A collision prediction method, device, medium and electronic device for a vehicle

Abstract

An embodiment of the present application provides a collision prediction method, device, medium, and electronic device for a vehicle. This method can be applied to scenarios such as ports, border crossings, mines, highways, or urban traffic. First, vehicle speed information, basic steering angle control information, and obstacle coordinate information are obtained. Then, based on the target steering angle control sequence information and vehicle speed information, the predicted pose of the vehicle within the target time domain is determined, and the target steering angle control sequence information is obtained based on the basic steering angle control information; Next, according to the predicted pose and vehicle body size information, the vehicle contour information within the target time domain is determined, and based on the vehicle contour information and obstacle coordinate information, the target collision prediction result of the vehicle is determined, and the target collision prediction result indicates whether the vehicle collides with an obstacle within the target time domain. This method can reduce the impact of positioning failure or abnormality on the collision prediction result and improve the accuracy of collision prediction.

Description

Technical Field

[0001] The embodiments of the present application relate to the field of autonomous driving technology, and in particular, to a collision prediction method, device, medium and electronic device for a vehicle. Background Technique

[0002] With the continuous development of autonomous driving technology, autonomous driving vehicles have attracted more and more attention. An autonomous driving vehicle is an intelligent vehicle that senses the road environment through an in-vehicle sensing system, automatically plans a driving route and controls the vehicle to reach a predetermined destination, and its functions are realized by an Autonomous Driving System (ADS).

[0003] The system safety of autonomous driving vehicles is the most important link in autonomous driving vehicle control technology. If safety cannot be guaranteed, driverless vehicles cannot be put into use. To improve the system safety of autonomous driving vehicles, Autonomous Emergency Braking (AEB) is usually adopted. AEB refers to the situation where when the driving environment around the vehicle changes, resulting in a possible rear-end collision or collision, endangering the driver, passengers and pedestrians, the AEB state will brake with a large deceleration, thereby avoiding or reducing the occurrence of traffic accidents.

[0004] However, in the related art, the collision prediction of autonomous emergency braking usually depends on positioning or path. When the positioning suddenly deviates or the tracking deviates from the path, anomalies or failures often occur, the collision prediction is inaccurate, and the vehicle safety is reduced. Summary of the Invention

[0005] The embodiments of the present application provide a collision prediction method, device, medium and electronic device for a vehicle, which can perform collision prediction without relying on positioning results, reduce the influence of positioning failure or anomaly on the collision prediction result, and improve the accuracy of collision prediction.

[0006] In a first aspect, the embodiments of the present application provide a collision prediction method for a vehicle, including:

[0007] Obtain the vehicle speed information, basic corner control information and obstacle coordinate information of the vehicle;

[0008] Based on the target corner control sequence information and the vehicle speed information, determine the predicted pose of the vehicle within the target time domain; the target corner control sequence information is obtained based on the basic corner control information; the target corner control sequence information includes an array of steering axis angles corresponding to multiple control moments arranged in chronological order within the target time domain;

[0009] Determine the vehicle contour information of the vehicle within the target time domain according to the predicted pose and the vehicle body size information;

[0010] Determine the target collision prediction result of the vehicle according to the vehicle contour information and the obstacle coordinate information; the target collision prediction result characterizes whether the vehicle collides with the obstacle within the target time domain.

[0011] The vehicle collision prediction method provided by the embodiments of the present application first obtains the vehicle speed information, the basic corner control information and the obstacle coordinate information of the vehicle, and then determines the predicted pose of the vehicle within the target time domain based on the target corner control sequence information and the vehicle speed information. The target corner control sequence information is obtained based on the basic corner control information. The target corner control sequence information includes an array of steering shaft angles corresponding to multiple control moments arranged in chronological order within the target time domain. Then, according to the predicted pose and the vehicle body size information, determine the vehicle contour information of the vehicle within the target time domain, and according to the vehicle contour information and the obstacle coordinate information, determine the target collision prediction result of the vehicle. The target collision prediction result characterizes whether the vehicle collides with the obstacle within the target time domain. This method does not depend on the positioning result, only relates to the possible motion trajectories within the future target time domain in the current state and the obstacles sensed by the vehicle, reduces the impact of positioning failure or abnormality on the collision prediction result, and improves the accuracy of collision prediction.

[0012] In an alternative embodiment, before determining the predicted pose of the vehicle within the target time domain based on the target corner control sequence information and the vehicle speed information, the method further includes:

[0013] Use the corner control sequence of the model predictive control (MPC) in the obtained basic corner control information as the target corner control sequence information.

[0014] In this embodiment, by using the corner control sequence of the model predictive control (MPC) in the obtained basic corner control information as the target corner control sequence information, a method for recursively predicting the pose of an autonomous vehicle within the target time domain based on the corner control sequence of MPC and performing collision prediction is provided. It can improve the trajectory prediction accuracy within the target time domain, does not depend on the positioning result, only relates to the possible motion trajectories within the future target time domain in the current state and the obstacles sensed by the vehicle, reduces the impact of positioning failure or abnormality on the collision prediction result, and further improves the accuracy of collision prediction.

[0015] In an alternative embodiment, the basic cornering control information further includes the vehicle wheel angle information of the vehicle; after using the cornering control sequence of the model predictive control (MPC) in the basic cornering control information as the target cornering control sequence information, before determining the predicted pose of the vehicle in the target time domain based on the target cornering control sequence information and the vehicle speed information, it further includes:

[0016] If the cornering control sequence of the MPC is not included in the basic cornering control information, then use the vehicle wheel angle information as the value of each set of steering axis angle arrays in the target cornering control sequence information.

[0017] In this embodiment, when the cornering control sequence of the MPC is not included in the basic cornering control information, by using the vehicle wheel angle information of the vehicle included in the basic cornering control information as the value of each set of steering axis angle arrays in the target cornering control sequence information, a mechanism is provided to determine the predicted pose of the autonomous vehicle in the target time domain based on the cornering control sequence of the MPC or the vehicle wheel angle and perform collision prediction, improving the accuracy of collision prediction and enhancing the safety of the vehicle at the same time.

[0018] In an alternative embodiment, the obstacle coordinate information is the coordinate of the obstacle in the vehicle body coordinate system at the moment when the obstacle coordinate information is obtained; the obstacle coordinate information is obtained by an in-vehicle sensor.

[0019] In this embodiment, by obtaining the coordinate of the obstacle in the vehicle body coordinate system at the moment when the obstacle coordinate information is obtained through the in-vehicle sensor, the obstacles sensed by the vehicle can be determined accurately and efficiently, improving the accuracy of collision prediction.

[0020] In an alternative embodiment, the predicted pose includes the abscissa of the predicted vehicle body position, the ordinate of the predicted vehicle body position, and the heading angle eigenvalue in the vehicle body coordinate system at the moment when the obstacle coordinate information is obtained; the heading angle eigenvalue represents the change value of the heading angle at the predicted vehicle body position at any control moment compared with the heading angle at the moment when the obstacle coordinate information is obtained.

[0021] In this embodiment, the predicted pose includes the abscissa of the predicted vehicle position, the ordinate of the predicted vehicle position, and the heading angle eigenvalue in the vehicle body coordinate system at the moment when the obstacle coordinate information is obtained; the heading angle eigenvalue represents the change value of the heading angle at the predicted vehicle position at any control moment compared with the heading angle at the moment when the obstacle coordinate information is obtained. This method can directly calculate the abscissa of the predicted vehicle position, the ordinate of the predicted vehicle position, and the heading angle eigenvalue in the vehicle body coordinate system; the heading angle eigenvalue represents the change value of the heading angle at the predicted vehicle position at any control moment compared with the heading angle at the moment when the obstacle coordinate information is obtained. This collision prediction process does not depend on the positioning result and is only related to the possible motion trajectories within the future target time domain in the current state and the obstacles sensed by the vehicle. By only analyzing the coordinates and heading angle changes in the vehicle body coordinate system, the result of collision prediction affected by positioning failure or abnormality is reduced, and the accuracy of collision prediction is further improved.

[0022] In an alternative embodiment, the steering axle of the vehicle includes a front steering axle and a rear steering axle; the steering axle angle array includes the front steering axle angle and the rear steering axle angle; determining the predicted pose of the vehicle within the target time domain based on the target steering angle control sequence information and the vehicle speed information includes:

[0023] Obtain the target steering angle array at the target control moment in the target steering angle control sequence information one by one in chronological order, and successively perform the following steps on the obtained target steering angle array at the target control moment to obtain the predicted pose of the vehicle at each target control moment:

[0024] Calculate the target sideslip angle of the center of mass at the target control moment according to the obtained target steering angle array and the preset first kingpin offset parameter and second kingpin offset parameter; the first kingpin offset parameter is the distance from the center of the front steering axle to the reference center of the vehicle body; the second kingpin offset parameter is the distance from the center of the rear steering axle to the reference center of the vehicle body; the reference center of the vehicle body is the center of the vehicle body.

[0025] Obtain the target heading angle change rate at the target control moment according to the target sideslip angle of the center of mass, the target steering angle array, the first kingpin offset parameter, and the second kingpin offset parameter;

[0026] Based on the target heading angle change rate determined within the target time domain, obtain the heading angle eigenvalue at the target control moment as the target heading angle eigenvalue;

[0027] Based on the target centroid sideslip angle, the target heading angle eigenvalue, the prediction interval threshold, and the vehicle speed information at the current moment, obtain the abscissa of the predicted vehicle position and the ordinate of the predicted vehicle position at the target control moment; the prediction interval threshold is the time interval between any two adjacent control moments in the target steering angle control sequence information.

[0028] In this embodiment, a mechanism is provided for determining the abscissa, ordinate, and heading angle change value of a vehicle with double steering axles in a vehicle body coordinate system within a target time domain based on target steering angle control sequence information and the vehicle speed information. This collision prediction process can accurately and efficiently recursively calculate the vehicle pose within the target time domain, without relying on the positioning result, only related to the possible motion trajectories within the future target time domain in the current state and the obstacles sensed by the vehicle, reducing the impact of positioning failure or abnormality on the collision prediction result and improving the accuracy of collision prediction.

[0029] In an alternative embodiment, according to the predicted pose and the vehicle body size information, determine the vehicle profile information of the vehicle within the target time domain, including:

[0030] Based on the abscissa of the predicted vehicle position, the ordinate of the predicted vehicle position, the abscissa of the vehicle body reference center, and the ordinate of the vehicle body reference center, obtain the abscissa of the predicted vehicle body reference center and the ordinate of the predicted vehicle body reference center at each control moment;

[0031] Based on the heading angle eigenvalue, determine the predicted vehicle body direction information at each control moment; the predicted vehicle body direction information represents the angle between the predicted vehicle body direction at the control moment and the vehicle body direction at the moment when the obstacle coordinate information is obtained;

[0032] Based on the abscissa of the predicted vehicle body reference center, the ordinate of the predicted vehicle body reference center, the predicted vehicle body direction information, and the vehicle body size information at each control moment, obtain the vehicle profile information of the vehicle within the target time domain.

[0033] In this embodiment, the abscissa of the predicted vehicle position, the ordinate of the predicted vehicle position, and the heading angle eigenvalue in the vehicle body coordinate system can be directly recursively calculated; the heading angle eigenvalue represents the change value of the heading angle at the predicted vehicle position at any control moment compared with the heading angle at the moment when the obstacle coordinate information is obtained. This collision prediction process does not rely on the positioning result. By using the abscissa of the predicted vehicle position, the ordinate of the predicted vehicle position, and the heading angle eigenvalue in the vehicle body coordinate system, determine the vehicle profile information of the vehicle within the target time domain, reduce the computational amount in the collision prediction process, shorten the prediction time, and improve the accuracy and prediction efficiency of collision prediction.

[0034] Second aspect, an embodiment of the present application further provides a collision prediction device for a vehicle, including:

[0035] An information acquisition unit, configured to acquire vehicle speed information, basic corner control information, and obstacle coordinate information of the vehicle;

[0036] A pose prediction unit, configured to determine a predicted pose of the vehicle within a target time domain based on target corner control sequence information and the vehicle speed information; the target corner control sequence information is obtained based on the basic corner control information; the target corner control sequence information includes an array of steering axis angles corresponding one by one to multiple control moments arranged in chronological order within the target time domain;

[0037] A contour determination unit, configured to determine vehicle contour information of the vehicle within the target time domain according to the predicted pose and body size information;

[0038] A collision detection unit, configured to determine a target collision prediction result of the vehicle according to the vehicle contour information and the obstacle coordinate information; the target collision prediction result indicates whether the vehicle collides with the obstacle within the target time domain.

[0039] In an optional embodiment, the pose prediction unit is further configured to:

[0040] Use the corner control sequence of model predictive control (MPC) in the obtained basic corner control information as the target corner control sequence information.

[0041] In an optional embodiment, the basic corner control information further includes vehicle wheel angle information of the vehicle; the pose prediction unit is further configured to:

[0042] If the corner control sequence of the MPC is not included in the basic corner control information, use the vehicle wheel angle information as the value of each group of the steering axis angle arrays in the target corner control sequence information.

[0043] In an optional embodiment, the obstacle coordinate information is the coordinate of the obstacle in the vehicle body coordinate system at the moment when the obstacle coordinate information is acquired; the obstacle coordinate information is acquired by an in-vehicle sensor.

[0044] In an optional embodiment, the predicted pose includes the abscissa of the predicted vehicle body position, the ordinate of the predicted vehicle body position, and the heading angle eigenvalue in the vehicle body coordinate system at the moment when the obstacle coordinate information is acquired; the heading angle eigenvalue indicates the change value of the heading angle at the predicted vehicle body position at any control moment compared with the heading angle at the moment when the obstacle coordinate information is acquired.

[0045] In an alternative embodiment, the steering axle of the vehicle includes a front steering axle and a rear steering axle; the steering axle angle array includes a front steering axle angle and a rear steering axle angle; the pose prediction unit is specifically configured to:

[0046] Obtain the target steering angle array at the target control moment in the target corner control sequence information one by one in chronological order, and successively perform the following steps on the obtained target steering angle array at the target control moment to obtain the predicted pose of the vehicle at each target control moment:

[0047] According to the obtained target steering angle array, the preset first axle center distance parameter and the second axle center distance parameter, calculate the target centroid side slip angle at the target control moment; the first axle center distance parameter is the distance from the center of the front steering axle to the vehicle body reference center; the second axle center distance parameter is the distance from the center of the rear steering axle to the vehicle body reference center; the vehicle body reference center is the center of the vehicle body;

[0048] According to the target centroid side slip angle, the target steering angle array, the first axle center distance parameter and the second axle center distance parameter, obtain the target heading angle change rate at the target control moment;

[0049] Based on the target heading angle change rate determined within the target time domain, obtain the heading angle eigenvalue at the target control moment as the target heading angle eigenvalue;

[0050] According to the target centroid side slip angle, the target heading angle eigenvalue, the prediction interval threshold and the vehicle speed information at the current moment, obtain the predicted horizontal coordinate of the vehicle body and the predicted vertical coordinate of the vehicle body at the target control moment; the prediction interval threshold is the time interval between any two adjacent control moments in the target corner control sequence information.

[0051] In an alternative embodiment, the contour determination unit is specifically configured to:

[0052] According to the predicted horizontal coordinate of the vehicle body, the predicted vertical coordinate of the vehicle body, the horizontal coordinate of the vehicle body reference center, and the vertical coordinate of the vehicle body reference center, obtain the predicted horizontal coordinate of the vehicle body reference center and the predicted vertical coordinate of the vehicle body reference center at each control moment;

[0053] According to the heading angle eigenvalue, determine the predicted vehicle body direction information at each control moment; the predicted vehicle body direction information represents the angle between the predicted vehicle body direction at the control moment and the vehicle body direction at the moment when the obstacle coordinate information is obtained;

[0054] Based on the abscissa of the predicted reference center of the vehicle body, the ordinate of the predicted reference center of the vehicle body, the predicted direction information of the vehicle body, and the vehicle body size information at each of the control times, the vehicle contour information of the vehicle within the target time domain is obtained.

[0055] In a third aspect, an embodiment of the present application further provides a computer-readable storage medium, in which a computer program is stored. When the computer program is executed by a processor, the vehicle collision prediction method in the first aspect is implemented.

[0056] In a fourth aspect, an embodiment of the present application further provides an electronic device, including a memory and a processor. A computer program that can run on the processor is stored on the memory. When the computer program is executed by the processor, the vehicle collision prediction method in the first aspect is implemented.

[0057] In a fifth aspect, an embodiment of the present application further provides a computer program product, which includes computer instructions. When the computer instructions are executed by a computing device, the computing device can execute the vehicle collision prediction method as described in any one of the first aspect.

[0058] For the technical effects brought by any implementation manner in the second aspect to the fifth aspect, reference can be made to the corresponding implementation manner in the first aspect, which will not be elaborated here. Description of the Drawings

[0059] Figure 1 It is a flowchart of a vehicle collision prediction method provided by an embodiment of the present application;

[0060] Figure 2 It is a flowchart of determining the predicted pose of a vehicle collision prediction method provided by an embodiment of the present application;

[0061] Figure 3 It is a flowchart of determining the vehicle contour information of a vehicle collision prediction method provided by an embodiment of the present application;

[0062] Figure 4 It is a schematic diagram of the vehicle contour information of a vehicle collision prediction method provided by an embodiment of the present application;

[0063] Figure 5 It is a schematic diagram of the structure of a vehicle collision prediction device provided by an embodiment of the present application;

[0064] Figure 6 It is a schematic diagram of the structure of an electronic device provided by an embodiment of the present application. Detailed Embodiments

[0065] To make the objectives, technical solutions and advantages of this application clearer, the following will further describe this application in detail with reference to the accompanying drawings. Obviously, the described embodiments are only a part rather than all of the embodiments of this application. All other embodiments obtained by those of ordinary skill in the art based on the embodiments in this application without creative efforts shall fall within the scope of protection of this application.

[0066] It should be noted that the terms "including" and "having" and their variations involved in the documents of this application are intended to cover non-exclusive inclusion. For example, a process, method, system, product or device that includes a series of steps or units does not necessarily limit to those steps or units clearly listed, but may include other steps or units not clearly listed or inherent to these processes, methods, products or devices.

[0067] The following explains some terms appearing in the text:

[0068] (1) Autonomous Emergency Braking (AEB): AEB refers to the situation where when the driving environment around the vehicle changes, which may lead to a rear-end collision or a crash, endangering the driver, passengers and pedestrians. In the AEB state, the vehicle will brake with a relatively large deceleration to avoid or mitigate the occurrence of a car accident.

[0069] (2) MPC (Model Predictive Control): MPC is a model-based closed-loop optimal control. Its main idea is: based on the current state and input of the model, as well as the constructed system model, to predict the system output within the next N sampling periods. Then, using the predicted output and the constraint conditions designed according to the actual situation and control requirements of the system to obtain the local optimal solution of the cost function, and get the input of the next control point of the system. In this way, a solution is obtained at each sampling period, and finally the system reaches the control target.

[0070] (3) Vehicle body coordinate system: The vehicle body coordinate system is also known as the vehicle coordinate system or the ego-vehicle coordinate system. The vehicle coordinate system is a special moving coordinate system used to describe the motion of an automobile. There are various ways to define the vehicle body coordinate system. Currently, there are several commonly used definitions of the vehicle body coordinate system in the academic and industrial circles, namely the international standard ISO definition, the Society of Automotive Engineers (SAE) definition, and the coordinate definition based on the Inertial Measurement Unit (IMU). For example, taking a dual-steering-axis autonomous vehicle as an example, the vehicle body coordinate system takes the center position of the rear steering axis as the origin of the vehicle body coordinate system. The X-axis points to the right along the axis of the dual-steering-axis autonomous vehicle, the Y-axis is parallel to the ground and points to the front of the vehicle, and the Z-axis passes through the origin of the vehicle body coordinate system and points to the sky. In the embodiments of the present application, since the collision prediction method of the vehicle is a motion analysis in a plane, the Z-axis coordinate of the vehicle body coordinate system can be not considered at this time, and only the X-axis coordinate and the Y-axis coordinate are considered.

[0071] With the continuous development of autonomous driving technology, autonomous vehicles have attracted more and more attention. An autonomous vehicle is an intelligent vehicle that senses the road environment through an on-vehicle sensing system, automatically plans a driving route, and controls the vehicle to reach a predetermined destination. It relies on an Autonomous Driving System (ADS) to achieve its functions.

[0072] The system safety of autonomous vehicles is the most important link in the control technology of autonomous vehicles. If safety cannot be guaranteed, driverless vehicles cannot be put into use. To improve the system safety of autonomous vehicles, Autonomous Emergency Braking (AEB) is usually adopted. AEB refers to the situation where when the driving environment around the vehicle changes, it may lead to a rear-end collision or a collision, endangering the driver, passengers, and pedestrians. In the AEB state, the vehicle will brake with a large deceleration, thereby avoiding or reducing the occurrence of traffic accidents.

[0073] However, in related technologies, the collision prediction of autonomous emergency braking usually relies on positioning or paths. When the positioning suddenly deviates or the tracking deviates from the path, anomalies or failures often occur, the collision prediction is inaccurate, and the vehicle safety is reduced.

[0074] To solve the above problems, an embodiment of the present application provides a collision prediction method for a vehicle. First, vehicle speed information, basic steering angle control information, and obstacle coordinate information are obtained. Then, based on the target steering angle control sequence information and the vehicle speed information, the predicted pose of the vehicle within the target time domain is determined. The target steering angle control sequence information is obtained based on the basic steering angle control information. The target steering angle control sequence information includes an array of steering shaft angles corresponding one-to-one with multiple control moments arranged in chronological order within the target time domain. Then, according to the predicted pose and the vehicle body size information, the vehicle contour information within the target time domain is determined, and based on the vehicle contour information and the obstacle coordinate information, the target collision prediction result of the vehicle is determined. The target collision prediction result indicates whether the vehicle collides with an obstacle within the target time domain. This method does not depend on the positioning result and is only related to the possible motion trajectories within the future target time domain in the current state and the obstacles perceived by the vehicle, reducing the impact of positioning failure or abnormality on the collision prediction result and improving the accuracy of collision prediction.

[0075] The collision prediction method for a vehicle provided by an embodiment of the present application is applicable to low-speed autonomous vehicles. Among them, low-speed operation is usually below 35 Km / h. This method can be applied to scenarios such as ports, ports, mines, highways, or urban traffic.

[0076] Next, the technical solution provided by the embodiment of the present application will be described in detail with reference to the accompanying drawings.

[0077] Figure 1 The flowchart of a collision prediction method for a vehicle provided by an embodiment of the present application is shown. As Figure 1 shown, the method may include the following steps:

[0078] Step S101, obtain the vehicle speed information, basic steering angle control information, and obstacle coordinate information of the vehicle.

[0079] Specifically, the vehicle is a low-speed autonomous vehicle running at a low speed. In some embodiments of the present application, the vehicle is a dual-steering-axis autonomous vehicle. It should be noted that the vehicle in the present application may also be other types of low-speed autonomous vehicles, such as single-steering-axis autonomous vehicles. The vehicle speed information is the vehicle's center-of-mass speed. The basic steering angle control information is the information used to control the steering angle of the vehicle.

[0080] Exemplarily, obtain the vehicle speed information Speed, basic steering angle control information Basic_Ang_C, and obstacle coordinate information obstacle_Inf of the autonomous vehicle.

[0081] When the speed of the autonomous vehicle does not exceed 35 Km / h, the collision prediction method according to the embodiments of the present application can improve the vehicle safety to ensure the safety to the greatest extent when the positioning suddenly deviates or the tracking deviates from the path.

[0082] In some embodiments of the present application, the obstacle coordinate information is the coordinate of the obstacle in the vehicle body coordinate system at the moment when the obstacle coordinate information is obtained; the obstacle coordinate information is obtained by an in-vehicle sensor.

[0083] Specifically, the in-vehicle sensor may be a lidar. The lidar can measure the coordinate of the obstacle in the vehicle body coordinate system at the moment when the obstacle coordinate information is obtained.

[0084] Exemplarily, assume that the moment when the obstacle coordinate information is obtained is moment Tn, and the vehicle body coordinate system at moment Tn is vehicle body coordinate system OXY. The obstacle coordinate information obstacle_Inf is the coordinate of obstacle obstacle1 in the vehicle body coordinate system OXY at moment Tn. Assume that the obstacle coordinate information obstacle_Inf is the coordinate of a point (Xp, Yp); the obstacle coordinate information (Xp, Yp) can be obtained by the lidar.

[0085] Step S102, determine the predicted pose of the vehicle in the target time domain based on the target corner control sequence information and the vehicle speed information.

[0086] Among them, the target corner control sequence information is obtained based on the basic corner control information; the target corner control sequence information includes an array of steering axis angles corresponding to multiple control moments arranged in chronological order in the target time domain.

[0087] Specifically, the target time domain may be a time range determined by a preset duration threshold.

[0088] In some embodiments of the present application, the duration threshold is taken as 2s to 3s. Assume that the duration threshold is 2s. For moment Tn, the value of the target time domain Tar_T_Sc is a time range with a duration of 2s starting from moment Tn. For another moment Tm, the value of the target time domain Tar_T_Sc is a time range with a duration of 2s starting from moment Tm.

[0089] Exemplarily, based on the target corner control sequence information Target_Ang_C and the vehicle speed information Speed, the predicted pose Po_Or of the autonomous vehicle within the target time domain Tar_T_Sc is determined. Among them, the target time domain Tar_T_Sc is a time range of 2 s starting from the moment Tn, and the target corner control sequence information Target_Ang_C is obtained based on the basic corner control information Basic_Ang_C; the target corner control sequence information Target_Ang_C includes a steering axis angle array corresponding one-to-one to 10 control moments (Po_C_1, Po_C_2, Po_C_3, Po_C_4, Po_C_5, Po_C_6, Po_C_7, Po_C_8, Po_C_9, Po_C_10) arranged in chronological order within the target time domain Tar_T_Sc. Assume that the steering axis angle arrays corresponding one-to-one to the 10 control moments (Po_C_1, Po_C_2, Po_C_3, Po_C_4, Po_C_5, Po_C_6, Po_C_7, Po_C_8, Po_C_9, Po_C_10) included in the target corner control sequence information Target_Ang_C are Angle_Data_1, Angle_Data_2, Angle_Data_3, Angle_Data_4, Angle_Data_5, Angle_Data_6, Angle_Data_7, Angle_Data_8, Angle_Data_9, and Angle_Data_10 respectively. Among them, the value of the control moment Po_C_1 can be the moment Tn.

[0090] In the embodiments of the present application, the basic corner control information can be a corner control sequence and / or vehicle wheel angle information.

[0091] In some embodiments of the present application, the basic corner control information includes the corner control sequence of the model predictive control MPC. First, the corner control sequence of the model predictive control MPC in the obtained basic corner control information is used as the target corner control sequence information, and then based on the obtained target corner control sequence information and the vehicle speed information, the predicted pose of the vehicle within the target time domain is determined.

[0092] In this embodiment, the basic corner control information includes the corner control sequence of the model predictive control (MPC). Exemplarily, it is assumed that the basic corner control information Basic_Ang_C at time Tn includes the corner control sequence Cmd_Sequence_1 of the model predictive control (MPC). Before determining the predicted pose Po_Or of the autonomous vehicle within the target time domain Tar_T_Sc based on the target corner control sequence information Target_Ang_C and the vehicle speed information Speed, the corner control sequence Cmd_Sequence_1 of the model predictive control (MPC) in the obtained basic corner control information Basic_Ang_C is used as the target corner control sequence information Target_Ang_C. Among them, the target time domain Tar_T_Sc is a time range with a duration of 2 s starting from time Tn.

[0093] In some embodiments of the present application, the basic corner control information further includes the vehicle wheel angle information of the vehicle; after using the corner control sequence of the model predictive control (MPC) in the basic corner control information as the target corner control sequence information, if the basic corner control information does not include the corner control sequence of the MPC, the vehicle wheel angle information is used as the value of each set of steering axis angle arrays in the target corner control sequence information. Then, based on this target corner control sequence information and the vehicle speed information, the predicted pose of the vehicle within the target time domain is determined.

[0094] In this embodiment, the basic corner control information includes the corner control sequence of the model predictive control (MPC) and the vehicle wheel angle information of the vehicle. Exemplarily, after performing the step of using the corner control sequence of the model predictive control (MPC) in the basic corner control information as the target corner control sequence information, if it is determined that the basic corner control information Basic_Ang_C at time Tn does not include the corner control sequence of the MPC but only includes the vehicle wheel angle information Wheel_Angle_1 of the autonomous vehicle, the vehicle wheel angle information Wheel_Angle_1 is used as the value of each set of steering axis angle arrays in the target corner control sequence information Target_Ang_C, that is, the values of Angle_Data_1, Angle_Data_2, Angle_Data_3, Angle_Data_4, Angle_Data_5, Angle_Data_6, Angle_Data_7, Angle_Data_8, Angle_Data_9, and Angle_Data_10 are all the vehicle wheel angle information Wheel_Angle_1. Then, based on the target corner control sequence information Target_Ang_C and the vehicle speed information Speed, the predicted pose Po_Or of the autonomous vehicle within the target time domain Tar_T_Sc is determined.

[0095] In some embodiments of the present application, the predicted pose includes the abscissa of the predicted vehicle position, the ordinate of the predicted vehicle position, and the heading angle eigenvalue in the vehicle body coordinate system at the moment when the obstacle coordinate information is obtained; the heading angle eigenvalue represents the change value of the heading angle at the predicted vehicle position at any control moment compared with the heading angle at the moment when the obstacle coordinate information is obtained.

[0096] Exemplarily, the predicted pose Po_Or can be (X, Y, ψ), where X and Y are respectively the abscissa of the predicted vehicle position and the ordinate of the predicted vehicle position in the vehicle body coordinate system at time Tn, and ψ is the heading angle eigenvalue in the vehicle body coordinate system at time Tn; the heading angle eigenvalue ψ represents the change value of the heading angle at the predicted vehicle position at any control moment compared with the heading angle at time Tn.

[0097] In the embodiments of the present application, at any moment Ti, if the vehicle speed information, the basic steering angle control information, and the obstacle coordinate information of the autonomous vehicle are obtained, and the predicted pose of the autonomous vehicle within the target time domain starting from time Ti is determined, then setting the pose of the autonomous vehicle at time Ti to (0, 0, 0) can reduce the computational amount for determining the predicted pose.

[0098] To more clearly illustrate the technical solution of the present application, taking a dual-steering-axis autonomous vehicle as an example, the process of determining the predicted pose of the vehicle within the target time domain will be introduced in detail below.

[0099] For a dual-steering-axis autonomous vehicle, the steering axes of the vehicle include a front steering axis and a rear steering axis; the steering axis angle array includes the front steering axis angle and the rear steering axis angle.

[0100] Figure 2 It is a schematic flow chart of determining the predicted pose of a vehicle collision prediction method provided in the embodiments of the present application. The steering axes of the vehicle include a front steering axis and a rear steering axis; the steering axis angle array includes the front steering axis angle and the rear steering axis angle; in the process of determining the predicted pose of the vehicle within the target time domain based on the target steering angle control sequence information and the vehicle speed information, in order to obtain the predicted pose of the vehicle at each target control moment, the target steering angle array at the target control moment in the target steering angle control sequence information can be obtained one by one in chronological order, and the obtained target steering angle array at the target control moment is sequentially processed as follows Figure 2 As shown, it is achieved through the following steps:

[0101] Step S201, calculate the target centroid side slip angle at the target control moment according to the obtained target steering angle array and the preset first pivot distance parameter and second pivot distance parameter.

[0102] Among them, the first axis distance parameter is the distance from the center of the front steering axle to the vehicle body reference center; the second axis distance parameter is the distance from the center of the rear steering axle to the vehicle body reference center.

[0103] In specific implementation, the vehicle body reference center is the center of the vehicle body.

[0104] According to the obtained target steering angle array and the preset first axis distance parameter and second axis distance parameter, calculate the target centroid side slip angle at the target control moment. Specifically, according to the front steering axle angle and rear steering axle angle included in the obtained target steering angle array and the preset first axis distance parameter and second axis distance parameter, calculate the target centroid side slip angle at the target control moment through the following formula:

[0105]

[0106] Among them, β is the centroid side slip angle, representing the included angle between the speed direction and the vehicle body direction;

[0107] l f is the distance from the center of the front steering axle to the vehicle body reference center;

[0108] l r is the distance from the center of the rear steering axle to the vehicle body reference center;

[0109] σ r is the rear steering axle angle;

[0110] σ f is the front steering axle angle.

[0111] Exemplarily, assume that the vehicle body of the autonomous driving vehicle is a regular rectangle in the vehicle body coordinate system OXY, and the vehicle body reference center can be the center M of the rectangle. The first axis distance parameter l f is the distance from the center of the front steering axle to the vehicle body reference center M; the second axis distance parameter l r is the distance from the center of the rear steering axle to the vehicle body reference center M. In order to obtain the predicted pose of the autonomous driving vehicle at each target control moment, the target steering angle array at the target control moment in the target corner control sequence information Target_Ang_C can be obtained one by one in chronological order, and the obtained target steering angle array at the target control moment is processed in sequence. Taking the target control moment as Po_C_2 and the obtained target steering angle array as Po_C_2 as an example, the target steering angle array Po_C_2 can include the front steering axle angle σ f_2 and the rear steering axle angle σ r_2 , then first calculate the target centroid side slip angle β 2 at the target control moment Po_C_2:

[0112]

[0113] Step S202: Obtain the target heading angle change rate at the target control moment according to the target centroid sideslip angle, the target steering angle array, the first pivot distance parameter, and the second pivot distance parameter.

[0114] Specifically, when implemented, the heading angle change rate is calculated by the following formula:

[0115]

[0116] where β is the centroid sideslip angle;

[0117] is the heading angle change rate, which is the derivative of the heading angle ψ with respect to time;

[0118] l f is the distance from the center of the front steering axle to the vehicle body reference center;

[0119] l r is the distance from the center of the rear steering axle to the vehicle body reference center;

[0120] σ r is the rear steering axle angle;

[0121] σ f is the front steering axle angle;

[0122] V is the vehicle speed.

[0123] Exemplarily, the heading angle change rate ψ of the target control moment Po_C_2 obtained 2 can be:

[0124]

[0125] Step S203: Based on the target heading angle change rate determined within the target time domain, obtain the heading angle eigenvalue at the target control moment as the target heading angle eigenvalue.

[0126] Among them, the heading angle eigenvalue represents the change value of the heading angle at the predicted position of the vehicle body at any control moment compared with the heading angle at the moment when the obstacle coordinate information is obtained. The prediction interval threshold is the time interval between any two adjacent control moments in the target steering angle control sequence information. In the embodiments of the present application, the predicted position of the vehicle body at the target control moment can be the predicted position after a time period of the length of the prediction interval threshold when the vehicle is controlled according to the target steering angle array at the target control moment.

[0127] In specific implementation, based on the determined target heading angle change rate within the target time domain, the heading angle eigenvalue at the predicted position after a time period with a length equal to the prediction interval threshold can be obtained when the vehicle is controlled according to the target steering angle array at the target control moment, and this eigenvalue is used as the target heading angle eigenvalue. It should be particularly noted that this target heading angle eigenvalue is obtained through cumulative calculation of each determined target heading angle change rate within the time period including the target control moment in the target time domain and the prediction interval threshold.

[0128] Exemplarily, assume that the prediction interval threshold is Δ t , the target heading angle change rate at the control moment Po_C_1 the target heading angle change rate at the target control moment Po_C_2 Based on the determined target heading angle change rate within the target time domain and the heading angle eigenvalue at the target control moment Po_C_2 is obtained, assumed to be ψ_2, and used as the target heading angle eigenvalue. The value ψ_2 of this target heading angle eigenvalue is obtained through cumulative calculation of each determined target heading angle change rate within the time period including the target control moment Po_C_2 in the target time domain Tar_T_Sc known and the prediction interval threshold Δ t .

[0129] Step S204: Obtain the abscissa and ordinate of the predicted vehicle body position at the target control moment according to the target centroid side slip angle, the target heading angle eigenvalue, the prediction interval threshold, and the vehicle speed information at the current moment.

[0130] Among them, the prediction interval threshold is the time interval between any two adjacent control moments in the target steering angle control sequence information.

[0131] In specific implementation, the abscissa and ordinate of the predicted vehicle body position at the target control moment are determined by combining the following formulas:

[0132]

[0133]

[0134] Among them, β is the centroid side slip angle;

[0135] ψ is the heading angle;

[0136] is the abscissa change rate of the predicted vehicle body position, which is the derivative of the abscissa X of the predicted vehicle body position with respect to time;

[0137] is the longitudinal coordinate change rate of the predicted vehicle body position, which is the derivative of the longitudinal coordinate Y of the predicted vehicle body position with respect to time;

[0138] V is the vehicle speed.

[0139] Exemplarily, based on the target centroid side slip angle β 2 , the target heading angle eigenvalue ψ_2, the prediction interval threshold Δ t and the vehicle speed information Speed at the current moment, the lateral coordinate change rate of the predicted vehicle body position at the target control moment Po_C_2 can be determined and the longitudinal coordinate change rate of the predicted vehicle body position and then the lateral coordinate X_2 of the predicted vehicle body position and the longitudinal coordinate Y_2 of the predicted vehicle body position at the target control moment Po_C_2 can be calculated.

[0140] In the embodiments of the present application, the process of calculating the lateral coordinate of the predicted vehicle body position and the longitudinal coordinate of the predicted vehicle body position based on the lateral coordinate change rate of the predicted vehicle body position and the longitudinal coordinate change rate of the predicted vehicle body position is similar to the process of calculating the target heading angle eigenvalue based on the target heading angle change rate, and the same parts will not be elaborated.

[0141] In some embodiments of the present application, based on the target steering angle control sequence information and the vehicle speed information, determining the predicted pose of the vehicle in the target time domain can also be achieved through a preset kinematic model.

[0142] Specifically, when implementing, the target steering angle control sequence information and the vehicle speed information are input into the preset kinematic model, so as to obtain the predicted pose of the vehicle in the target time domain.

[0143] In an embodiment of the present application, for an autonomous vehicle with dual steering axles, the kinematic model adopted is as follows:

[0144]

[0145] where β is the centroid side slip angle, representing the angle between the speed direction and the vehicle body direction; l f is the distance from the center of the front steering axle to the vehicle body reference center, l r is the distance from the center of the rear steering axle to the vehicle body reference center; σ r is the rear steering axle angle; σ f is the front steering axle angle; is the heading angle change rate, which is the derivative of the heading angle ψ with respect to time; ψ is the heading angle; is the lateral coordinate change rate of the predicted vehicle body position, which is the derivative of the lateral coordinate X of the predicted vehicle body position with respect to time; is the longitudinal coordinate change rate of the predicted vehicle body position, which is the derivative of the longitudinal coordinate Y of the predicted vehicle body position with respect to time; V is the vehicle speed; the vehicle body reference center is the vehicle body center.

[0146] Step S103: Determine the vehicle contour information of the vehicle within the target time domain according to the predicted pose and vehicle body size information.

[0147] In specific implementation, according to the predicted pose and vehicle body size information, the abscissa and ordinate of the predicted reference center of the vehicle body at each control moment, and the predicted direction information of the vehicle body at each control moment can be determined first, and then according to the abscissa of the predicted reference center of the vehicle body, the ordinate of the predicted reference center of the vehicle body, the predicted direction information of the vehicle body, and the vehicle body size information at each control moment, the vehicle contour information of the vehicle within the target time domain can be obtained.

[0148] Figure 3 It is a schematic flow diagram for determining the vehicle contour information of a vehicle collision prediction method provided by an embodiment of the present application. As Figure 3 shown, according to the predicted pose and vehicle body size information, to determine the vehicle contour information of the vehicle within the target time domain, it can be achieved through the following steps:

[0149] Step S301: Obtain the abscissa and ordinate of the predicted reference center of the vehicle body at each control moment according to the abscissa of the predicted position of the vehicle body, the ordinate of the predicted position of the vehicle body, the abscissa of the reference center of the vehicle body, and the ordinate of the reference center of the vehicle body.

[0150] Among them, the reference center of the vehicle body is the center of the vehicle body.

[0151] Exemplarily, taking the target control moment Po_C_2 as an example, assuming that the abscissa of the reference center of the vehicle body is a and the ordinate of the reference center of the vehicle body is b, according to the abscissa X_2 of the predicted position of the vehicle body, the ordinate Y_2 of the predicted position of the vehicle body, the abscissa a of the reference center of the vehicle body, and the ordinate b of the reference center of the vehicle body, the abscissa X_2_C and the ordinate Y_2_C of the predicted reference center of the vehicle body at the target control moment Po_C_2 are obtained:

[0152] X_2_C = X_2 + a,

[0153] Y_2_C = Y_2 + b.

[0154] Step S302: Determine the predicted direction information of the vehicle body at each control moment according to the heading angle eigenvalue.

[0155] Among them, the predicted direction information of the vehicle body represents the included angle between the predicted vehicle body direction at the control moment and the vehicle body direction at the moment of obtaining the obstacle coordinate information.

[0156] In specific implementation, to determine the predicted direction information of the vehicle body at each control moment according to the heading angle eigenvalue, the heading angle eigenvalue at each control moment can be used as the predicted direction information of the vehicle body at this control moment.

[0157] Exemplarily, the heading angle eigenvalue of each control moment (Po_C_1, Po_C_2, Po_C_3, Po_C_4, Po_C_5, Po_C_6, Po_C_7, Po_C_8, Po_C_9, Po_C_10) is obtained in sequence as the vehicle body prediction direction information Orie of the control moment (Po_C_1, Po_C_2, Po_C_3, Po_C_4, Po_C_5, Po_C_6, Po_C_7, Po_C_8, Po_C_9, Po_C_10). In the vehicle body prediction direction information Orie, the value of the vehicle body prediction direction information Orie at the control moment Po_C_2 is the heading angle eigenvalue ψ_2.

[0158] Step S303: Obtain the vehicle contour information of the vehicle within the target time domain according to the abscissa of the vehicle body prediction reference center, the ordinate of the vehicle body prediction reference center, the vehicle body prediction direction information, and the vehicle body size information at each control moment.

[0159] Exemplarily, according to the abscissa X_n_C of the vehicle body prediction reference center, the ordinate Y_n_C of the vehicle body prediction reference center, the vehicle body prediction direction information Orie, and the vehicle body size information at each control moment (Po_C_1, Po_C_2, Po_C_3, Po_C_4, Po_C_5, Po_C_6, Po_C_7, Po_C_8, Po_C_9, Po_C_10), the vehicle contour information of the autonomous vehicle within the target time domain Tar_T_Sc is obtained. For example, the vehicle contour information can be Figure 4 the multiple vehicle contours 400 shown in

[0160] Step S104: Determine the target collision prediction result of the vehicle according to the vehicle contour information and the obstacle coordinate information.

[0161] Among them, the target collision prediction result characterizes whether the vehicle collides with an obstacle within the target time domain.

[0162] Exemplarily, assume that the vehicle contour information is Figure 4 the multiple vehicle contours 400 shown in, RESULT_A characterizes that the autonomous vehicle will collide with an obstacle within the target time domain, and RESULT_B characterizes that the autonomous vehicle will not collide with an obstacle within the target time domain. If the obstacle coordinate information (Xp, Yp) falls inside at least one vehicle contour 400 or on at least one vehicle contour 400, then it is determined that the target collision prediction result of the autonomous vehicle is RESULT_A, and RESULT_A characterizes that the autonomous vehicle will collide with an obstacle within the target time domain.

[0163] Understandably, in some embodiments of the present application, when determining the target collision prediction result of the autonomous vehicle based on the vehicle contour information and the obstacle coordinate information, it may also be determined that the target collision prediction result of the autonomous vehicle is a possible collision when the obstacle coordinate information (Xp, Yp) falls inside or within the safety range of at least one vehicle contour. Among them, the safety range of the vehicle contour can be obtained by expanding the vehicle contour by a multiple of a preset proportional threshold. For example, the proportional threshold can be 1.2.

[0164] In some embodiments of the present application, after determining the target collision prediction result of the vehicle, it is further determined whether to perform emergency braking on the vehicle based on the target collision prediction result.

[0165] In this embodiment, based on the target collision prediction result, it is determined whether to perform emergency braking on the vehicle. Therefore, the AEB link does not depend on positioning\path, but only on the possible motion trajectories in the very short future, such as 2-3 seconds, in the current state and the obstacles sensed by the vehicle. If the trajectory in the next 2-3 seconds collides with the obstacle, then stop, which can ensure the safety of the vehicle to a greater extent.

[0166] The collision prediction method for a vehicle provided by the embodiments of the present application first obtains the vehicle speed information, the basic corner control information, and the obstacle coordinate information of the vehicle, and then determines the predicted pose of the vehicle in the target time domain based on the target corner control sequence information and the vehicle speed information. The target corner control sequence information is obtained based on the basic corner control information, and the target corner control sequence information includes an array of steering shaft angles corresponding to multiple control moments arranged in chronological order in the target time domain. Then, according to the predicted pose and the vehicle body size information, the vehicle contour information of the vehicle in the target time domain is determined, and based on the vehicle contour information and the obstacle coordinate information, the target collision prediction result of the vehicle is determined; the target collision prediction result represents whether the vehicle collides with an obstacle in the target time domain. This method does not depend on the positioning result, but only on the possible motion trajectories in the future target time domain in the current state and the obstacles sensed by the vehicle, reducing the occurrence of the result of the collision prediction being affected by positioning failure or abnormality, and improving the accuracy of the collision prediction.

[0167] Although the embodiments of the present application provide the operation steps of the method as shown in the above embodiments or the drawings, more or fewer operation steps may be included in the above method based on routine or non-creative labor. In the steps where there is no necessary causal relationship logically, the execution order of these steps is not limited to the execution order provided by the embodiments of the present application. When the above method is actually processed or executed by the device, it can be executed in the order shown in the embodiments or the drawings or executed in parallel.

[0168] Based on the same inventive concept, an embodiment of the present application further provides a collision prediction device for a vehicle. Since this device corresponds to the vehicle collision prediction method in the embodiment of the present application, and the principle of this device to solve problems is similar to that of the method, the implementation of this device can refer to the implementation process of the above method embodiment, and the repeated parts will not be elaborated.

[0169] Figure 5 The structural schematic diagram of a collision prediction device for a vehicle provided by an embodiment of the present application is shown. This collision prediction device for a vehicle, as Figure 5 shown, includes: an information acquisition unit 501, a pose prediction unit 502, a contour determination unit 503, and a collision detection unit 504; wherein,

[0170] The information acquisition unit 501 is configured to acquire the vehicle speed information, the basic steering angle control information, and the obstacle coordinate information of the vehicle;

[0171] The pose prediction unit 502 is configured to determine the predicted pose of the vehicle within the target time domain based on the target steering angle control sequence information and the vehicle speed information; the target steering angle control sequence information is obtained based on the basic steering angle control information; the target steering angle control sequence information includes an array of steering axis angles corresponding to multiple control moments arranged in chronological order within the target time domain;

[0172] The contour determination unit 503 is configured to determine the vehicle contour information of the vehicle within the target time domain according to the predicted pose and the vehicle body size information;

[0173] The collision detection unit 504 is configured to determine the target collision prediction result of the vehicle according to the vehicle contour information and the obstacle coordinate information; the target collision prediction result indicates whether the vehicle collides with an obstacle within the target time domain.

[0174] In an optional embodiment, the pose prediction unit 502 is further configured to:

[0175] Use the steering angle control sequence of the model predictive control (MPC) in the acquired basic steering angle control information as the target steering angle control sequence information.

[0176] In an optional embodiment, the basic steering angle control information further includes the vehicle wheel angle information of the vehicle; the pose prediction unit 502 is further configured to:

[0177] If the steering angle control sequence of the MPC is not included in the basic steering angle control information, use the vehicle wheel angle information as the value of each group of steering axis angle arrays in the target steering angle control sequence information.

[0178] In an optional embodiment, the obstacle coordinate information is the coordinate of the obstacle in the vehicle body coordinate system at the moment when the obstacle coordinate information is acquired; the obstacle coordinate information is obtained through an in-vehicle sensor.

[0179] In an alternative embodiment, the predicted pose includes the abscissa of the predicted vehicle position, the ordinate of the predicted vehicle position, and the heading angle eigenvalue in the vehicle body coordinate system at the moment of obtaining the obstacle coordinate information; the heading angle eigenvalue represents the change value of the heading angle at the predicted vehicle position at any control moment compared with the heading angle at the moment of obtaining the obstacle coordinate information.

[0180] In an alternative embodiment, the steering axles of the vehicle include a front steering axle and a rear steering axle; the steering axle angle array includes the front steering axle angle and the rear steering axle angle; the pose prediction unit 502 is specifically configured to:

[0181] Obtain the target steering angle array at the target control moment in the target corner control sequence information one by one in chronological order, and sequentially perform the following steps on the obtained target steering angle array at the target control moment to obtain the predicted pose of the vehicle at each target control moment:

[0182] According to the obtained target steering angle array and the preset first pivot distance parameter and second pivot distance parameter, calculate the target centroid side slip angle at the target control moment; the first pivot distance parameter is the distance from the center of the front steering axle to the vehicle body reference center; the second pivot distance parameter is the distance from the center of the rear steering axle to the vehicle body reference center; the vehicle body reference center is the center of the vehicle body;

[0183] According to the target centroid side slip angle, the target steering angle array, the first pivot distance parameter, and the second pivot distance parameter, obtain the target heading angle change rate at the target control moment;

[0184] Based on the target heading angle change rate determined within the target time domain, obtain the heading angle eigenvalue at the target control moment as the target heading angle eigenvalue;

[0185] According to the target centroid side slip angle, the target heading angle eigenvalue, the prediction interval threshold, and the vehicle speed information at the current moment, obtain the abscissa of the predicted vehicle position and the ordinate of the predicted vehicle position at the target control moment; the prediction interval threshold is the time interval between any two adjacent control moments in the target corner control sequence information.

[0186] In an alternative embodiment, the contour determination unit 503 is specifically configured to:

[0187] According to the abscissa of the predicted vehicle position, the ordinate of the predicted vehicle position, the abscissa of the vehicle body reference center, and the ordinate of the vehicle body reference center, obtain the abscissa of the predicted vehicle body reference center and the ordinate of the predicted vehicle body reference center at each control moment;

[0188] Determine the vehicle body prediction direction information at each control moment according to the heading angle eigenvalue; the vehicle body prediction direction information represents the included angle between the predicted vehicle body direction at the control moment and the vehicle body direction at the moment when the obstacle coordinate information is obtained.

[0189] According to the abscissa of the predicted reference center of the vehicle body, the ordinate of the predicted reference center of the vehicle body, the vehicle body prediction direction information and the vehicle body size information at each control moment, obtain the vehicle contour information of the vehicle within the target time domain.

[0190] Based on the same inventive concept as the above method embodiments, an electronic device is also provided in the embodiments of the present application. This electronic device can be used for vehicle collision prediction. In one embodiment, the electronic device can be a server, or a terminal device or other electronic devices. In this embodiment, the structure of the electronic device can be as Figure 6 shown, including a memory 601, a communication module 603, and one or more processors 602.

[0191] The memory 601 is used to store the computer program executed by the processor 602. The memory 601 may mainly include a program storage area and a data storage area. Among them, the program storage area may store an operating system and programs required to run the instant messaging function, etc.; the data storage area may store various instant messaging information and operation instruction sets, etc.

[0192] The memory 601 can be a volatile memory, such as a random-access memory (RAM); the memory 601 can also be a non-volatile memory, such as a read-only memory, a flash memory, a hard disk drive (HDD) or a solid-state drive (SSD), or the memory 601 is any other medium that can be used to carry or store the desired program code in the form of instructions or data structures and can be accessed by a computer, but is not limited thereto. The memory 601 can be a combination of the above memories.

[0193] The processor 602 can include one or more central processing units (CPUs) or be a digital processing unit, etc. The processor 602 is used to implement the above vehicle collision prediction method when calling the computer program stored in the memory 601.

[0194] The communication module 603 is used to communicate with the terminal device and other servers.

[0195] In the embodiments of the present application, the specific connection medium between the above-mentioned memory 601, communication module 603 and processor 602 is not limited. In the embodiments of the present disclosure Figure 6 it is shown that the memory 601 and the processor 602 are connected through a bus 604, and the bus 604 is represented by a thick line in Figure 6 The connection manners between other components are only for illustrative purposes and are not limiting. The bus 604 can be divided into an address bus, a data bus, a control bus, etc. For the sake of convenience of representation, Figure 6 it is only represented by a thick line in, but it does not mean that there is only one bus or one type of bus.

[0196] The embodiments of the present application also provide a computer storage medium, in which computer-executable instructions are stored, and the computer-executable instructions are used to implement the vehicle collision prediction method of any embodiment of the present application.

[0197] According to one aspect of the present application, there is provided a computer program product or a computer program, the computer program product or the computer program includes computer instructions, and the computer instructions are stored in a computer-readable storage medium. The processor of the computer device reads the computer instructions from the computer-readable storage medium, and the processor executes the computer instructions, so that the computer device executes the vehicle collision prediction method in the above embodiments. The program product may adopt any combination of one or more readable media. The readable media may be a readable signal medium or a readable storage medium. The readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination of the above. More specific examples (non-exhaustive list) of the readable storage medium include: an electrical connection with one or more wires, a portable disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above.

[0198] The above is only the specific implementation manner of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art can easily think of changes or substitutions within the technical scope disclosed by the present application, and all should be covered by the protection scope of the present application.

Claims

1. A collision prediction method for a vehicle, characterized in that, the method includes: Obtaining the vehicle speed information, basic corner control information and obstacle coordinate information of the vehicle; the vehicle is a low-speed autonomous vehicle operating at a low speed; the vehicle speed information is the vehicle centroid speed of the vehicle; Based on the target corner control sequence information and the vehicle speed information, determining the predicted pose of the vehicle within the target time domain; the target corner control sequence information is obtained based on the basic corner control information; the target corner control sequence information includes an array of steering axis angles corresponding one-to-one with multiple control times arranged in chronological order within the target time domain; According to the predicted pose and the vehicle body size information, determining the vehicle contour information of the vehicle within the target time domain; According to the vehicle contour information and the obstacle coordinate information, determining the target collision prediction result of the vehicle; the target collision prediction result indicates whether the vehicle collides with an obstacle within the target time domain; Before determining the predicted pose of the vehicle within the target time domain based on the target corner control sequence information and the vehicle speed information, the method further includes: Taking the corner control sequence of the model predictive control (MPC) in the obtained basic corner control information as the target corner control sequence information; The basic corner control information further includes the vehicle wheel angle information of the vehicle; after taking the corner control sequence of the model predictive control (MPC) in the basic corner control information as the target corner control sequence information, and before determining the predicted pose of the vehicle within the target time domain based on the target corner control sequence information and the vehicle speed information, it further includes: If the corner control sequence of the MPC is not included in the basic corner control information, taking the vehicle wheel angle information as the value of each group of the steering axis angle arrays in the target corner control sequence information.

2. The method according to claim 1, characterized in that, The obstacle coordinate information is the coordinate of the obstacle in the vehicle body coordinate system at the moment when the obstacle coordinate information is obtained; the obstacle coordinate information is obtained through an in-vehicle sensor.

3. The method according to claim 1, characterized in that, The predicted pose includes the abscissa of the predicted vehicle body position, the ordinate of the predicted vehicle body position and the heading angle eigenvalue in the vehicle body coordinate system at the moment when the obstacle coordinate information is obtained; the heading angle eigenvalue indicates the change value of the heading angle at the predicted vehicle body position at any control time compared with the heading angle at the moment when the obstacle coordinate information is obtained.

4. The method according to claim 3, characterized in that, The steering axis of the vehicle includes a front steering axis and a rear steering axis; the steering axis angle array includes a front steering axis angle and a rear steering axis angle; determining the predicted pose of the vehicle within the target time domain based on the target corner control sequence information and the vehicle speed information includes: Obtain the target steering angle array at the target control moment in the target corner control sequence information one by one in chronological order, and successively perform the following steps on the obtained target steering angle array at the target control moment to obtain the predicted pose of the vehicle at each target control moment: Calculate the target centroid side slip angle at the target control moment according to the obtained target steering angle array and the preset first kingpin offset parameter and second kingpin offset parameter; the first kingpin offset parameter is the distance from the center of the front steering axle to the body reference center; the second kingpin offset parameter is the distance from the center of the rear steering axle to the body reference center; the body reference center is the center of the vehicle body; Obtain the target heading angle change rate at the target control moment according to the target centroid side slip angle, the target steering angle array, the first kingpin offset parameter and the second kingpin offset parameter; Based on the target heading angle change rate determined within the target time domain, obtain the heading angle eigenvalue at the target control moment as the target heading angle eigenvalue; Obtain the predicted abscissa of the vehicle body and the predicted ordinate of the vehicle body at the target control moment according to the target centroid side slip angle, the target heading angle eigenvalue, the prediction interval threshold and the vehicle speed information at the current moment; the prediction interval threshold is the time interval between any two adjacent control moments in the target corner control sequence information.

5. The method according to claim 4, wherein, According to the predicted pose and the vehicle body size information, determine the vehicle contour information of the vehicle within the target time domain, including: Obtain the predicted reference center abscissa of the vehicle body and the predicted reference center ordinate of the vehicle body at each control moment according to the predicted abscissa of the vehicle body, the predicted ordinate of the vehicle body, the abscissa of the body reference center, and the ordinate of the body reference center; Determine the predicted body direction information at each control moment according to the heading angle eigenvalue; the predicted body direction information represents the angle between the predicted body direction at the control moment and the body direction at the moment when the obstacle coordinate information is obtained; Obtain the vehicle contour information of the vehicle within the target time domain according to the predicted reference center abscissa of the vehicle body, the predicted reference center ordinate of the vehicle body, the predicted body direction information and the vehicle body size information at each control moment.

6. A vehicle collision prediction device, wherein, comprising: An information acquisition unit for acquiring the vehicle speed information, the basic corner control information and the obstacle coordinate information of the vehicle; The vehicle is a low-speed autonomous vehicle running at a low speed; The vehicle speed information is the vehicle centroid speed of the vehicle; A pose prediction unit for determining the predicted pose of the vehicle within the target time domain based on the target corner control sequence information and the vehicle speed information; the target corner control sequence information is obtained based on the basic corner control information; the target corner control sequence information includes a steering axle angle array corresponding one by one to a plurality of control moments arranged in chronological order within the target time domain; A contour determination unit, configured to determine vehicle contour information of the vehicle within the target time domain according to the predicted pose and vehicle dimension information; A collision detection unit, configured to determine a target collision prediction result of the vehicle according to the vehicle contour information and the obstacle coordinate information; the target collision prediction result indicates whether the vehicle collides with an obstacle within the target time domain; The pose prediction unit is further configured to: Use the corner control sequence of the model predictive control (MPC) in the obtained basic corner control information as the target corner control sequence information; The basic corner control information further includes vehicle wheel angle information of the vehicle; the pose prediction unit is further configured to: If the basic corner control information does not include the corner control sequence of the MPC, use the vehicle wheel angle information as the value of each steering axis angle array in the target corner control sequence information.

7. A computer-readable storage medium, in which a computer program is stored, Characterized in that: When the computer program is executed by a processor, the method according to any one of claims 1 to 5 is implemented.

8. An electronic device, Characterized in that, It includes a memory and a processor, and a computer program that can run on the processor is stored on the memory. When the computer program is executed by the processor, the method according to any one of claims 1 to 5 is implemented.

Images