, mobile phones, etc.

[0172] The present application also provides a computer program product, which, when executed on a data processing device, is suitable for executing a program for initializing at least the following method steps:

[0173] Step S101: During the driving of the autonomous driving vehicle, if the autonomous driving vehicle meets predetermined conditions, calibrate predetermined parameters of the target sensor to obtain target parameters, wherein the predetermined parameters include at least one of a yaw angle and a pitch angle. If the predetermined parameters include the yaw angle, the predetermined conditions include: a rate of change of the yaw angle of the autonomous driving vehicle is less than a first threshold, and a rate of change of a predetermined acceleration is less than a second threshold, and the predetermined acceleration is an acceleration of the autonomous driving vehicle perpendicular to the driving direction. If the predetermined parameters include the pitch angle, the predetermined conditions include: a rate of change of the slope of the road where the autonomous driving vehicle is located is less than a third threshold.

[0174] Step S102, when the target parameter is within the first predetermined range, correcting the position information of the obstacle measured by the target sensor according to the target parameter;

[0175] Step S103: When the target parameter is not within the first predetermined range, lowering the autonomous driving level of the autonomous driving vehicle.

[0176] According to another aspect of an embodiment of the present application, an autonomous driving system is also provided, including an autonomous driving vehicle, one or more processors, a memory and one or more programs, wherein the one or more programs are stored in the memory and are configured to be executed by the one or more processors, and the one or more programs include methods for executing any one of the above methods.

[0177] In the automatic driving system of the present application, the processor is used to execute any control method, which first calibrates the yaw angle of the sensor when the automatic driving vehicle satisfies the change rate of the yaw angle of the automatic driving vehicle is less than the first threshold and the change rate of the predetermined acceleration is less than the second threshold during the driving process of the automatic driving vehicle, and / or calibrates the pitch angle of the sensor when the automatic driving vehicle satisfies the change rate of the slope of the road where the automatic driving vehicle is located is less than the third threshold, and obtains the target parameter, ensuring that the attitude information of the calibrated sensor is relatively accurate; then, when the target parameter is within the first predetermined range, the position information of the obstacle measured by the sensor is corrected according to the target parameter, ensuring that the sensor accurately locates the obstacles around the vehicle, thereby ensuring the driving safety of the automatic driving vehicle; when the target parameter is not within the first predetermined range, it means that the positioning of the sensor is inaccurate and the risk of automatic driving is high. At this time, the automatic driving level of the automatic driving vehicle is reduced to avoid driving accidents caused by inaccurate positioning of the sensor. The present application effectively solves the problem of inaccurate sensor positioning of automatic driving vehicles in the prior art as the driving time increases.

[0178] In addition, the processor of the above-mentioned autonomous driving system can improve the accurate positioning of the surrounding obstacles of the autonomous driving vehicle by calibrating the yaw angle of the sensor; by calibrating the pitch angle of the sensor, the reliability of the obstacle height information obtained by the autonomous driving vehicle can be greatly improved, which is conducive to the autonomous driving vehicle to determine whether the vehicle can safely pass certain height-restricted areas to ensure the safety of autonomous driving. It can also greatly reduce the dependence of sensor calibration on high-precision and high-cost calibration equipment and reduce the cost of static calibration equipment; at the same time, the above method can replace the dynamic calibration process of the vehicle offline, saving the cost of building a dynamic calibration site.

[0179] In order to enable those skilled in the art to more clearly understand the technical solution of the present application, it will be described below in conjunction with specific embodiments.

[0180] Example

[0181] The control method of the autonomous driving vehicle is applied to the multi-sensor fusion algorithm module, such as Figure 6 As shown, the multi-sensor fusion algorithm module includes a coordinate system conversion module, an obstacle preliminary screening module, a horizontal calibration module, a pitch calibration module and a decision processing module. The specific processing process is as follows:

[0182] Step 1: The coordinate system conversion module is used to convert the obstacle information (x, y, z) detected by the sensor to the vehicle coordinate system when the vehicle speed is greater than the set speed threshold during vehicle driving. According to its installation angle Rota_angle, the quaternion transformation matrix Rotation_Matrix is ​​calculated. Combined with its installation position Pos_Tra, the obstacle position (x, y, z) is converted and translated to its specific position (X, Y, Z) in the vehicle coordinate system:

[0183] (X,Y,Z)=(x,y,z)×Rotation_Matrix+Pos_Tra

[0184] The speed information, yaw_rate and lateral acceleration (i.e. the above-mentioned predetermined acceleration) of the autonomous driving vehicle are filtered to obtain stable speed information, yaw_rate and lateral acceleration. At the same time, the obstacle position information (X, Y, Z) is slightly adjusted in real time in combination with its own motion posture to obtain a new position (X0, Y0, Z0) = (X, Y, Z) + (△x, △y, △z); at the same time, the effective and available status of the high-precision map, road boundary lines, road edge types (flower beds, walls, metal railings, etc.), road slopes at different distances ahead, and gantry or road signs and other information are recorded;

[0185] Step 2: The obstacle preliminary screening module is used to preliminarily screen out the stationary obstacles among the above obstacles based on the speed information of the obstacles and the speed information of the autonomous driving vehicle to obtain the initial stationary obstacles;

[0186] Step 3: When the yaw angle yaw_rate and lateral acceleration of the autonomous driving vehicle are in a stable state (fluctuating within a very small range), and the slope information of the high-precision map is also in a stable state (fluctuating within a very small range), enter the horizontal calibration module;

[0187] Step 4: When the high-precision map feedback slope information at different longitudinal distances in front of the vehicle is in a stable state (fluctuating within a very small range), and at the same time the front feedback shows that there is an obstacle of fixed height such as a gantry or road sign in front, enter the pitch calibration module;

[0188] Step 5: The decision processing module determines whether the final yaw angle yaw_angle and the final pitch angle pitch_angle of the dynamically calibrated sensor are within the allowable working range (i.e., the first predetermined range mentioned above). If they are within the allowable working range, the installation angle Rota angle is updated, the transformation matrix Rotation_Matrix is ​​recalculated, and the obstacle position information is compensated and corrected; if they exceed the allowable working range, the driver / vehicle manager is promptly informed that the sensor requires physical mechanical adjustment, and the vehicle is downgraded to or even exits the autonomous driving.

[0189] For the specific execution process of the horizontal calibration module, see Figure 7 The horizontal calibration process diagram shown in the figure is explained in detail:

[0190] S1: According to the boundary line equation of the high-precision map road and the detection error of the sensor itself, an appropriate lateral distance error Dx (i.e., the sixth threshold) is set to further screen the initially screened absolute stationary obstacles to obtain effective stationary obstacles, i.e., the first stationary obstacles mentioned above;

[0191] S2: When the road boundary line equation of the high-precision map is a straight line, similar to Figure 2 The horizontal straight road calibration scene shown in the figure enters S3 horizontal straight road calibration; when the road boundary line equation of the high-precision map is not a straight line, similar to Figure 3 The horizontal calibration non-straight road scenario shown enters S4 horizontal non-straight road calibration;

[0192] S3: Horizontal straight road calibration, determine whether the yaw angle yaw_rate and lateral acceleration of the autonomous driving vehicle are stable and fluctuate within a small range around 0. If so, enter S31 and continue calibration; otherwise, end the calibration.

[0193] S31: Determine whether the number of effective stationary obstacles M_Line is greater than the set threshold (i.e. the fourth threshold mentioned above) N_Threshold_Line. The settings of M_Line and N_Threshold_Line here depend on the road edge type (flower beds, walls, metal railings, etc.) provided by the high-precision map. If so, enter S32 and continue calibration; otherwise, end calibration.

[0194] S32: According to all valid stationary obstacle position information (X0, Y0), least square fitting is performed on all position points, and the angle between the obtained straight line equation, i.e., the first fitting line equation, and the boundary line equation is the new yaw angle yaw_angle_new;

[0195] S33: Count the yaw angle yaw_angle_new_left of the left boundary line and the yaw angle yaw_angle_new_right of the right boundary line respectively;

[0196] S34: According to the difference between yaw_angle_new_left and yaw_angle_new_right, the calibrated yaw angle yaw_angle_new is obtained through comprehensive judgment. The logic is as follows:

[0197] Both are invalid angles, no output;

[0198] If there is only one valid angle between the two, the output will be the valid angle;

[0199] When both are valid angles and the difference is not large, take the average value for output;

[0200] Both are valid angles, but the difference is large. The effective angle of the boundary line that the vehicle approaches is used as the output.

[0201] S35: To ensure the stability of the yaw angle, multiple frames are accumulated and calculated to obtain the final yaw_angle, and the horizontal straight calibration process ends.

[0202] S4: Horizontal non-straight road calibration, determine whether the yaw_rate and lateral acceleration of the vehicle are stable and fluctuate within a small range around a non-zero value. If so, enter S41 and continue calibration; otherwise, end calibration.

[0203] S41: Determine whether the number of valid stationary obstacles M_No_Line is greater than the set threshold N_Threshold_No_Line. The settings of M_No_Line and N_Threshold_No_Line here depend on the road edge type (flower bed, wall, metal railing, etc.) provided by the high-precision map; if so, enter S42 and continue calibration; otherwise, end calibration.

[0204] S42: The currently valid stationary obstacle position points are formed into a set D, and the traversal polling flag Flag=0 is recorded;

[0205] S43: Calculate the distances between all the position points in the set and the boundary line equation to form an array Dis_n, and obtain its average value Dis_Ave;

[0206] S44: Set the minimum distance Dis_Threshold (ie, the fifth threshold value) for judging whether the curves are parallel or overlapped, and determine whether Dis_Ave is less than Dis_Threshold. If so, proceed to S45; otherwise, proceed to S46.

[0207] S45: Determine whether the traversal polling flag Flag is 0. If so, it indicates that the current yaw angle is normal and no further calibration is required, and the calibration ends; otherwise, enter S48;

[0208] S46: within the allowed angle working range (i.e. the third predetermined range mentioned above), traverse with the set angle step value (i.e. the preset angle step value mentioned above) angle_step, generate a new rotation matrix with yaw_angle_new, re-convert, translate and screen the initially screened absolute stationary obstacles, form a new set D', and set the traversal polling flag Flag=1;

[0209] S47: Determine whether the traversal is completed. If so, it means that the traversal is completed but the appropriate installation angle is still not found, indicating that the current allowable working range is exceeded and the calibration is completed; otherwise, enter S43, recalculate the minimum distance between all position points in the set D' and the boundary line equation, and construct an array Dis_n at the same time, and calculate its average value Dis_Ave;

[0210] S48: Determine the updated angle as the calibrated yaw angle.

[0211] S49: To ensure the stability of the yaw angle, after accumulating statistics over multiple frames, the final yaw angle yaw_angle is obtained, and the horizontal non-straight track calibration process ends.

[0212] like Figure 8 As shown in the figure, the pitch calibration module is described in detail:

[0213] S5: According to the height information of the gantry / road sign provided by the high-precision map, the stationary obstacle is screened out to obtain the second stationary obstacle;

[0214] S51: To ensure the validity and reliability of the height information, select a suitable longitudinal distance interval (i.e., the fourth predetermined range), count the height information corresponding to the stationary obstacles falling within the longitudinal distance interval, and count the number of the stationary obstacles;

[0215] S52: Determine whether the number of stationary obstacles is greater than a set threshold M (i.e., the seventh threshold mentioned above). If so, proceed to S53; otherwise, the pitch calibration ends.

[0216] S53: Establish a coordinate system with the longitudinal distance X between the obstacle and the gantry / road sign as the horizontal axis and the height Z as the vertical axis to form a (Z, X) point set;

[0217] S54: Least squares fitting is used, and the angle between the obstacle fitting line and the gantry / road sign fitting line is the final pitch angle pitch_angle.

[0218] In the above embodiments of the present application, the description of each embodiment has its own emphasis. For parts that are not described in detail in a certain embodiment, please refer to the relevant description of other embodiments.

[0219] In the several embodiments provided in this application, it should be understood that the disclosed technical content can be implemented in other ways. Among them, the device embodiments described above are only schematic. For example, the division of the above-mentioned units can be a logical function division. There may be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of units or modules, which can be electrical or other forms.

[0220] The units described above as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple units. Some or all of the units may be selected according to actual needs to achieve the purpose of the present embodiment.

[0221] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0222] If the above-mentioned integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application is essentially or the part that contributes to the prior art or all or part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including a number of instructions to enable a computer device (which can be a personal computer, a server or a network device, etc.) to perform all or part of the steps of the above-mentioned methods of each embodiment of the present application. The aforementioned storage medium includes: U disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), mobile hard disk, disk or optical disk and other media that can store program codes.

[0223] From the above description, it can be seen that the above embodiments of the present application achieve the following technical effects:

[0224] 1) The control method of the autonomous driving vehicle of the present application, first, during the driving process of the autonomous driving vehicle, when the autonomous driving vehicle satisfies the change rate of the yaw angle of the autonomous driving vehicle is less than the first threshold and the change rate of the predetermined acceleration is less than the second threshold, the yaw angle of the sensor is calibrated, and / or when the autonomous driving vehicle satisfies the change rate of the slope of the road where the autonomous driving vehicle is located is less than the third threshold, the pitch angle of the sensor is calibrated to obtain the target parameter, ensuring that the attitude information of the calibrated sensor is relatively accurate; then, when the target parameter is within the first predetermined range, the position information of the obstacle measured by the sensor is corrected according to the target parameter, ensuring that the sensor accurately locates the obstacles around the vehicle, thereby ensuring the driving safety of the autonomous driving vehicle; when the target parameter is not within the first predetermined range, it means that the positioning of the sensor is inaccurate and the risk of autonomous driving is high. At this time, the autonomous driving level of the autonomous driving vehicle is reduced to avoid driving accidents caused by inaccurate positioning of the sensor. The present application effectively solves the problem of inaccurate sensor positioning of autonomous driving vehicles in the prior art as the driving time increases.

[0225] 2) The control device of the autonomous driving vehicle of the present application, through the calibration unit, during the driving process of the autonomous driving vehicle, when the autonomous driving vehicle satisfies the change rate of the yaw angle of the autonomous driving vehicle is less than the first threshold and the change rate of the predetermined acceleration is less than the second threshold, the yaw angle of the sensor is calibrated, and / or when the autonomous driving vehicle satisfies the change rate of the slope of the road where the autonomous driving vehicle is located is less than the third threshold, the pitch angle of the sensor is calibrated to obtain the target parameter, thereby ensuring that the attitude information of the calibrated sensor is relatively accurate; through the correction unit, when the target parameter is within the first predetermined range, the position information of the obstacle measured by the sensor is corrected according to the target parameter, thereby ensuring that the sensor accurately locates the obstacles around the vehicle, thereby ensuring the driving safety of the autonomous driving vehicle; when the target parameter is not within the first predetermined range, it means that the positioning of the sensor is inaccurate and the risk of autonomous driving is high. The autonomous driving level of the autonomous driving vehicle is reduced by the reduction unit, thereby avoiding driving accidents caused by inaccurate positioning of the sensor. The present application effectively solves the problem of inaccurate sensor positioning of autonomous driving vehicles in the prior art as the driving time increases.

[0226] 3) In the automatic driving system of the present application, the processor is used to execute any one of the control methods. The method first calibrates the yaw angle of the sensor when the automatic driving vehicle satisfies the change rate of the yaw angle of the automatic driving vehicle is less than the first threshold and the change rate of the predetermined acceleration is less than the second threshold during the driving process of the automatic driving vehicle, and / or calibrates the pitch angle of the sensor when the automatic driving vehicle satisfies the change rate of the slope of the road where the automatic driving vehicle is located is less than the third threshold, and obtains the target parameter, ensuring that the attitude information of the calibrated sensor is relatively accurate; then, when the target parameter is within the first predetermined range, the position information of the obstacle measured by the sensor is corrected according to the target parameter, ensuring that the sensor accurately locates the obstacles around the vehicle, thereby ensuring the driving safety of the automatic driving vehicle; when the target parameter is not within the first predetermined range, it means that the positioning of the sensor is inaccurate and the risk of automatic driving is high. At this time, the automatic driving level of the automatic driving vehicle is reduced to avoid driving accidents caused by inaccurate positioning of the sensor. The present application effectively solves the problem of inaccurate sensor positioning of automatic driving vehicles in the prior art as the driving time increases.

[0227] The above description is only the preferred 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 modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application shall be included in the protection scope of the present application.

Claims

1. A control method for an autonomous driving vehicle, characterized in that: include: During the driving of an autonomous driving vehicle, when the autonomous driving vehicle meets predetermined conditions, calibrating predetermined parameters of a target sensor to obtain target parameters, wherein the predetermined parameters include at least one of a yaw angle and a pitch angle, and when the predetermined parameters include the yaw angle, the predetermined conditions include: a rate of change of the yaw angle of the autonomous driving vehicle is less than a first threshold, and a rate of change of a predetermined acceleration is less than a second threshold, and the predetermined acceleration is an acceleration of the autonomous driving vehicle perpendicular to the driving direction, and when the predetermined parameters include the pitch angle, the predetermined conditions include: a rate of change of the slope of the road where the autonomous driving vehicle is located is less than a third threshold; When the target parameter is within a first predetermined range, correcting the position information of the obstacle measured by the target sensor according to the target parameter; When the target parameter is not within the first predetermined range, the autonomous driving level of the autonomous driving vehicle is reduced.

2. The method according to claim 1, characterized in that Calibrate the predetermined parameters of the target sensor to obtain the target parameters, including: Calibrate the yaw angle according to a frame of data collected by the target sensor and a boundary line equation to obtain a calibrated yaw angle, wherein the boundary line equation is an equation of a boundary line of the road where the autonomous driving vehicle is located determined according to a high-precision map; A statistical characteristic analysis is performed on the calibrated yaw angle corresponding to a plurality of consecutive frames of data to obtain a final yaw angle.

3. The method according to claim 1, characterized in that The yaw angle is calibrated according to a frame of data collected by the target sensor and a boundary line equation to obtain a calibrated yaw angle, including: Determining, according to one frame of the data, location information and types of a plurality of first stationary obstacles, wherein the first stationary obstacles are obstacles located on both sides of the road and do not move; When the boundary line equation is a straight line equation and the number of first target obstacles is greater than a fourth threshold, determining the calibrated yaw angle according to the boundary line equation and position information of a plurality of first target obstacles, wherein the first target obstacle is the first stationary obstacle of the same type; When the boundary line equation is not the straight line equation and the number of the first stationary obstacles of the same type is greater than the fourth threshold, the calibrated yaw angle is determined according to position information of a plurality of the first target obstacles.

4. The method according to claim 3, characterized in that Determining the calibrated yaw angle according to the boundary line equation and the position information of the plurality of first target obstacles includes: Performing least square fitting on the position information of the first target obstacle located on the same side of the road to obtain a first fitting line equation; The calibrated yaw angle is determined according to the first fitting line equation and the boundary line equation.

5. The method according to claim 4, characterized in that The road includes a first boundary line and a second boundary line relative to each other, the boundary line equation includes a first boundary line sub-equation and a second boundary line sub-equation, and when a plurality of the first target obstacles are located on different sides of the road, the first fitting line equation includes a first fitting line sub-equation and a second fitting line sub-equation, the first boundary line sub-equation and the first fitting line sub-equation correspond to the first boundary line respectively, the second boundary line sub-equation and the second fitting line sub-equation correspond to the second boundary line respectively, and the calibrated yaw angle is determined according to the first fitting line equation and the boundary line equation, including: In a case where the first target obstacle is located on a different side of the road, determining an angle between the first boundary line sub-equation and the first fitting line sub-equation as a first angle, and an angle between the second boundary line sub-equation and the second fitting line sub-equation as a second angle; In a case where one of the first angle and the second angle is an invalid angle, determining a valid angle between the first angle and the second angle as the calibrated yaw angle, the number of the first target obstacles corresponding to the invalid angle is less than or equal to the fourth threshold, and the number of the first target obstacles corresponding to the valid angle is greater than the fourth threshold; When both the first angle and the second angle are the valid angles and the difference between the first angle and the second angle is within a second predetermined range, determining an average value of the first angle and the second angle as the calibrated yaw angle; When the first angle and the second angle are both the valid angles, and the difference between the first angle and the second angle is not within the second predetermined range, the boundary line distance between the autonomous driving vehicle and the first boundary line and the boundary line distance between the autonomous driving vehicle and the second boundary line are respectively obtained, and the angle corresponding to the minimum value of the boundary line distance between the first angle and the second angle is determined to be the calibrated yaw angle.

6. The method according to claim 4, characterized in that The road includes a first boundary line and a second boundary line relative to each other, the boundary line equation includes a first boundary line sub-equation corresponding to the first boundary line and a second boundary line sub-equation corresponding to the second boundary line, and determining the calibrated yaw angle according to the first fitting line equation and the boundary line equation includes: When all the first target obstacles are located on the same side of the road, a first distance and a second distance are obtained, wherein the first distance is the distance between the first boundary line sub-equation and the first fitting line equation, and the second distance is the distance between the second boundary line sub-equation and the first fitting line equation; Determine, in the first boundary line sub-equation and the second boundary line sub-equation, the sub-equation corresponding to the minimum value of the first distance and the second distance as the target sub-boundary line equation; The angle between the first fitting line equation and the target sub-boundary line equation is determined as the calibrated yaw angle.

7. The method according to claim 3, characterized in that Determining the calibrated yaw angle according to the position information of the plurality of first target obstacles includes: Determine the distance between the position information of the second target obstacle and the boundary line equation as a third distance, and calculate the average value of each of the third distances, wherein the second target obstacle is all the first target obstacles located on the same side of the road; determining whether the average value is less than a fifth threshold; When the average value is greater than or equal to the fifth threshold, determining the calibrated yaw angle according to the yaw angle, a third predetermined range, and a preset angle step; When the average value is less than the fifth threshold, the yaw angle is determined to be the calibrated yaw angle.

8. The method according to claim 7, characterized in that When the average value is greater than or equal to the fifth threshold, determining the calibrated yaw angle according to the yaw angle, the third predetermined range, and the preset angle step includes: an updating step, when the average value is greater than or equal to the fifth threshold, starting from the yaw angle, increasing or decreasing the yaw angle stepwise by the preset angle to obtain an updated angle, wherein the updated angle is within the third predetermined range; a correction step, correcting the position information of each of the second target obstacles according to the update angle to obtain corrected position information; A calculation step, calculating an average distance between the corrected position information and the boundary line equation; A first determination step is to determine whether the distance average value is less than the fifth threshold; The second determination step is to determine the updated angle as the calibrated yaw angle when the distance average value is less than the fifth threshold value, and to sequentially perform the updating step, the correcting step, the calculating step and the first determination step when the distance average value is greater than or equal to the fifth threshold value until the distance average value is less than the fifth threshold value or the updated angle reaches the boundary of the third predetermined range.

9. The method according to claim 3, characterized in that: Before calibrating the predetermined parameters of the target sensor to obtain the target parameters, the method further includes: Obtaining a sixth threshold, a speed of the autonomous driving vehicle, and a speed of each of the obstacles on both sides of the road; According to the vehicle speed and each of the speeds, determining that the obstacle whose speed is opposite to the vehicle speed and whose predetermined difference is within a fifth predetermined range is an initial stationary obstacle, wherein the predetermined difference is an absolute value of a difference between an absolute value of the speed and an absolute value of the vehicle speed; Calculating the distance between each of the initial stationary obstacles and the boundary line equation according to the position information of the initial stationary obstacles and the boundary line equation to obtain a plurality of fourth distances; Determine that the initial stationary obstacle corresponding to the fourth distance smaller than the sixth threshold is the first stationary obstacle.

10. The method according to any one of claims 1 to 9, characterized in that Calibrate the predetermined parameters of the target sensor to obtain the target parameters, including: Extracting a second stationary obstacle from the continuous multiple frames of data collected by the target sensor, and acquiring a first height, where the first height is height information of the second stationary obstacle measured by the target sensor, and the second stationary obstacle is an obstacle located on the road or on both sides of the road, within a fourth predetermined range from the autonomous driving vehicle, and does not move; Obtaining a fitting relationship between a second height and a fifth distance to obtain a second fitting line equation, wherein the second height is height information of the second stationary obstacle determined according to a high-precision map, and the fifth distance is a real-time distance between the second stationary obstacle and the autonomous driving vehicle in the driving direction; When the number of the second stationary obstacles is greater than a seventh threshold, fitting the first height and the fifth distance to obtain a third fitting line equation; Determine an angle between the second fitting line equation and the third fitting line equation as a final pitch angle.

11. A control device for an autonomous driving vehicle, characterized in that: include: a calibration unit, configured to calibrate predetermined parameters of a target sensor to obtain target parameters during driving of an autonomous driving vehicle, when the autonomous driving vehicle satisfies predetermined conditions, wherein the predetermined parameters include at least one of a yaw angle and a pitch angle, and when the predetermined parameters include the yaw angle, the predetermined conditions include: a rate of change of the yaw angle of the autonomous driving vehicle is less than a first threshold, and a rate of change of a predetermined acceleration is less than a second threshold, and the predetermined acceleration is an acceleration of the autonomous driving vehicle perpendicular to a driving direction, and when the predetermined parameters include the pitch angle, the predetermined conditions include: a rate of change of a slope of a road on which the autonomous driving vehicle is located is less than a third threshold; a correction unit, configured to correct the position information of the obstacle measured by the target sensor according to the target parameter when the target parameter is within a first predetermined range; A reducing unit is used to reduce the autonomous driving level of the autonomous driving vehicle when the target parameter is not within the first predetermined range.

12. A computer-readable storage medium, characterized in that: The computer-readable storage medium includes a stored program, wherein the program executes the method of any one of claims 1 to 10.

13. A processor, characterized in that: The processor is used to run a program, wherein the program executes the method according to any one of claims 1 to 10 when running.

14. An automatic driving system, characterized in that: include: Autonomous vehicles; One or more processors, a memory, and one or more programs, wherein the one or more programs are stored in the memory and are configured to be executed by the one or more processors, and the one or more programs include methods for executing any one of claims 1 to 10.

Images