Method and apparatus for an autonomous vehicle to determine a driving strategy

Abstract

A method and device for determining a driving strategy of an autonomous vehicle, the method comprising: obtaining dynamic information, static information, and spatial information of a first vehicle; determining a unit time collision reaction space of the vehicle according to the dynamic information of the vehicle and the static information of the vehicle; calculating a spatial interaction entropy between the first vehicle and other vehicles within the unit time collision reaction space of the vehicle; and executing a predetermined driving strategy if the spatial interaction entropy is less than a set threshold, wherein the predetermined driving strategy is determined according to the spatial information.

Description

Technical Field

[0001] This article relates to the field of autonomous driving technology, and in particular to a method and apparatus for determining driving strategies for autonomous vehicles. Background Technology

[0002] Defensive driving refers to the ability of a vehicle to anticipate dangers arising from other vehicles, the surrounding environment, and current road conditions, and to take appropriate driving decisions to prevent accidents and plan routes. Defensive driving is included within autonomous driving decision-making technology, and its characteristic should be that the vehicle has the ability to anticipate potential accidents and make corresponding decisions.

[0003] Some autonomous driving decision-making technologies include those using genetic algorithms, ant colony optimization, ID3 decision trees, neural networks, and artificial potential field methods. Among these, neural network methods require extensive data training; however, conventional datasets often fail to handle the vast complexity of different environments, making it difficult to cover all vehicle conditions. Artificial potential field methods have been applied to low-level autonomous driving assistance, but they still suffer from local optima, collisions, and failure to reach target points. Throughout the autonomous driving decision-making process, numerous unknown behavioral factors influence decision-making; these factors are diverse and lack hierarchy, leading to insufficient information acquisition and representation. Furthermore, the aforementioned autonomous driving decision-making methods cannot assess the complexity of the surrounding environment, necessitating a method that can determine driving strategies based on spatial interaction entropy and spatial repulsive forces. Summary of the Invention

[0004] This application provides a method and apparatus for determining driving strategies for autonomous vehicles. In an emergency, the vehicle can assess the degree of danger of the surrounding environment through spatial interaction entropy, thereby executing a better driving strategy, which can improve the vehicle's collision avoidance capabilities and its comprehensive scene adaptability in autonomous driving environments.

[0005] This application provides a method for determining a driving strategy for an autonomous vehicle, the method comprising:

[0006] Acquire dynamic information, static information, collaborative space and warning space information of the first vehicle;

[0007] The vehicle's collision response space per unit time is determined based on the vehicle's dynamic information and static information.

[0008] Within the collision response space of the vehicle per unit time, calculate the spatial interaction entropy between the first vehicle and other vehicles;

[0009] If the spatial interaction entropy is less than the set threshold, then the predetermined driving strategy is executed;

[0010] The predetermined driving strategy is determined based on the vector sum of the cooperative space repulsive force and the warning space repulsive force;

[0011] The cooperative space is the minimum spatial area required for the first vehicle and other vehicles to cooperate in collision avoidance; the warning space is the maximum spatial area that the sensors of the first vehicle can perceive.

[0012] In one exemplary embodiment, the static information includes: acquiring environmental information within the collaboration space and the warning space, calculating the static environment reaction space, and determining the location information of static obstacles located within the static environment reaction space or affecting vehicle collision avoidance operations;

[0013] The dynamic information includes: vehicle speed, vehicle instantaneous acceleration, vehicle centrifugal acceleration, vehicle adhesion coefficient to the surrounding ground, and gravitational acceleration of the area where the vehicle is located.

[0014] In one exemplary embodiment, the collision response space of the vehicle per unit time is a two-dimensional region composed of a first curve, a second curve, a third curve, and a fourth straight line.

[0015] Among them, the first curve and the second curve are the circular arc curves of the maximum steering trajectory;

[0016] The third curve is elliptical in shape, with the vertices of the first curve, the second curve, and the distance point from which the first vehicle stops at its maximum deceleration along a straight line, respectively. The major axis of the ellipse is half the distance between the first curve and the second curve, and the minor axis of the ellipse is the distance from the distance point to the first curve and the second curve.

[0017] The fourth straight line is the trajectory of the vehicle traveling at its maximum deceleration when its centrifugal acceleration is 0.

[0018] In one exemplary embodiment, the method for determining the first curve and the second curve is as follows:

[0019] Based on the vehicle's speed, instantaneous acceleration, adhesion coefficient between the vehicle and the surrounding ground, and gravitational acceleration of the area where the vehicle is located, the radius of curvature and arc length of the maximum steering trajectory of the vehicle are determined using a preset calculation formula.

[0020] The preset calculation formula is as follows:

[0021]

[0022] In the above formula, v is the vehicle speed, and a r Let be the instantaneous acceleration of the vehicle, μ be the adhesion coefficient between the vehicle and the surrounding ground, g be the gravitational acceleration of the area where the vehicle is located, r be the radius of curvature of the circular arc, and s be the arc length of the circular arc.

[0023] In one exemplary embodiment, the method for determining the third curve is as follows:

[0024] Determined based on the turning apex of the first curve and the second curve, and the distance from the point where the first vehicle stops at its maximum deceleration along a straight line;

[0025] Wherein, the distance x from the point where the first vehicle stops after its maximum deceleration along the straight line is:

[0026]

[0027] In the above formula, x is the distance from the point of maximum deceleration along a straight line to a stop, and a r Let T be the instantaneous acceleration of the vehicle, T be the time it takes for the first vehicle to come to a stop along a straight line using maximum deceleration, and t be the travel time of the first vehicle.

[0028] In one exemplary embodiment, the method for determining the fourth straight line is as follows:

[0029] Acquire the vehicle's decision-making response time from receiving surrounding information to initiating collision avoidance maneuvers;

[0030] The fourth set of lines is the signal response space, which is the region enclosed by the starting points of the first and second curves and the left and right front vertices of the first vehicle, and the displacement curve of the first vehicle within the decision response time.

[0031] In one exemplary embodiment, the spatial interaction entropy between vehicles includes: lateral spatial interaction entropy and longitudinal spatial interaction entropy;

[0032] The calculation of the spatial interaction entropy between the vehicle and other vehicles within the vehicle's collision response space per unit time includes:

[0033] Based on grid partitioning, the vehicle collision response space per unit time is divided into j 2D sub-regions; the ratio P(ψ) of the intersection area of ​​each sub-region within the overlapping area of ​​the second vehicle's reaction space and the first vehicle's reaction space per unit time to the original area of ​​the sub-region is calculated. j );

[0034] Calculate the ratio P(υ) of the empty (non-intersecting) area of ​​each sub-region within the overlapping area of ​​the reaction spaces of the second and first vehicles in the collision reaction space per unit time to the original area of ​​the sub-region. j );

[0035] Calculate the lateral spatial interaction entropy and longitudinal spatial interaction entropy of the vehicle based on the ratios.

[0036] The other vehicles include multiple second vehicles.

[0037] In one exemplary embodiment, the formula for calculating the longitudinal spatial interaction entropy is:

[0038]

[0039]

[0040] The formula for calculating the transverse spatial interaction entropy is:

[0041]

[0042] in, The minimum value is -1, ln(P(ψ) j The minimum value of )) is -1, k is the total number of the second vehicles in the warning space, j is the number of 2D regions that the collision reaction space is divided into per unit time, m and n are the starting values ​​for calculating spatial interaction entropy, and ε is the area of ​​the collision reaction space of the first vehicle per unit time.

[0043] In one exemplary embodiment, the repulsive forces include longitudinal repulsive forces and transverse repulsive forces;

[0044] Among them, the longitudinal repulsive force F z for:

[0045]

[0046] The lateral repulsive force F h for:

[0047]

[0048] In the above formula, F z For longitudinal repulsive forces, F h For transverse repulsive forces, w j To indicate the repulsive force weight in the warning space, w x V represents the repulsive force weight in the cooperative space. h V represents the velocity component perpendicular to the lane line (lateral direction). z The velocity component is in the direction parallel to the lane line (longitudinal). denoted as the lateral and longitudinal velocity components of the second vehicle, d represents the distance between vehicles, and k represents the total number of the second vehicles within the warning space.

[0049] In one exemplary embodiment, the predetermined driving strategy includes adjusting the vehicle steering wheel angle or adjusting the vehicle acceleration;

[0050] in,

[0051] The steering wheel angle is: θ = μF h ;

[0052] The acceleration is: a = δF z ;

[0053] In the above formula, θ is the steering wheel angle, a is the vehicle acceleration, and μ and δ are both pre-calibrated values.

[0054] This application also provides an apparatus for determining a driving strategy based on spatial interaction entropy and spatial repulsion, the apparatus comprising: a memory and a processor, the memory being used to store a program for determining a driving strategy for an autonomous vehicle, and the processor being used to read and execute the program for determining a driving strategy for an autonomous vehicle, and to execute the method described in any of the above embodiments.

[0055] Compared with related technologies, this application provides a method and apparatus for determining a driving strategy for an autonomous vehicle. The method includes: acquiring dynamic information, static information, and spatial information of a first vehicle; determining the vehicle's collision response space per unit time based on the vehicle's dynamic information and static information; calculating the spatial interaction entropy between the first vehicle and other vehicles within the vehicle's collision response space per unit time; and executing a predetermined driving strategy if the spatial interaction entropy is less than a set threshold. The predetermined driving strategy is determined based on the spatial information. Through the technical solution of this invention, in an emergency situation, this method can assess the degree of danger of the surrounding environment through spatial interaction entropy, thereby executing a better driving strategy, which can improve the vehicle's collision avoidance capability and its comprehensive scene adaptability in an autonomous driving environment.

[0056] Other features and advantages of this application will be set forth in the following description, and will be apparent in part from the description, or may be learned by practicing the application. Other advantages of this application can be realized and obtained by means of the solutions described in the description and the accompanying drawings. Attached Figure Description

[0057] The accompanying drawings are used to provide an understanding of the technical solutions of this application and constitute a part of the specification. They are used together with the embodiments of this application to explain the technical solutions of this application and do not constitute a limitation on the technical solutions of this application.

[0058] Figure 1 A flowchart illustrating a method for determining a driving strategy for an autonomous vehicle according to an embodiment of this application;

[0059] Figure 2 A schematic diagram of a device for determining a driving strategy for an autonomous vehicle according to an embodiment of this application;

[0060] Figure 3 These are methods for defensive driving strategies for autonomous vehicles based on spatial interaction entropy in some exemplary embodiments;

[0061] Figure 4 This is a schematic diagram of the longitudinal and lateral safety control response space of an autonomous vehicle in some exemplary embodiments;

[0062] Figure 5 These are schematic diagrams illustrating the interaction entropy of the safety response space of autonomous vehicles in some exemplary embodiments;

[0063] Figure 6 This is a schematic diagram of the repulsive forces between the warning space and the cooperative space of an autonomous vehicle in some exemplary embodiments. Detailed Implementation

[0064] This application describes several embodiments, but these descriptions are exemplary and not restrictive, and it will be apparent to those skilled in the art that many more embodiments and implementations are possible within the scope of the embodiments described herein. Although many possible combinations of features are shown in the drawings and discussed in the detailed description, many other combinations of the disclosed features are also possible. Unless specifically limited, any feature or element of any embodiment may be used in combination with, or may replace, any feature or element of any other embodiment.

[0065] This application includes and contemplates combinations of features and elements known to those skilled in the art. The embodiments, features, and elements disclosed in this application may also be combined with any conventional features or elements to form a unique inventive scheme as defined by the claims. Any feature or element of any embodiment may also be combined with features or elements from other inventive schemes to form another unique inventive scheme as defined by the claims. Therefore, it should be understood that any feature shown and / or discussed in this application may be implemented individually or in any suitable combination. Therefore, the embodiments are not limited except by the limitations imposed by the appended claims and their equivalents. Furthermore, various modifications and changes may be made within the scope of the appended claims.

[0066] Furthermore, in describing representative embodiments, the specification may have presented methods and / or processes as a specific sequence of steps. However, the method or process should not be limited to the specific order of steps described herein, to the extent that it does not depend on such a specific order. As will be understood by those skilled in the art, other sequences of steps are also possible. Therefore, the specific order of steps set forth in the specification should not be construed as a limitation of the claims. Moreover, the claims concerning the method and / or process should not be limited to the steps performed in the written order, and those skilled in the art will readily understand that these orders can be varied and still remain within the spirit and scope of the embodiments of this application.

[0067] Information entropy, as a scalar, does not provide information about what behavior or reaction the vehicle will perform. Entropy primarily represents the vehicle's decision-making needs and specifies how many reactions are feasible at any given moment during driving. Therefore, entropy can assess the complexity of the vehicle's surrounding environment. In an emergency, the vehicle can assess the degree of danger in the surrounding environment through spatial interaction entropy, thereby executing the optimal driving strategy. This invention combines the vehicle's reaction space and information entropy to propose the concept of spatial interaction entropy and applies it to defensive driving decisions. This technology can significantly improve the vehicle's collision avoidance capabilities and overall scenario adaptability in autonomous driving environments. When the vehicle is not in an emergency, it calculates a repulsive force vector sum through the cooperative space and warning space, performs a new assessment of the current driving route, and makes a higher-level defensive driving strategy, improving driving stability in autonomous driving environments.

[0068] This disclosure provides a method for determining a driving strategy for an autonomous vehicle, such as... Figure 1 As shown, the method includes steps S100-S130, as detailed below:

[0069] S100. Obtain dynamic information, static information, collaborative space and warning space information of the first vehicle;

[0070] S110. Determine the vehicle's collision response space per unit time based on the vehicle's dynamic information and static information;

[0071] S120. Within the collision response space of the vehicle per unit time, calculate the spatial interaction entropy between the vehicle and other vehicles;

[0072] S130. If the interaction entropy of the reaction space is less than the set threshold, then the predetermined driving strategy is executed;

[0073] In this embodiment, the predetermined driving strategy is determined based on the vector sum of the cooperative space-like repulsive force and the warning space-like repulsive force; wherein, the cooperative space is the minimum spatial area required for the first vehicle and other vehicles to cooperate in collision avoidance; and the warning space is the maximum spatial area that the sensors of the first vehicle can perceive. The size relationship between the cooperative space, the warning space, and the collision response space per unit time is as follows: Figure 3 As shown.

[0074] In one exemplary embodiment, the static information includes: static obstacles around the vehicle, typically considered as: walls, large rocks, stationary vehicles, construction barriers, etc.; and the static information mainly considers: location information, category information, size information, etc.

[0075] The dynamic information includes: vehicle speed, vehicle instantaneous acceleration, vehicle centrifugal acceleration, vehicle adhesion coefficient to the surrounding ground, and gravitational acceleration of the area where the vehicle is located.

[0076] In one exemplary embodiment, the vehicle's collision response space per unit time is determined based on the vehicle's dynamic information and static information; wherein, the vehicle's collision response space per unit time is a two-dimensional region composed of a first curve, a second curve, a third curve, and a fourth straight line.

[0077] Among them, the first curve and the second curve are the circular arc curves of the maximum steering trajectory;

[0078] The third curve is elliptical in shape, with the vertices of the first curve, the second curve, and the distance point from which the first vehicle stops at its maximum deceleration along a straight line, respectively. The major axis of the ellipse is half the distance between the first curve and the second curve, and the minor axis of the ellipse is the distance from the distance point to the first curve and the second curve.

[0079] The fourth straight line is the trajectory of the vehicle traveling at its maximum deceleration when its centrifugal acceleration is 0.

[0080] In one exemplary embodiment, the method for determining the third curve is as follows:

[0081] Determined based on the turning apex of the first curve and the second curve, and the distance from the point where the first vehicle stops at its maximum deceleration along a straight line;

[0082] Wherein, the distance x from the point where the first vehicle stops after its maximum deceleration along the straight line is:

[0083]

[0084] In the above formula, x is the distance from the point of maximum deceleration along a straight line to a stop, and a rLet T be the instantaneous acceleration of the vehicle, T be the time it takes for the first vehicle to come to a stop along a straight line using maximum deceleration, and t be the travel time of the first vehicle.

[0085] In one exemplary embodiment, the method for determining the fourth straight line is as follows:

[0086] Acquire the vehicle's decision-making response time from receiving surrounding information to initiating collision avoidance maneuvers;

[0087] The fourth set of lines is the signal response space, which is the region enclosed by the starting points of the first and second curves and the left and right front vertices of the first vehicle, and the displacement curve of the first vehicle within the decision response time.

[0088] In step S120, the spatial interaction entropy between vehicles includes: lateral spatial interaction entropy and longitudinal spatial interaction entropy. Calculating the spatial interaction entropy between the vehicle and other vehicles within the vehicle's collision response space per unit time includes:

[0089] The first step is to calculate the collision response space per unit time and the surrounding static environment response space, where the static environment response space is the 2D area occupied by the static environment. Since in actual driving scenarios, the vehicle cannot use the static environment response space to avoid collisions (it is occupied by static obstacles), subtracting the static environment response space yields the vehicle's effective collision response space.

[0090] The second step is to divide the vehicle's collision response space per unit time into j 2D sub-regions based on grid partitioning.

[0091] The third step is to calculate the ratio P(ψ) of the area of ​​the intersection of each sub-region in the reaction space of the second vehicle and the reaction space of the first vehicle per unit time to the area of ​​the sub-region. j ); wherein, the other vehicles include multiple second vehicles.

[0092] Step 4: Calculate the ratio P(υ) of the empty (non-intersecting) area of ​​each sub-region within the overlapping area of ​​the reaction space of the second vehicle and the reaction space of the first vehicle per unit time to the original area of ​​the sub-region. j ):

[0093]

[0094] Step 5: Calculate the lateral spatial interaction entropy and longitudinal spatial interaction entropy of the vehicle based on the ratio.

[0095] Calculate the vertical spatial interaction entropy. The minimum value is -1, as shown in the formula below:

[0096]

[0097] Calculate the transverse spatial interaction entropy, ln(P(ψ) j The minimum value of )) is -1, and the formula is as follows:

[0098]

[0099] in, The minimum value is -1, ln(P(ψ) j The minimum value of )) is -1, k is the number of surrounding second vehicles, j is the number of 2D regions that the collision reaction space is divided into per unit time, m and n are the starting values ​​for calculating spatial interaction entropy, and ε is the area of ​​the collision reaction space of the first vehicle per unit time.

[0100] In one exemplary embodiment, after calculating the spatial interaction entropy between the vehicle and other vehicles, the value of the spatial interaction entropy is compared with a pre-set threshold (different thresholds are set according to different vehicle types and different vehicle speeds) to execute different driving strategies:

[0101] The first category: Cases where the spatial interaction entropy is greater than or equal to the set threshold.

[0102] If both the lateral spatial interaction entropy and the longitudinal spatial interaction entropy are greater than or equal to the threshold, calculate the deceleration of the first vehicle and the steering wheel angle.

[0103] The deceleration of the first vehicle is calculated as follows:

[0104] Let γ be the deceleration coefficient, and the deceleration calculation formula is as follows:

[0105]

[0106] In the above formula, a is the vehicle deceleration, and γ is the deceleration coefficient, which is a constant. It represents the vertical spatial interaction entropy.

[0107] The steering wheel angle is calculated as follows:

[0108] In the above formula, θ is the steering wheel angle, ξ is the steering coefficient, P(ψ1), P(ψ2)... It is the ratio of the area of ​​the intersection of each sub-region in the reaction space of the second vehicle on the left side of the vehicle's main axle and the reaction space of the first vehicle to the area of ​​the sub-region. …P(ψ k This is the ratio of the area of ​​the intersection of each sub-region in the reaction space of the second vehicle and the reaction space of the first vehicle to the area of ​​the sub-region itself. The vehicle adjusts its driving strategy in the next unit of time based on deceleration and steering wheel angle. For example... Figure 5 This shows that the vehicle will adjust its driving strategy based on the interaction entropy of the reaction space.

[0109] The second category: Both the horizontal spatial interaction entropy and the vertical spatial interaction entropy are less than the set threshold.

[0110] If both the lateral and longitudinal spatial interaction entropies are less than the set thresholds, the repulsive force vectors of the cooperative space and the warning space are calculated separately. Based on the different weights of the repulsive forces in the cooperative and warning spaces, the sum of the repulsive force vectors of the cooperative and warning spaces is calculated. The driving strategy is then determined based on this sum. Figure 6 As shown.

[0111] First, calculate the longitudinal and transverse repulsive forces;

[0112] Repulsive forces are classified into longitudinal (i.e., lane line direction) repulsive forces and lateral repulsive forces (i.e., perpendicular to lane line direction); the magnitude of the repulsive force is related to the distance and relative speed between vehicles.

[0113] Longitudinal repulsive forces are used to control vehicle acceleration, while lateral repulsive forces are used to control the vehicle's steering wheel angle. For example, the velocity component of the first vehicle in the direction of its current lane is V. z The velocity component perpendicular to the lane line is V h The velocity component of the second vehicle within the collision reaction space per unit time is: and The repulsive force weights for the warning space and the collaboration space are w respectively. j w x The longitudinal and lateral components of the second and first vehicles within the collision reaction space per unit time are F1 and F2, respectively. z and F h .

[0114] The longitudinal repulsive force F in the sensing space z The formula is as follows:

[0115]

[0116] Transverse repulsive force F in sensing space h The formula is as follows:

[0117]

[0118] Secondly, based on the lateral and longitudinal repulsive forces in the sensing space, the steering wheel angle and acceleration of the first vehicle are calculated.

[0119] Based on the lateral and longitudinal repulsive forces in the sensing space, the formula for calculating the acceleration of the first vehicle is as follows:

[0120] aδF z

[0121] Based on the lateral and longitudinal repulsive forces in the sensing space, the formula for calculating the steering wheel angle of the first vehicle is as follows:

[0122] θ=μE h

[0123] Wherein, δ and μ are both manually calibrated values, with μ ranging from 2.2 to 5.3 and δ ranging from 0.01 to 0.1.

[0124] Based on the different repulsive force weights in the cooperative space and the warning space, the sum of the repulsive force vectors in the cooperative space and the warning space is calculated. The magnitude of the sum of the repulsive force vectors determines the specific operational level of defensive driving decisions. For example, the longitudinal repulsive force vector sum controls the vehicle's acceleration, and the lateral repulsive force vector sum controls the vehicle's steering wheel angle. The formula for calculating the sum of the repulsive force vectors is:

[0125]

[0126] Among them, w j w x V represents the repulsive force weights of the warning space and the cooperation space, respectively. i V represents the speed of the second vehicle, and V represents the speed of the first vehicle. Figure 6 This demonstrates a scenario where a vehicle is about to make defensive driving decisions based on the sum of repulsive force vectors in the sensed space. The sum of these repulsive force vectors reflects the complexity of the required defensive driving decisions. Longitudinal repulsive forces control vehicle acceleration, while lateral repulsive forces control the steering wheel angle. Driving decisions such as acceleration or steering are made based on the sum of these repulsive force vectors. The demonstration shows the different weights of the repulsive forces in the vehicle's cooperative and warning zones.

[0127] This application discloses a defensive driving decision-making method based on spatial interaction entropy and spatial repulsive forces, which can assess the complexity of the vehicle's surrounding environment. In an emergency, the vehicle can assess the degree of danger in the surrounding environment through spatial interaction entropy, thereby executing the optimal driving strategy. This technology combines the vehicle's reaction space and information entropy to propose the concept of spatial interaction entropy and applies it to defensive driving decision-making. This technology can significantly improve the vehicle's collision avoidance capabilities and overall scenario adaptability in autonomous driving environments. When the vehicle is not in an emergency, it calculates the repulsive force vector sum through the cooperative space and warning space, performs a new assessment of the current driving route, and formulates a higher-level defensive driving strategy, improving driving stability in autonomous driving environments.

[0128] This disclosure also provides an apparatus for determining a driving strategy for an autonomous vehicle, such as... Figure 2As shown, the device includes a memory 210 and a processor 220; the memory 210 is used to store a program for determining a driving strategy for an autonomous vehicle, and the processor 220 is used to read and execute the program for determining a driving strategy for an autonomous vehicle, and execute any of the methods for determining a driving strategy for an autonomous vehicle in the above embodiments.

[0129] Example 1

[0130] For the first vehicle, the process of calculating the vehicle's collision response space per unit time is as follows:

[0131] Step 1. Obtain the dynamic information of the first vehicle;

[0132] The dynamic information includes vehicle speed V and vehicle instantaneous acceleration α. r , μ is the coefficient of adhesion between the vehicle and the surrounding ground, and g is the gravitational acceleration of the current area where the first vehicle is located.

[0133] Step 2. Set a unit time T and calculate the collision response space of the first vehicle per unit time.

[0134] The collision response space of a vehicle per unit time is the envelope of a series of complex curves, such as... Figure 4 As shown, the vehicle's collision response space per unit time is enclosed by curve 1, curve 2, curve 3, and line segment 4.

[0135] Curves 1 and 2 represent the maximum combined centrifugal acceleration and deceleration caused by the resistance of the first vehicle. In other words, this curve is one whose radius of curvature *r* decreases with the length of curve *s*. Figure 4 In the diagram, curve 1 represents the left turn curve for a vehicle, and curve 2 represents the right turn curve.

[0136] The calculation formulas for curve 1 and curve 2 are as follows:

[0137]

[0138]

[0139] Curve number 3 is an ellipse shape, as shown below. Figure 4 As shown, the vertices of the ellipse are the turning vertices of curves 1 and 2, and the distance P from the point where the vehicle stops at its maximum deceleration along the straight line. The major axis of the ellipse is half the distance between the turning vertices of curves 1 and 2, and the minor axis is the distance from point P to the turning vertices of curves 1 and 2. The formula for the distance x from the vehicle to point P is as follows:

[0140]

[0141] In the above formula, x is the distance from the point of maximum deceleration along a straight line to a stop, and ar Let T be the instantaneous acceleration of the vehicle, T be the time it takes for the first vehicle to come to a stop along a straight line using maximum deceleration, and t be the travel time of the first vehicle.

[0142] The 4th line segment set encompasses the vehicle's signal response space, which is the speed at which the vehicle makes a decision after receiving information from its surroundings. This speed depends on the vehicle's hardware. The 4th line segment set is defined by the starting points of curves 1 and 2, as well as the vehicle's left and right front vertices. Based on the above calculations, the collision response space per unit time includes the vehicle's signal response space.

[0143] Step 3. Calculate the collision response space ε of other vehicles per unit time within the first vehicle warning space according to the relevant calculation method in Step 2.

[0144] The method for determining driving strategies for autonomous vehicles disclosed in this embodiment calculates the spatial interaction entropy and the sum of repulsive force vectors of the currently driven vehicle. Since both spatial interaction entropy and the sum of repulsive force vectors can reflect the ability to predict the probability of potential accidents in a defensive driving strategy, defensive driving decisions are made accordingly to achieve driving decisions for the next unit of time, thereby improving driving safety and driving stability.

[0145] It will be understood by those skilled in the art that all or some of the steps, systems, or apparatuses disclosed above, and their functional modules / units, can be implemented as software, firmware, hardware, or suitable combinations thereof. In hardware implementations, the division between functional modules / units mentioned above does not necessarily correspond to the division of physical components; for example, a physical component may have multiple functions, or a function or step may be performed collaboratively by several physical components. Some or all components may be implemented as software executed by a processor, such as a digital signal processor or microprocessor, or as hardware, or as an integrated circuit, such as an application-specific integrated circuit (ASIC). Such software may be distributed on a computer-readable medium, which may include computer storage media (or non-transitory media) and communication media (or transient media). As is known to those skilled in the art, the term computer storage media includes volatile and non-volatile, removable and non-removable media implemented in any method or technology for storing information (such as computer-readable instructions, data structures, program modules, or other data). Computer storage media include, but are not limited to, RAM, ROM, EEPROM, flash memory or other memory technologies, CD-ROM, digital versatile disc (DVD) or other optical disc storage, magnetic cartridges, magnetic tape, disk storage or other magnetic storage devices, or any other medium that can be used to store desired information and can be accessed by a computer. Furthermore, it is well known to those skilled in the art that communication media typically contain computer-readable instructions, data structures, program modules, or other data in modulated data signals such as carrier waves or other transmission mechanisms, and may include any information delivery medium.

Claims

1. A method for determining a driving strategy for an autonomous vehicle, characterized in that, The method includes: Obtain dynamic, static, and spatial information of the first vehicle; The collision response space of the first vehicle per unit time is determined based on the dynamic information and static information of the first vehicle. Within the collision reaction space of the first vehicle per unit time, the spatial interaction entropy between the first vehicle and other vehicles is calculated, and the spatial interaction entropy includes: lateral spatial interaction entropy and longitudinal spatial interaction entropy; the other vehicles include multiple second vehicles; If the spatial interaction entropy is less than the set threshold, then the predetermined driving strategy is executed; The predetermined driving strategy is determined based on the spatial information; The step of calculating the spatial interaction entropy between the first vehicle and other vehicles within the collision response space of the first vehicle per unit time includes: Based on grid partitioning, the unit-time collision response space of the first vehicle is divided into... A 2D sub-region; Calculate the ratio of the intersection area of ​​the sub-regions in the overlapping interval between the collision response space of the second vehicle and the collision response space of the first vehicle per unit time to the original area of ​​the sub-region. This ratio is expressed as... ; Calculate the ratio of the non-intersecting area of ​​each sub-region within the overlapping region of the collision response space of the second vehicle and the collision response space of the first vehicle per unit time to the original area of ​​the sub-region. This ratio is expressed as... ; The lateral spatial interaction entropy and longitudinal spatial interaction entropy of the first vehicle are calculated based on the ratio of the intersection area of ​​the sub-regions in the overlapping area between the collision response space of the second vehicle and the collision response space of the first vehicle per unit time to the original area of ​​the sub-regions, and the ratio of the non-intersecting area of ​​each sub-region in the overlapping area between the collision response space of the second vehicle and the collision response space of the first vehicle per unit time to the original area of ​​the sub-regions. The formula for calculating the vertical spatial interaction entropy is: ； The formula for calculating the transverse spatial interaction entropy is: ； in, For vertical spatial interaction entropy, For horizontal spatial interaction entropy, The minimum value is -1. The minimum value is -1, k is the total number of second vehicles in the warning space, j is the number of 2D regions divided into the collision reaction space per unit time, and m and n are the starting values ​​for calculating spatial interaction entropy. The area of ​​the collision response space of the first vehicle per unit time; the warning space is the maximum spatial area that the sensors of the first vehicle can perceive.

2. The method for determining a driving strategy for an autonomous vehicle according to claim 1, characterized in that, The static information includes: acquiring environmental information within the collaborative space and warning space, calculating the static environmental reaction space, and determining the location information of static obstacles located within the static environmental reaction space or affecting vehicle collision avoidance operations. The static environmental reaction space is the 2D area occupied by the static obstacles. The dynamic information includes: vehicle speed, vehicle instantaneous acceleration, vehicle centrifugal acceleration, vehicle adhesion coefficient to the surrounding ground, and gravitational acceleration of the area where the vehicle is located. The spatial information includes: information about the collaborative space and information about the warning space; The cooperative space is the minimum spatial area required for the first vehicle and other vehicles to cooperate in completing a collision avoidance maneuver.

3. The method for determining a driving strategy for an autonomous vehicle according to claim 2, characterized in that, The collision response space of the first vehicle per unit time is a two-dimensional region composed of the first curve, the second curve, the third curve, and the fourth straight line; Among them, the first curve and the second curve are the circular arc curves of the maximum steering trajectory; The third curve is elliptical in shape, with the vertices of the first curve, the turning vertices of the second curve, and the distance point at which the first vehicle stops along a straight line with maximum deceleration, respectively. The major axis of the ellipse is half the distance between the turning vertices of the first and second curves, and the minor axis of the ellipse is the distance from the distance point to the turning vertices of the first and second curves. The fourth straight line is the trajectory of the vehicle when its centrifugal acceleration is 0 and it travels at its maximum deceleration.

4. The method for determining a driving strategy for an autonomous vehicle according to claim 3, characterized in that, The method for determining the first curve and the second curve is as follows: Based on the vehicle's speed, instantaneous acceleration, adhesion coefficient between the vehicle and the surrounding ground, and gravitational acceleration of the area where the vehicle is located, the radius of curvature and arc length of the maximum steering trajectory of the vehicle are determined using a preset calculation formula. The preset calculation formula is as follows: ， ； In the above formula, For vehicle speed, For the instantaneous acceleration of the vehicle, The coefficient of adhesion between the vehicle and the surrounding ground. The gravitational acceleration of the area where the vehicle is located. Let be the radius of curvature of the arc. Let be the arc length of the circle.

5. The method for determining a driving strategy for an autonomous vehicle according to claim 4, characterized in that, The method for determining the third curve is as follows: The distance is determined based on the turning apex of the first curve and the second curve, and the distance from which the first vehicle stops along a straight line with maximum deceleration. Among them, the first vehicle travels along a straight line at the distance from the point where it stops with maximum deceleration. for: In the above formula, Let this be the distance the vehicle travels along a straight line until it stops with maximum deceleration. For the instantaneous acceleration of the vehicle, T For the first vehicle traveling in a straight line, using maximum deceleration, the time from the start of deceleration to coming to a complete stop is given. t This refers to the time it takes for the first vehicle to travel.

6. The method for determining a driving strategy for an autonomous vehicle according to claim 5, characterized in that, The method for determining the fourth straight line is as follows: Acquire the vehicle's decision-making response time from receiving surrounding information to initiating collision avoidance maneuvers; The fourth set of straight lines is the signal response space, which is the area enclosed by the starting points of the first and second curves and the left and right front vertices of the first vehicle. The longitudinal distance of the enclosed area is the width of the first vehicle, and the lateral distance is the displacement of the first vehicle within the decision response time.

7. The method for determining a driving strategy for an autonomous vehicle according to claim 6, characterized in that, The predetermined driving strategy is determined based on the spatial information and includes: The predetermined driving strategy is determined based on the vector sum of the cooperative space repulsive force and the warning space repulsive force; The repulsive forces include longitudinal repulsive forces and transverse repulsive forces. The longitudinal repulsive force for: ； The lateral repulsive force for: ； In the above formula, It is a longitudinal repulsive force. It is a transverse repulsive force. To warn of the repulsive force weights in the space, For the repulsive force weights in the cooperative space, The velocity component perpendicular to the lane line. The velocity component is in the direction parallel to the lane line. , These are the lateral and longitudinal velocity components of the second vehicle, respectively. d Indicates the distance between vehicles. k This is to indicate the total number of second vehicles in the warning space.

8. The method for determining a driving strategy for an autonomous vehicle according to claim 7, characterized in that, The predetermined driving strategy includes adjusting the vehicle's steering wheel angle or adjusting the vehicle's acceleration; The steering wheel angle is: ; The acceleration is: ; In the above formula, For steering wheel angle, To accelerate the vehicle, It is a longitudinal repulsive force. It is a transverse repulsive force. and All of these are pre-calibrated values.

9. A device for determining a driving strategy for an autonomous vehicle, characterized in that, The apparatus includes a memory and a processor; wherein the memory is used to store a program for a method of determining a driving strategy for an autonomous vehicle, and the processor is used to execute the method according to any one of claims 1-8.

Images