Parking collision detection method and device, vehicle and storage medium

Abstract

The embodiment of the invention discloses a parking collision detection method and device, a vehicle and a storage medium. The method comprises the following steps: acquiring a deviation between a vehicle and a current tracking path; if the deviation is greater than a preset deviation, determining a predicted driving track of the vehicle based on the current pose of the vehicle and a preset kinematic model; if the deviation is smaller than or equal to the preset deviation, determining a predicted driving track of the vehicle based on a projection point of the vehicle on the current tracking path and a road point set of the current tracking path; and acquiring a collision risk level corresponding to the predicted driving track. According to the method, the driving track is predicted, collision detection is carried out, the system is allowed to make a response when the vehicle is not close to the obstacle, possible hard braking under the emergency condition is avoided, and the risk of accidental collision is reduced.

Description

Technical Field

[0001] The present application belongs to the field of intelligent driving technology, and specifically relates to a parking collision detection method, device, vehicle and storage medium. Background Art

[0002] As an important part of intelligent driving technology, automatic parking systems have been widely used in the market. In the process of controlling the vehicle to park, the existing automatic parking system may deviate from the planned trajectory due to factors such as speed changes, environmental perception errors, and tracking deviations, resulting in the existing collision detection methods misjudging or missing the vehicle collision risk. Summary of the invention

[0003] In view of the above problems, the present application proposes a parking collision detection method, device, vehicle and storage medium to improve the above problems.

[0004] In a first aspect, an embodiment of the present application provides a parking collision detection method, the method comprising: obtaining a deviation between a vehicle and a current tracking path; if the deviation is greater than a preset deviation, determining a predicted driving trajectory of the vehicle based on the current position of the vehicle and a preset kinematic model; if the deviation is less than or equal to the preset deviation, determining a predicted driving trajectory of the vehicle based on a projection point of the vehicle on the current tracking path and a set of waypoints of the current tracking path; and obtaining a collision risk level corresponding to the predicted driving trajectory.

[0005] In a second aspect, an embodiment of the present application provides a parking collision detection device, which includes: a deviation acquisition unit, used to acquire the deviation between a vehicle and a current tracking path; a first determination unit, used to determine the predicted driving trajectory of the vehicle based on the current posture of the vehicle and a preset kinematic model if the deviation is greater than a preset deviation; a second determination unit, used to determine the predicted driving trajectory of the vehicle based on the projection point of the vehicle on the current tracking path and the set of waypoints of the current tracking path if the deviation is less than or equal to the preset deviation; and a level acquisition unit, used to acquire the collision risk level corresponding to the predicted driving trajectory.

[0006] In a third aspect, an embodiment of the present application provides a vehicle, comprising one or more processors and a memory; one or more programs, wherein the one or more programs are stored in the memory and configured to be executed by the one or more processors, and the one or more programs are configured to execute the above-mentioned method.

[0007] In a fourth aspect, an embodiment of the present application provides a computer-readable storage medium, in which a program code is stored, wherein the above method is executed when the program code is run.

[0008] The embodiments of the present application provide a parking collision detection method, device, vehicle and storage medium. First, the deviation between the vehicle and the current tracking path is obtained. If the deviation is greater than the preset deviation, the predicted driving trajectory of the vehicle is determined based on the current position and preset kinematic model of the vehicle. If the deviation is less than or equal to the preset deviation, the predicted driving trajectory of the vehicle is determined based on the projection point of the vehicle on the current tracking path and the set of waypoints of the current tracking path, and then the collision risk level corresponding to the predicted driving trajectory is obtained. Through the above method, by predicting the driving trajectory and performing collision detection, the system is allowed to react before the vehicle approaches an obstacle, avoiding possible hard braking in an emergency and reducing the risk of accidental collisions. BRIEF DESCRIPTION OF THE DRAWINGS

[0009] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For those skilled in the art, other drawings can be obtained based on these drawings without creative work.

[0010] Figure 1 A flow chart of a parking collision detection method proposed in one embodiment of the present application is shown;

[0011] Figure 2 A flowchart of a parking collision detection method proposed in another embodiment of the present application is shown;

[0012] Figure 3 A schematic diagram showing the process described in step S210 to step S230 in another embodiment of the present application is shown;

[0013] Figure 4 A structural block diagram of a parking collision detection device proposed in an embodiment of the present application is shown;

[0014] Figure 5 A structural block diagram of a vehicle for executing a parking collision detection method according to an embodiment of the present application is shown;

[0015] Figure 6 A storage unit in an embodiment of the present application is shown, which is used to store or carry program codes for implementing the parking collision detection method according to an embodiment of the present application. DETAILED DESCRIPTION

[0016] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions 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 are within the scope of protection of this application.

[0017] The embodiments of the present application propose a parking collision detection method, device, vehicle, and storage medium. First, the deviation between the vehicle and the current tracking path is obtained. If the deviation is greater than the preset deviation, the predicted driving trajectory of the vehicle is determined based on the current position and preset kinematic model of the vehicle. If the deviation is less than or equal to the preset deviation, the predicted driving trajectory of the vehicle is determined based on the projection point of the vehicle on the current tracking path and the set of waypoints of the current tracking path, and then the collision risk level corresponding to the predicted driving trajectory is obtained. Through the above method, by predicting the driving trajectory and performing collision detection, the system is allowed to react when the vehicle has not yet approached an obstacle, avoiding possible hard braking in an emergency and reducing the risk of accidental collisions.

[0018] The embodiments of the present application will be described in detail below with reference to the accompanying drawings.

[0019] See also Figure 1 , an embodiment of the present application provides a parking collision detection method, the method comprising:

[0020] Step S110: Obtain the deviation between the vehicle and the current tracking path.

[0021] In the embodiment of the present application, the vehicle is a vehicle that is currently undergoing automatic parking. The current tracking path is an automatic parking path generated by the automatic parking system based on the parking start point to the parking end point of the vehicle. The current tracking path can be understood as a path composed of multiple waypoints. Each waypoint can be represented by (x, y, yaw), wherein x represents the horizontal coordinate of the waypoint, which is used to indicate the horizontal position of the waypoint on a two-dimensional plane or a map; y represents the vertical coordinate of the waypoint, which is used to indicate the vertical position of the waypoint on a two-dimensional plane or a map; yaw represents the yaw angle (or heading angle) of the waypoint, which is used to indicate the direction the vehicle should face when arriving at the waypoint. In other words, the waypoints in the waypoint set of the current tracking path can be used not only to indicate the geographical location that the vehicle should pass through, but also to indicate the correct orientation that the vehicle should adjust to when arriving at each waypoint.

[0022] The deviation between the vehicle and the current tracking path can be understood as the degree of deviation between the current vehicle and the current tracking path. The greater the deviation between the vehicle and the current tracking path, the greater the degree of deviation between the vehicle and the current tracking path; conversely, the smaller the deviation between the vehicle and the current tracking path, the smaller the degree of deviation between the vehicle and the current tracking path.

[0023] In the embodiment of the present application, the deviation between the vehicle and the current tracking path can be calculated based on the current posture of the vehicle and the coordinates of a target waypoint in the waypoint set included in the current tracking path. The target waypoint can be understood as the waypoint in the waypoint set with the smallest Euclidean distance from the current posture of the vehicle; the current posture of the vehicle can also be represented by (x, y, yaw) or (x, y, yaw, curv), which is used to characterize the current geographical location and orientation of the vehicle and represent the current curvature of the vehicle, that is, the curvature of the vehicle's driving path.

[0024] Step S120: If the deviation is greater than a preset deviation, a predicted driving trajectory of the vehicle is determined based on the current posture of the vehicle and a preset kinematic model.

[0025] In the embodiment of the present application, the preset deviation is a preset minimum deviation between the vehicle and the current tracking path for indicating the switching mode of determining the predicted driving trajectory. The predicted driving trajectory can be understood as a set of trajectory points that the vehicle will travel within a preset time in the future. Each trajectory point can also be represented by (x, y, yaw).

[0026] After obtaining the deviation between the current vehicle and the current tracking path, the deviation can be compared with the preset deviation to determine which method should be selected to determine the predicted driving trajectory. Among them, the method of determining the predicted driving trajectory can include at least two methods, one is to determine it through the kinematic model of the vehicle, and the other is to determine it by calculating the distance between each waypoint.

[0027] As a method, if the deviation between the current vehicle and the current tracking path is greater than a preset deviation, it can be determined to choose to determine the vehicle's predicted driving trajectory through the vehicle's kinematic model; conversely, if the deviation between the current vehicle and the current tracking path is less than or equal to the preset deviation, it can be determined to choose to determine the vehicle's predicted driving trajectory by calculating the distance between each waypoint.

[0028] Among them, in the method of determining the predicted driving trajectory of the vehicle through the kinematic model of the vehicle, the set of trajectory points that the vehicle will travel within a preset time can be predicted based on the current vehicle posture (x, y, yaw, curv) and the kinematic model of the vehicle, which is called the predicted driving trajectory. Among them, when parking, the vehicle generally moves in a plane, and the kinematic model of the vehicle can be represented by the following motion equation. The motion equation is as follows:

[0029] Step S130: If the deviation is less than or equal to the preset deviation, the predicted driving trajectory of the vehicle is determined based on the projection point of the vehicle on the current tracking path and the set of waypoints of the current tracking path.

[0030] In an embodiment of the present application, if the deviation between the current vehicle and the current tracking path is less than or equal to a preset deviation, the predicted driving trajectory of the vehicle can be determined by calculating the distance between each waypoint.

[0031] When determining the predicted driving path of the vehicle by calculating the distance between each waypoint, the distance between each waypoint can be calculated one by one from the projection point of the current vehicle on the current tracking path until the cumulative distance is greater than the preset distance, and this section of the path is recorded as the predicted driving trajectory. The projection point of the vehicle on the current tracking path refers to the waypoint with the smallest Euclidean distance from the current position of the vehicle in the set of waypoints included in the current tracking path.

[0032] Step S140: Obtaining the collision risk level corresponding to the predicted driving trajectory.

[0033] In the embodiment of the present application, the collision risk refers to the probability that a collision may occur between a vehicle and an obstacle in the predicted driving trajectory; the collision risk level can be understood as the risk degree of a possible collision of the vehicle in the predicted driving trajectory. Different predicted driving trajectories may correspond to different collision risk levels. Among them, different risk collision levels may include multiple risk collision levels divided based on preset division rules. For example, the risk collision level may include five safety levels such as kSafe, kSpdLimit, kSoftStop, kHardStop, and kReplan. Among them, the preset division rule is a pre-set rule for dividing the collision risk level. The preset division rule can be any rule that can be used to divide the collision risk level, and is not specifically limited here.

[0034] After the predicted driving trajectory of the vehicle is determined, the collision risk level corresponding to the predicted driving trajectory can be determined based on a preset rule. The preset rule is a rule pre-set to determine the collision risk level corresponding to the predicted driving path. For example, the preset trajectory can determine the corresponding collision risk level based on the corresponding relationship between the predicted driving trajectory and the collision risk level.

[0035] The present application provides a parking collision detection method, which first obtains the deviation between the vehicle and the current tracking path. If the deviation is greater than the preset deviation, the predicted driving trajectory of the vehicle is determined based on the current position and the preset kinematic model of the vehicle. If the deviation is less than or equal to the preset deviation, the predicted driving trajectory of the vehicle is determined based on the projection point of the vehicle on the current tracking path and the set of waypoints of the current tracking path, and then the collision risk level corresponding to the predicted driving trajectory is obtained. Through the above method, by predicting the driving trajectory and performing collision detection, the system is allowed to react when the vehicle has not yet approached an obstacle, avoiding possible hard braking in an emergency and reducing the risk of accidental collision.

[0036] See also Figure 2 , an embodiment of the present application provides a parking collision detection method, the method comprising:

[0037] Step S210: Acquire the projection point of the vehicle on the current tracking path.

[0038] In the embodiment of the present application, the projection point of the vehicle on the current tracking path refers to the waypoint with the smallest Euclidean distance from the current position and posture of the vehicle in the set of waypoints included in the current tracking path.

[0039] As one approach, step S210 may include: obtaining a set of waypoints of a current tracking path and a current posture of the vehicle; obtaining the Euclidean distance between the current posture and each waypoint in the waypoint set; and determining the waypoint in the waypoint set with the smallest corresponding Euclidean distance as the projection point of the vehicle on the current tracking path.

[0040] The waypoint set of the current tracking path refers to the set of all waypoints included in the current tracking path. The waypoint set of the current tracking path can be obtained when the current tracking path is just generated, that is, the waypoint set can be obtained by traversing each waypoint in the current tracking path. The current position of the vehicle refers to the current geographical location of the vehicle, the vehicle orientation, and the current curvature of the vehicle.

[0041] After obtaining the waypoint set of the current tracking path and the current posture of the vehicle, the Euclidean distance between all waypoints in the waypoint set and the current posture of the vehicle can be calculated according to the current posture of the vehicle, that is, the Euclidean distance between each waypoint in the waypoint set and the current posture of the vehicle is calculated.

[0042] After calculating the Euclidean distance between each waypoint and the current position of the vehicle, all waypoints in the waypoint set can be sorted according to the size of the Euclidean distance corresponding to each waypoint to obtain a sorted waypoint set, and then the waypoints in the sorted waypoint set whose corresponding Euclidean distance is ranked at a specified position are used as projection points of the vehicle on the current tracking path. The specified position can be the front position or the last position in the waypoint set, or other positions, which are not specifically limited here.

[0043] Optionally, in an embodiment of the present application, a waypoint in the waypoint set having the smallest Euclidean distance to the current position of the vehicle is selected as the projection point of the vehicle on the current tracking path.

[0044] Step S220: Obtain a difference vector between the projection point and the current posture.

[0045] In an embodiment of the present application, the difference vector between the projection point of the vehicle on the current tracking path and the current posture can be understood as the error between the projection point and the current posture, and the error can be expressed in the form of a vector.

[0046] Step S230: Projecting the difference vector onto the normal vector of the projection point to determine the deviation between the vehicle and the current tracking path.

[0047] In the embodiment of the present application, the degree of deviation between the vehicle and the current tracking path can be measured by a lateral deviation, where the lateral deviation refers to the projection of the difference vector on the normal vector of the projection point.

[0048] Exemplarily, the process described in step S210 to step S230 can be as follows: Figure 3 As shown, in Figure 3 In the figure, it is assumed that the solid line is the current tracking path of the vehicle, point c is the center point of the rear axle of the vehicle, and point m is the projection point calculated in step S210. Then vector x represents the position of the vehicle, and vector x m represents the projection point, vector τ represents the tangent vector, vector n represents the normal vector, and vector θ represents the angle between the vector and the positive direction of the x-axis. According to vector knowledge, the error between the vehicle pose point c and the matching point m is the vector xx m . The lateral deviation is the vector xx m In the normal vector n m The projection on the vertical deviation is the vector xx m In the tangent vector τ m The degree of deviation between the vehicle and the current tracking path should be measured by the lateral deviation, so the lateral deviation e1 between the vehicle and the current tracking path is e1 = (xx m )·n m .

[0049] Step S240: If the deviation is greater than a preset deviation, a predicted driving trajectory of the vehicle is determined based on the current posture of the vehicle and a preset kinematic model.

[0050] As a method, the current posture of the vehicle is input into the preset kinematic model to obtain the predicted driving trajectory of the vehicle output by the preset kinematic model, wherein the predicted driving trajectory is a set of trajectory points that the vehicle will travel within a preset time output by the preset kinematic model.

[0051] In an embodiment of the present application, the preset kinematic model may be a pre-trained neural network model, which may be used to predict the trajectory points that the vehicle will travel within a preset time based on the current position of the input vehicle. The preset time may be determined based on the length of the training data input when training the preset kinematic model. Optionally, the preset time may also be a custom time length. For example, the preset time may be 2s in the future, which is not specifically limited here. The training data and training methods used in training the preset kinematic model are common methods in deep learning and will not be described in detail here.

[0052] When the preset kinematic model is a neural network model, after obtaining the current posture of the vehicle, the current posture of the vehicle can be input into the preset kinematic model, so that the preset kinematic model can output a set of trajectory points that the vehicle will travel within a preset time, that is, the preset kinematic model can output a predicted driving trajectory.

[0053] Step S250: If the deviation is less than or equal to the preset deviation, the predicted driving trajectory of the vehicle is determined based on the projection point of the vehicle on the current tracking path and the set of waypoints of the current tracking path.

[0054] As a method, starting from the projection point in the waypoint set, the distances between two adjacent waypoints are calculated in sequence to obtain multiple distances; the sum of the distances of the multiple distances is calculated in sequence; if the sum of the distances is greater than a preset distance, the multiple waypoints corresponding to the distance sum are used as the predicted driving trajectory of the vehicle.

[0055] In the embodiment of the present application, the distance between two adjacent waypoints refers to the straight-line distance between the two adjacent waypoints, which can be Euclidean distance, Manhattan distance, or Chebyshev distance, and is not specifically limited here.

[0056] As can be known from the above content, the projection point is also a waypoint in the waypoint set corresponding to the current tracking path. Therefore, what is calculated in this step is the distance between each two adjacent waypoints in the waypoint set starting from the projection point. For example, if the waypoint set includes waypoint 1, waypoint 2, waypoint 3, waypoint 4, waypoint 5 and waypoint 6, and the projection point is waypoint 2, then the distances between the two adjacent waypoints to be calculated are: distance 1 between waypoint 2 and waypoint 3, distance 2 between waypoint 3 and waypoint 4, distance 3 between waypoint 4 and waypoint 5, and distance 4 between waypoint 5 and waypoint 6, that is, 4 distances can be calculated.

[0057] When calculating the sum of distances of multiple distances in sequence, the sum of distances may be calculated once after calculating one distance, or the sum of distances of multiple distances may be calculated in sequence after calculating multiple distances. For example, as mentioned above, the sum of distances may be calculated once when distance 1 is calculated, and once again after distance 2 is calculated, and so on; or, after distance 1, distance 2, distance 3, and distance 4 are calculated, the sum of distances may be calculated in sequence, that is, the sum of distance 1 and distance 2 is calculated, and the sum of distance 1 and distance 3 is calculated, and so on, and no specific limitation is made here.

[0058] In the embodiment of the present application, the preset distance is a preset path length that the vehicle may travel during future driving. For example, the preset distance can be set to 2.5m.

[0059] Each time a distance sum is calculated, the distance sum is compared with the preset distance. If the distance sum is less than the preset distance, the distance sum is continued to be calculated. If the distance sum is greater than the preset distance, the distance sum is stopped from being calculated, and the multiple road points corresponding to the distance sum greater than the preset distance are used as the predicted driving trajectory of the vehicle. For example, as in the above example, after calculating distance 1, distance 2, distance 3, and distance 4, the distance sum 1 of distance 1 is first calculated, and the distance sum 1 is compared with the preset distance. If the distance sum 1 is greater than the preset distance, the road points 2 and 3 corresponding to the distance sum 1 are determined as the predicted driving trajectory of the vehicle; if the distance sum 1 is less than the preset distance, the distance sum 2 of the distance sum 1 and the distance sum 2 is continued to be calculated, and the distance sum 2 is compared with the preset distance. If the distance sum is greater than the preset distance, the road points 2, 3, and 4 corresponding to the distance sum 2 are used as the predicted driving distance of the vehicle. Otherwise, the distance sum 3 is continued to be calculated according to the above method. ..., and so on, until the distance sum is greater than the preset distance, the distance sum is stopped from being calculated.

[0060] Step S260: Obtain the distance between the predicted driving trajectory and the obstacle.

[0061] In the embodiment of the present application, the distance between the predicted driving trajectory and the obstacle refers to the distance between each trajectory point in the predicted driving trajectory and the obstacle closest to it.

[0062] As a method, a grid map and high and low obstacle information of the vehicle's current environment are obtained; based on the grid map and the high and low obstacle information, an obstacle distance map is generated; from the obstacle distance map, a minimum distance value of each trajectory point in the predicted driving trajectory is obtained, and the minimum distance value is the distance between each trajectory point and its nearest obstacle; different minimum distance values ​​correspond to different collision risk levels.

[0063] In an embodiment of the present application, the grid map and high and low obstacle information of the vehicle's current environment can be obtained through the perception module in the vehicle. Among them, the grid map is usually a two-dimensional array or matrix, in which each element represents a grid (or pixel) on the map; the value of the grid may represent the probability of the location being occupied, height information or other related attributes. The high and low obstacle information is an array or matrix corresponding to the grid map, which provides the height of the obstacle on each grid or other vertical dimension information.

[0064] When an obstacle distance map is generated based on the grid map and the high and low obstacle information, the grid map and the high and low obstacle information are binarized and scanned forward and backward to form an obstacle distance map.

[0065] The binarization process may include the following steps:

[0066] Set thresholds: Based on the height or other attributes of obstacles, set one or more thresholds to distinguish obstacles from non-obstacles. For example, locations with a height greater than a certain value can be considered obstacles.

[0067] Binarization: Traverse the grid map and obstacle information. For each grid, if its obstacle attribute exceeds the set threshold, it will be marked as an obstacle (such as 1) in the binary map, otherwise it will be marked as a non-obstacle (such as 0).

[0068] Before performing forward and backward scanning, the distance map can be initialized. Specifically, a matrix of the same size as the grid map is created as the distance map, and the initial value can be set to infinity or a large number to indicate that the distance is unknown at the beginning.

[0069] The forward and backward scanning may include forward scanning and backward scanning. Among them, the forward scanning may include starting from the starting position of the map (such as the current position of the vehicle) and scanning row by row or column by column along the vehicle's forward direction; for each grid, if it is an obstacle, calculate its distance from the vehicle's current position (or the last known distance point), and update the value of the position in the distance map; if it is not an obstacle, but there is an obstacle with a known distance ahead, the distance value of the position may be updated as needed (for example, set to the distance to the nearest obstacle).

[0070] Backward scanning may include scanning from the end of the map in the opposite direction to update or correct the distance value, which in some cases may help to more accurately reflect the distance of obstacles behind the vehicle.

[0071] When calculating the distance in the above way, it is necessary to consider the height of the obstacle or other vertical information. For example, if the obstacle is high, it may block the vehicle's line of sight or path, which needs to be reflected in the distance calculation. Optionally, the distance calculation method can be adjusted or additional weights can be introduced to reflect the impact of high and low obstacles on the distance.

[0072] After the above processing, a map containing the distance from each grid to the nearest obstacle can be obtained, that is, an obstacle distance map. This obstacle distance map can be used in tasks such as path planning and obstacle avoidance navigation.

[0073] The obstacle distance map includes the distance from each grid to the nearest obstacle, and each grid can be understood as a waypoint or trajectory point. After obtaining the obstacle distance map, the distance between each trajectory point in the predicted driving trajectory and its nearest obstacle can be obtained from the obstacle map.

[0074] After obtaining the minimum distance of each trajectory point, the collision risk level corresponding to each trajectory point can be determined, so that the vehicle can be controlled to perform different operations at each trajectory point.

[0075] Step S270: Based on the distance, obtaining a collision risk level corresponding to the preset driving trajectory.

[0076] As one approach, based on the minimum distance value of each trajectory point, the collision risk level corresponding to the predicted driving trajectory is determined from a plurality of pre-set collision risk levels.

[0077] The collision risk level corresponding to the predicted driving trajectory refers to the collision risk level corresponding to each trajectory point in the predicted driving trajectory, that is, the collision risk level corresponding to each trajectory point can be determined according to the minimum distance value corresponding to each trajectory point.

[0078] Step S280: controlling the vehicle to execute an operation corresponding to the collision risk level, wherein different collision risk levels correspond to different operations.

[0079] In the embodiment of the present application, different collision risk levels correspond to different operations, which means that at different collision risk levels, the vehicle can perform different operations. For example, when the collision risk level is kSafe, the vehicle can be controlled not to perform any operation; when the collision risk level is kSpdLimit, it is determined that the vehicle is close to the obstacle, and the vehicle can be controlled to perform speed limit operations to ensure the user's sense of security; when the collision risk level is kSoftStop, it is determined that the vehicle will collide with the obstacle in the future, but the distance is greater than 0.5m, and the vehicle can be controlled to perform slow braking operations; when the collision risk level is kHardStop, it is determined that the current distance between the vehicle and the obstacle is very close, and the vehicle can be controlled to perform emergency braking operations; when the collision risk level is kReplan, the future path is no longer available, and the path can be replanned. Controlling the vehicle to perform operations corresponding to the collision risk level means that, at each trajectory point, the vehicle is controlled to perform operations corresponding to the corresponding collision risk level.

[0080] The parking collision detection method provided by the present application can identify potential collision risks in advance by predicting the possible future driving trajectory of the vehicle and combining the current motion state and kinematic model of the vehicle. This not only enhances the early warning capability of the system, but also enables the system to take preventive measures before an emergency occurs, effectively avoiding problems such as sudden braking of the vehicle, and significantly improving the parking experience. Furthermore, based on dynamic response strategies of different risk levels, by converting obstacle distances into risk levels, the system can take different levels of response measures such as speed limit, slow braking or emergency braking according to actual conditions, which not only improves the flexibility of the system, but also further ensures safety and comfort during parking. Optionally, the aforementioned method also makes full use of the perception map information updated in real time, combined with the grid map and obstacle information, to achieve accurate perception of the vehicle's surrounding environment, so that the collision risk can be assessed more accurately, improving the intelligence level of the automatic parking system.

[0081] See also Figure 4 , an embodiment of the present application provides a parking collision detection device 300, the device 300 comprising:

[0082] The deviation obtaining unit 310 is used to obtain the deviation between the vehicle and the current tracking path.

[0083] As a method, the deviation acquisition unit 310 is specifically used to obtain the projection point of the vehicle on the current tracking path; obtain the difference vector between the projection point and the current posture; project the difference vector onto the normal vector of the projection point to determine the deviation between the vehicle and the current tracking path.

[0084] Furthermore, the deviation acquisition unit 310 is specifically used to obtain a set of waypoints of the current tracking path and the current posture of the vehicle; obtain the Euclidean distance between the current posture and each waypoint in the waypoint set; and determine the waypoint with the smallest corresponding Euclidean distance in the waypoint set as the projection point of the vehicle on the current tracking path.

[0085] The first determination unit 320 is used to determine the predicted driving trajectory of the vehicle based on the current posture of the vehicle and a preset kinematic model if the deviation is greater than a preset deviation.

[0086] As a method, the first determination unit 320 is specifically used to input the current posture of the vehicle into the preset kinematic model to obtain the predicted driving trajectory of the vehicle output by the preset kinematic model, wherein the predicted driving trajectory is a set of trajectory points that the vehicle will travel within a preset time output by the preset kinematic model.

[0087] The second determining unit 330 is used to determine the predicted driving trajectory of the vehicle based on the projection point of the vehicle on the current tracking path and the set of waypoints of the current tracking path if the deviation is less than or equal to the preset deviation.

[0088] As a method, the second determination unit 330 is specifically used to calculate the distance between two adjacent waypoints in sequence from the projection point in the waypoint set to obtain multiple distances; calculate the distance sum of the multiple distances in sequence; if the distance sum is greater than a preset distance, use the distance sum and the multiple waypoints corresponding to it as the predicted driving trajectory of the vehicle.

[0089] The level acquisition unit 340 is used to acquire the collision risk level corresponding to the predicted driving trajectory.

[0090] As a method, the level acquisition unit 340 is specifically used to obtain the distance between the predicted driving trajectory and the obstacle; based on the distance, the collision risk level corresponding to the preset driving trajectory is obtained.

[0091] Furthermore, the level acquisition unit 340 is specifically used to obtain a grid map and high and low obstacle information of the vehicle's current environment; generate an obstacle distance map based on the grid map and the high and low obstacle information; obtain the minimum distance value of each trajectory point in the predicted driving trajectory from the obstacle distance map, and the minimum distance value is the distance between each trajectory point and its nearest obstacle; different minimum distance values ​​correspond to different collision risk levels; based on the minimum distance value of each trajectory point, determine the collision risk level corresponding to the predicted driving trajectory from a plurality of pre-set collision risk levels.

[0092] It should be noted that the device embodiment in the present application corresponds to the aforementioned method embodiment. The specific principles in the device embodiment can be found in the contents of the aforementioned method embodiment and will not be repeated here.

[0093] The following will be combined Figure 5 A vehicle provided in the present application is described.

[0094] See also Figure 5 Based on the above parking collision detection method and device, the embodiment of the present application also provides another vehicle 800 that can execute the above parking collision detection method. The vehicle 800 includes one or more (only one is shown in the figure) processors 802, a memory 804, a millimeter wave radar 806, and a network module 806 that are coupled to each other. Among them, the memory 804 stores a program that can execute the content of the above embodiment, and the processor 802 can execute the program stored in the memory 804.

[0095] Among them, the processor 802 may include one or more processing cores. The processor 802 uses various interfaces and lines to connect various parts of the entire vehicle 800, and executes various functions and processes data of the vehicle 800 by running or executing instructions, programs, code sets or instruction sets stored in the memory 804, and calling data stored in the memory 804. Optionally, the processor 802 can be implemented in at least one hardware form of digital signal processing (Digital Signal Processing, DSP), field programmable gate array (Field-Programmable Gate Array, FPGA), and programmable logic array (Programmable Logic Array, PLA). The processor 802 can integrate one or a combination of a central processing unit (Central Processing Unit, CPU), a graphics processing unit (Graphics Processing Unit, GPU) and a modem. Among them, the CPU mainly processes the operating system, user interface and application programs; the GPU is responsible for rendering and drawing display content; and the modem is used to process wireless communications. It can be understood that the above-mentioned modem may not be integrated into the processor 802, but may be implemented separately through a communication chip.

[0096] The memory 804 may include a random access memory (RAM) or a read-only memory (ROM). The memory 804 may be used to store instructions, programs, codes, code sets or instruction sets. The memory 804 may include a program storage area and a data storage area, wherein the program storage area may store instructions for implementing an operating system, instructions for implementing at least one function (such as a touch function, a sound playback function, an image playback function, etc.), instructions for implementing the following various method embodiments, etc. The data storage area may also store data created by the vehicle 800 during use (such as a phone book, audio and video data, chat record data), etc.

[0097] The network module 806 is used to receive and send electromagnetic waves, realize the mutual conversion between electromagnetic waves and electrical signals, and thus communicate with a communication network or other devices, such as communicating with a vehicle. The network module 806 may include various existing circuit elements for performing these functions, such as antennas, radio frequency transceivers, digital signal processors, encryption / decryption chips, user identity modules (SIM) cards, memories, etc. The network module 806 may communicate with various networks such as the Internet, corporate intranets, wireless networks, or communicate with other devices via wireless networks. The above-mentioned wireless networks may include cellular telephone networks, wireless local area networks, or metropolitan area networks. For example, the network module 806 may exchange information with a base station.

[0098] Please refer to Figure 6 , which shows a structural block diagram of a computer-readable storage medium provided in an embodiment of the present application. The computer-readable storage medium 900 stores program codes, which can be called by a processor to execute the method described in the above method embodiment.

[0099] The computer-readable storage medium 900 may be an electronic memory such as a flash memory, an EEPROM (electrically erasable programmable read-only memory), an EPROM, a hard disk, or a ROM. Optionally, the computer-readable storage medium 900 includes a non-transitory computer-readable storage medium. The computer-readable storage medium 900 has storage space for program code 910 that performs any method step in the above method. These program codes can be read from or written to one or more computer program products. The program code 910 can be compressed, for example, in an appropriate form.

[0100] The present application provides a parking collision detection method, device, vehicle and storage medium, which first obtains the deviation between the vehicle and the current tracking path. If the deviation is greater than the preset deviation, the predicted driving trajectory of the vehicle is determined based on the current position and preset kinematic model of the vehicle. If the deviation is less than or equal to the preset deviation, the predicted driving trajectory of the vehicle is determined based on the projection point of the vehicle on the current tracking path and the set of waypoints of the current tracking path, and then the collision risk level corresponding to the predicted driving trajectory is obtained. Through the above method, by predicting the driving trajectory and performing collision detection, the system is allowed to react before the vehicle approaches an obstacle, avoiding possible hard braking in an emergency and reducing the risk of accidental collision.

[0101] The embodiments of the present invention are described above in conjunction with the accompanying drawings, but the present invention is not limited to the above-mentioned specific implementation modes, which are merely illustrative rather than restrictive. Under the guidance of the present invention, ordinary technicians in this field can also make many forms without departing from the scope of protection of the present invention and the claims, all of which are within the protection of the present invention.

Claims

1. A parking collision detection method, characterized in that: The method comprises: Get the deviation between the vehicle and the current tracking path; If the deviation is greater than a preset deviation, determining a predicted driving trajectory of the vehicle based on the current posture of the vehicle and a preset kinematic model; If the deviation is less than or equal to the preset deviation, determining a predicted driving trajectory of the vehicle based on a projection point of the vehicle on the current tracking path and a set of waypoints of the current tracking path; Obtain a collision risk level corresponding to the predicted driving trajectory.

2. The method according to claim 1, characterized in that The step of obtaining the deviation between the vehicle and the current tracking path includes: Obtaining a projection point of the vehicle on the current tracking path; Obtaining a difference vector between the projection point and the current pose; The difference vector is projected onto the normal vector of the projection point to determine the deviation between the vehicle and the current tracking path.

3. The method according to claim 2, characterized in that The obtaining of the projection point of the vehicle on the current tracking path comprises: Obtaining a set of waypoints of a current tracking path and a current position and posture of the vehicle; Obtaining the Euclidean distance between the current position and each waypoint in the waypoint set; The waypoint with the smallest corresponding Euclidean distance in the waypoint set is determined as the projection point of the vehicle on the current tracking path.

4. The method according to claim 1, characterized in that The step of determining a predicted driving trajectory of the vehicle based on the current posture of the vehicle and a preset kinematic model includes: The current posture of the vehicle is input into the preset kinematic model to obtain the predicted driving trajectory of the vehicle output by the preset kinematic model, wherein the predicted driving trajectory is a set of trajectory points that the vehicle will travel within a preset time output by the preset kinematic model.

5. The method according to claim 1, characterized in that The step of determining the predicted driving trajectory of the vehicle based on the projection point of the vehicle on the current tracking path and the set of waypoints of the current tracking path includes: Starting from the projection point in the waypoint set, the distances between two adjacent waypoints are calculated in sequence to obtain a plurality of distances; Calculating the distance sums of the multiple distances in sequence; If the sum of the distances is greater than a preset distance, the distances and the multiple waypoints corresponding to the distances are used as the predicted driving trajectory of the vehicle.

6. The method according to claim 1, characterized in that The obtaining of the collision risk level corresponding to the predicted driving trajectory includes: Obtaining the distance between the predicted driving trajectory and the obstacle; Based on the distance, a collision risk level corresponding to the preset driving trajectory is obtained.

7. The method according to claim 6, characterized in that The obtaining of the distance between the predicted driving trajectory and the obstacle includes: Obtaining a grid map and high and low obstacle information of the vehicle's current environment; Generate an obstacle distance map based on the grid map and the high and low obstacle information; Obtaining, from the obstacle distance map, a minimum distance value for each trajectory point in the predicted driving trajectory, the minimum distance value being the distance between each trajectory point and its nearest obstacle; different minimum distance values ​​corresponding to different collision risk levels; The acquiring, based on the distance, a collision risk level corresponding to the predicted driving trajectory includes: Based on the minimum distance value of each trajectory point, a collision risk level corresponding to the predicted driving trajectory is determined from a plurality of pre-set collision risk levels.

8. A parking collision detection device, characterized in that: The device comprises: A deviation acquisition unit, used to acquire the deviation between the vehicle and the current tracking path; A first determining unit, configured to determine a predicted driving trajectory of the vehicle based on a current posture of the vehicle and a preset kinematic model if the deviation is greater than a preset deviation; a second determining unit, configured to determine a predicted driving trajectory of the vehicle based on a projection point of the vehicle on the current tracking path and a set of waypoints of the current tracking path if the deviation is less than or equal to the preset deviation; The level acquisition unit is used to acquire the collision risk level corresponding to the predicted driving trajectory.

9. A vehicle, characterized in that: The method comprises one or more processors and a memory; one or more programs are stored in the memory and are configured so that the one or more processors execute the method according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores program code, wherein when the program code is executed by a processor, the method according to any one of claims 1 to 7 is executed.

Images