Speed decision-making and planning method and device based on interaction intention prediction, mobile tool and storage medium

Abstract

The invention discloses a speed decision-making and planning method and device based on interaction intention prediction, and the method comprises the steps: obtaining an obstacle ST risk field, and determining an interaction obstacle according to the obstacle ST risk field; the behavior intention of the interaction obstacle is analyzed, and behavior intention information of the interaction obstacle is determined; and updating the ST risk field of the obstacle according to the behavior intention information of the interactive obstacle, and performing speed decision and planning based on the updated ST risk field of the obstacle. According to the scheme of the invention, speed decision making and planning are carried out in combination with scene characteristics in a scene in which interaction with traffic participants exists, and the influence of interaction obstacles possibly existing in the current scene and behavior intentions of the interaction obstacles on speed decision making and planning results is fully considered. The adverse effect on the speed planning result caused by real-time prediction of the trajectory and intention jitter is avoided, and the interaction ability of the system and the comfort and intelligence of automatic driving are improved.

Description

Technical Field

[0001] The present invention relates to the field of autonomous driving technology, and in particular to a speed decision-making and planning method based on interaction intention prediction, a speed decision-making and planning device based on interaction intention prediction, and a mobile tool and a storage medium including the speed decision-making and planning device based on interaction intention prediction. Background Art

[0002] Artificial intelligence is the theory, method, technology and application system that uses digital computers or machines controlled by digital computers to simulate, extend and expand human intelligence, perceive the environment, acquire knowledge and use knowledge to obtain the best results. Autonomous driving is a mainstream application in the field of artificial intelligence. Autonomous driving technology relies on the collaboration of computer vision, radar, monitoring devices and global positioning systems to enable motor vehicles to achieve autonomous driving without the need for active human operation.

[0003] In the field of autonomous driving technology, the main key technologies currently include autonomous positioning, environmental perception, behavioral decision-making, trajectory planning, motion control, etc. Among these key technologies, speed decision-making and planning, as an important part of autonomous driving technology, directly affect the safety and comfort of autonomous driving. At present, speed decision-making and planning are usually implemented by using search or optimization algorithms based on the ST graph or risk field generated by the predicted trajectory of obstacles. However, the existing prediction algorithms mainly predict the intention and trajectory of each obstacle based on the perception data at the current and historical moments, the positioning information of the vehicle, and the high-precision map. In fact, during the driving process of the vehicle, the vehicle often interacts frequently with other traffic participants, such as in the scenarios of the front vehicle cutting in, pedestrians crossing, and crossing the intersection. In this scenario, traffic participants often adjust their behavior according to the intention of the interactor to avoid conflicts with other dynamic targets. However, the existing prediction algorithms do not consider the interaction scenarios between the vehicle and other traffic participants, which does not conform to the real driving habits. Therefore, speed decision-making and planning based directly on the ST graph or risk field generated by such prediction results may affect the comfort of the speed decision-making and planning results due to the prediction intention or trajectory jitter, and there is even a risk of not being able to search for a feasible solution, affecting the safety of autonomous driving. Summary of the invention

[0004] An embodiment of the present invention provides a speed decision and planning solution based on interactive intention prediction to solve the problem in the prior art that speed decision and planning directly based on the ST map or risk field generated by the obstacle prediction trajectory may affect the comfort of the speed decision and planning results and the safety of autonomous driving due to the predicted intention or trajectory jitter.

[0005] In a first aspect, an embodiment of the present invention provides a speed decision and planning method based on interaction intention prediction, which includes:

[0006] Obtaining an obstacle ST risk field, and determining an interactive obstacle according to the obstacle ST risk field;

[0007] Analyze the behavioral intention of the interactive obstacle and determine the behavioral intention information of the interactive obstacle;

[0008] According to the behavior intention information of the interactive obstacle, the obstacle ST risk field is updated, and speed decision and planning are performed based on the updated obstacle ST risk field.

[0009] In a second aspect, an embodiment of the present invention provides another speed decision and planning method based on interaction intention prediction, which includes:

[0010] Obtaining an obstacle ST risk field, and determining an interactive obstacle according to the obstacle ST risk field;

[0011] Analyze the behavioral intention of the interactive obstacle and determine the behavioral intention information of the interactive obstacle;

[0012] Determine the behavior intention acceleration of the ego vehicle under different behavior intentions;

[0013] According to the behavior intention information of the interactive obstacle and the behavior intention acceleration of the ego vehicle, speed decision and planning are performed directly based on the obstacle ST risk field or based on the updated obstacle ST risk field.

[0014] In a third aspect, an embodiment of the present invention provides a speed decision and planning device based on interaction intention prediction, which includes:

[0015] An interactive obstacle screening module, used to obtain an obstacle ST risk field and determine interactive obstacles according to the obstacle ST risk field;

[0016] The obstacle intention determination module is used to analyze the behavior intention of the interactive obstacle and determine the behavior intention information of the interactive obstacle;

[0017] The ego vehicle intention determination module is used to determine the behavior intention acceleration of the ego vehicle under different behavior intentions;

[0018] The speed decision and planning module is used to make speed decisions and plans directly based on the obstacle ST risk field or based on the updated obstacle ST risk field according to the behavior intention information of the interactive obstacle and the behavior intention acceleration of the vehicle.

[0019] In a fourth aspect, an embodiment of the present invention provides another speed decision and planning device based on interaction intention prediction, which includes:

[0020] A memory for storing executable instructions; and

[0021] The processor is used to execute the executable instructions stored in the memory, and the executable instructions implement the steps of the above method when executed by the processor.

[0022] In a fifth aspect, the present invention provides a storage medium having a computer program stored thereon, which implements the steps of the above method when executed by a processor.

[0023] In a sixth aspect, the present invention provides a mobile tool, which includes the speed decision and planning device based on interaction intention prediction as described in the fourth aspect above.

[0024] In the seventh aspect, the present invention provides a computer program product comprising a computer program stored on a non-volatile computer-readable storage medium, wherein the computer program comprises program instructions, and when the program instructions are executed by a computer, the computer executes the above-mentioned speed decision and planning method based on interaction intention prediction.

[0025] The beneficial effect of the embodiments of the present invention lies in that: the method provided by the embodiments of the present invention screens out interactive obstacles from the obstacle ST risk field, and based on the behavioral intention analysis of the interactive obstacles, updates the already constructed obstacle ST risk field when necessary, thereby realizing speed decision and planning based on a pre-constructed obstacle ST risk field or an updated obstacle ST risk field in combination with the characteristics of the scene in a scenario where there is interaction with traffic participants, fully considering the impact of possible interactive obstacles and their behavioral intentions on the obstacle ST risk field and speed decision and planning results in the current scenario, avoiding the adverse effects on the speed planning results due to jitter of real-time predicted trajectory and intention, and improving the interactive capability of the system and the comfort and intelligence of autonomous driving. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying any creative work.

[0027] Figure 1 A flow chart of a speed decision and planning method based on interaction intention prediction according to an embodiment of the present invention;

[0028] Figure 2 One embodiment of the present invention Figure 1 Flow chart of a method for implementing operation S11 in FIG.

[0029] Figure 3 One embodiment of the present invention Figure 2 Flow chart of a method for implementing operation S111 in FIG.

[0030] Figure 4 One embodiment of the present invention Figure 3 Flow chart of a method for implementing operation S111B in FIG.

[0031] Figure 5 One embodiment of the present invention Figure 3 FIG. 1 is a diagram showing a method for implementing operation S111C;

[0032] Figure 6 Another embodiment of the present invention Figure 1 Flow chart of a method for implementing operation S11 in FIG.

[0033] Figure 7 A flow chart of a method for updating the obstacle ST risk field according to the behavior intention information of the interactive obstacle according to an embodiment of the present invention is shown;

[0034] Figure 8 A flow chart showing a method for updating the obstacle ST risk field according to the behavior intention information of the interactive obstacle according to another embodiment of the present invention;

[0035] Fig. 9 A flowchart of a method for updating the obstacle ST risk field according to the behavior intention information of the interactive obstacle according to another embodiment of the present invention is shown;

[0036] Fig.10 This is a flow chart of a speed decision and planning method based on interaction intention prediction according to another embodiment of the present invention;

[0037] Fig.11 This is a display effect diagram of the ST diagram of the vehicle in the obstacle frenet coordinate system constructed according to an embodiment of the present invention;

[0038] Fig.12 The principle block diagram of a speed decision and planning device based on interaction intention prediction according to an embodiment of the present invention is schematically shown;

[0039] Fig.13 The principle block diagram of a speed decision and planning device based on interaction intention prediction according to another embodiment of the present invention is schematically shown;

[0040] Fig.14 The principle block diagram of a speed decision and planning device based on interaction intention prediction according to another embodiment of the present invention is schematically shown;

[0041] Fig.15 A principle block diagram of a control device according to an embodiment of the present invention;

[0042] Fig.16It is a structural schematic diagram of an embodiment of a speed decision and planning device based on interaction intention prediction of the present invention. DETAILED DESCRIPTION

[0043] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0044] It should be noted that, in the absence of conflict, the embodiments and features in the embodiments of the present application may be combined with each other.

[0045] The present invention may be described in the general context of computer-executable instructions executed by a computer, such as program modules. Generally, program modules include routines, programs, objects, components, data structures, etc. that perform specific tasks or implement specific abstract data types. The present invention may also be practiced in distributed computing environments where tasks are performed by remote processing devices connected through a communications network. In a distributed computing environment, program modules may be located in local and remote computer storage media, including storage devices.

[0046] In the present invention, "module", "device", "system" and the like refer to related entities applied to computers, such as hardware, a combination of hardware and software, software or software in execution, etc. In detail, for example, an element can be, but is not limited to, a process, a processor, an object, an executable element, an execution thread, a program and / or a computer running on a processor. In addition, an application or script program running on a server, a server can all be an element. One or more elements can be in an execution process and / or thread, and an element can be localized on a computer and / or distributed between two or more computers, and can be operated by various computer-readable media. An element can also communicate through local and / or remote processes according to a signal with one or more data packets, for example, a signal from a data that interacts with another element in a local system, a distributed system, and / or a network on the Internet through a signal to interact with other systems.

[0047] Finally, it should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include" and "comprise" include not only those elements, but also other elements that are not explicitly listed, or also include elements inherent to such processes, methods, articles or equipment. In the absence of further restrictions, the elements defined by the statement "include..." do not exclude the existence of other identical elements in the process, method, article or equipment that includes the elements.

[0048] The speed decision and planning method based on interaction intention prediction in the embodiment of the present invention can be applied to a speed decision and planning device based on interaction intention prediction, so that a user or an autonomous driving device can use the speed decision and planning device based on interaction intention prediction to obtain the decision result and planning trajectory for the current moment, so as to control the autonomous driving device to perform corresponding driving actions according to the decision result and planning trajectory.

[0049] These speed decision-making and planning devices based on interaction intention prediction can be specifically applied to mobile tools in the form of autonomous driving domain controllers. For example, the mobile tool is equipped with various sensors for sensing environmental information and the mobile tool's own information (including but not limited to laser radar, camera, millimeter wave radar, ultrasonic radar, inertial measurement unit, wheel speed meter, etc.), and the autonomous driving domain controller is connected to these sensors for communication, and performs path planning for the mobile tool based on the environmental information collected by these sensors and the mobile tool's own information, so as to achieve the purpose of controlling the mobile tool to travel automatically. Among them, the mobile tool may include but is not limited to autonomous driving vehicles (such as passenger cars, buses, minibuses, buses, trucks, sanitation vehicles, sweepers, scrubbers, etc.), sweeping robots, motorcycles, electric vehicles and other autonomous driving equipment, and the present invention does not limit this.

[0050] Figure 1 The schematic diagram shows a speed decision and planning method based on interaction intention prediction according to an embodiment of the present invention. The execution subject of the method can be a planner or controller on an autonomous driving vehicle, or a speed decision and planning device based on interaction intention prediction, or a processor of a mobile tool such as an unmanned cleaning vehicle, an unmanned sweeping vehicle, a sweeping robot, an autonomous driving vehicle, etc. The embodiment of the present invention is not limited to this. The embodiment of the present invention is first described in detail by taking the execution subject as a planner on an autonomous driving vehicle as an example. Figure 1 As shown, the method of the embodiment of the present invention includes:

[0051] Operation S10, obtaining an obstacle ST risk field, and determining an interactive obstacle according to the obstacle ST risk field;

[0052] Operation S11, analyzing the behavior intention of the interactive obstacle to determine the behavior intention information of the interactive obstacle; and

[0053] Operation S12: updating the obstacle ST risk field according to the behavior intention information of the interactive obstacle, and performing speed decision and planning based on the updated obstacle ST risk field.

[0054] In operation S10, the obstacle ST risk field may refer to a pre-constructed obstacle ST risk field obtained from a front module, or may refer to an obstacle ST risk field formed by updating the method of the embodiment of the present invention. The front module may be, for example, a controller connected to a planner on an autonomous driving vehicle or a submodule or processor of the planner for constructing an obstacle ST risk field. As a preferred embodiment, when the method of the embodiment of the present invention is initially executed, the obstacle ST risk field may be a pre-constructed obstacle ST risk field of all dynamic obstacles around the self-vehicle directly obtained from the front module. In subsequent processing, the obstacle ST risk field preferably refers to an obstacle ST risk field of all dynamic obstacles around the self-vehicle formed by updating the method of the embodiment of the present invention. Since the obstacle ST risk field formed in the previous round of processing is already a priori information for the next round of processing, the updated obstacle ST risk field may be stored in the previous round of processing to serve as input information for the next round of processing. Among them, in the embodiment of the present invention, the obstacle ST risk field is an ST probability grid map marked with the occupancy probability of each grid point occupied by a dynamic obstacle, and the ST probability grid map can be identified by (s, t, P(s, t)), wherein (s, t) represents the grid point in the ST risk field, and P(s, t) represents the occupancy probability of the grid point (s, t) in the ST risk field being occupied by a dynamic obstacle. Therefore, the obstacle ST risk field of the embodiment of the present invention can effectively describe the relationship between each grid point (s, t) and the occupancy probability P of the dynamic obstacle in the obstacle ST calculation result area. Exemplarily, the obstacle ST risk field of the embodiment of the present invention can be specifically implemented as an obstacle ST probability grid map established based on the Frenet coordinate system, that is, it is implemented as a grid method representation based on the Frenet coordinate system (also called the road coordinate system). Therefore, the obstacle ST risk field in the embodiment of the present invention specifically uses a grid to represent the spatial position of the dynamic obstacle in the grid map, that is, the position in the structured road. It should be noted that in an embodiment of the present invention, the constructed obstacle ST risk field and the method of constructing the obstacle ST risk field can be implemented using existing technologies. For example, the method for constructing a dynamic obstacle ST risk field described in the invention patent application with application number 202210272795.6 can be used to construct the obstacle ST risk field of an embodiment of the present invention.

[0055] In the embodiment of the present invention, interactive obstacles refer to conflicting obstacles that can observe the state of the ego vehicle and adjust the behavior intention according to the ego vehicle behavior, and after eliminating the conflicting obstacles that are not affected by the ego vehicle and high road rights obstacles, the dynamic obstacles here refer to traffic participants other than the ego vehicle that will move, which can be pedestrians, animals, or other vehicles. As an implementation method, in operation S10, conflicting obstacles can be first screened out according to the position of the obstacles in the obstacle ST risk field and their risk level, and then obstacles that need to be interactively predicted are determined as interactive obstacles based on the screened conflicting obstacles. As mentioned above, since the obstacle ST risk field is the ST probability grid map of the obstacle, it is actually a risk field described by several grids with occupancy probabilities. The occupancy probability of each grid can describe the risk level of the obstacle at different grid positions. Therefore, screening out conflicting obstacles according to the position of the obstacle in the obstacle ST risk field and its risk level can be specifically implemented as follows: judging whether there are grids with an occupancy probability greater than a set threshold in the ST risk field of the corresponding obstacle and whether there are grids located in the first quadrant of the Frenet coordinate system of the ST risk field among the grids with an occupancy probability greater than the set threshold. If such grids exist, the obstacle is determined as a conflicting obstacle. Exemplarily, the threshold value can be set to 0.8. As a possible implementation, determining obstacles that need to be interacted based on the screened conflicting obstacles can be specifically implemented by removing obstacles that are not affected by the vehicle and have high road rights from the screened conflicting obstacles, and treating the remaining conflicting obstacles as interactive obstacles. Among them, an obstacle with high road rights refers to an obstacle with a road right higher than that of the vehicle itself determined according to traffic rules. For example, according to traffic rules, the road right of a straight-moving vehicle is higher than that of a left-turning vehicle. Then, when the obstacle is a straight-moving vehicle and the vehicle itself is turning left, the obstacle is an obstacle with high road rights. It can be determined whether the selected conflicting obstacles are obstacles with high road rights based on traffic rules, the planned trajectory of the vehicle itself and the predicted trajectory of the obstacle. Obstacles not affected by the vehicle itself refer to obstacles that will not actively pay attention to the vehicle itself, such as the vehicle in front of the vehicle itself in the lane of the vehicle itself is an obstacle not affected by the vehicle itself. It can be determined whether each conflicting obstacle is an obstacle not affected by the vehicle itself according to traffic rules and driving habits, combined with the planned trajectory of the vehicle itself and the predicted trajectory of the obstacle. In another embodiment, interactive obstacles can also be determined from the selected conflicting obstacles according to civilized driving rules, wherein the civilized driving rules can be set as not to actively compete with pedestrians, non-motorized vehicles and vehicles with high road rights. Among them, vehicles with high road rights are vehicles with road rights higher than that of the vehicle itself determined according to traffic rules.It should be noted that when there is more than one conflicting obstacle determined in operation S10, it is necessary to perform the above-mentioned judgment on each conflicting obstacle separately to further screen out the obstacles to be interactively predicted, and when there is more than one interactive obstacle determined, it is necessary to perform behavioral intention analysis on each interactive obstacle separately and update the obstacle ST risk field of the corresponding interactive obstacle based on the results of the behavioral intention analysis.

[0056] In operation S11, the behavior intention information may be implemented as a behavior intention probability including the interactive obstacle. In the embodiment of the present invention, the behavior intention probability is used to characterize the similarity between the current acceleration of the interactive obstacle and the expected acceleration for realizing the corresponding behavior intention. The higher the behavior intention probability, the closer the current acceleration is to the expected acceleration required for realizing the corresponding behavior intention, which means that the corresponding behavior intention is more likely to be achieved. In some preferred implementations, the behavior intention of the interactive obstacle can be set to include yielding to the own vehicle and overtaking the own vehicle. Taking this as an example, in operation S11, the behavior intention probability specifically includes a first behavior intention probability for characterizing the behavior intention of yielding to the own vehicle and a second behavior intention probability for characterizing the behavior intention of overtaking the own vehicle. The first behavior intention probability characterizes the similarity between the acceleration of the interactive obstacle at the current moment and the expected acceleration required to realize the behavior intention of yielding to the own vehicle, reflecting the possibility that the interactive obstacle can achieve the behavior intention of yielding to the own vehicle based on the acceleration at the current moment. The higher the value, the greater the possibility that the interactive obstacle will execute the behavior intention of yielding to the own vehicle. The second behavior intention probability characterizes the similarity between the acceleration of the interactive obstacle at the current moment and the expected acceleration required to realize the behavior intention of overtaking the own vehicle, reflecting the possibility that the interactive obstacle can achieve the behavior intention of overtaking the own vehicle based on the acceleration at the current moment. The higher the value, the greater the possibility that the interactive obstacle will execute the behavior intention of overtaking the own vehicle. Therefore, by analyzing the behavioral intentions of interactive obstacles and determining their behavioral intention information, it is possible to predict the interactive intentions of interactive obstacles, and then adjust the speed decision and planning results based on the prediction of the interactive intentions, so that the speed decision and planning methods are closer to the human processing methods, thereby improving the intelligence and rationality of speed decision and planning.

[0057] As a possible implementation, in operation S11, the behavior intention analysis of the interactive obstacle can be implemented based on the predicted trajectory of the interactive obstacle and the planned trajectory of the vehicle. For example, Figure 2 The specific implementation process of operation S11 in one embodiment is schematically shown. Figure 2 As shown, its implementation includes:

[0058] Operation S111, determining a first minimum acceleration required for the interactive obstacle to safely overtake the ego vehicle and a first maximum acceleration required for safely giving way to the ego vehicle based on the predicted trajectory of the interactive obstacle and the planned trajectory of the ego vehicle; and

[0059] Operation S112: determining a behavior intention probability of the interactive obstacle at a current moment according to the current acceleration, the first minimum acceleration, and the first maximum acceleration of the interactive obstacle.

[0060] Among them, the predicted trajectory of the interactive obstacle and the planned trajectory of the self-vehicle can be obtained from the front module, such as the planner, so it will not be elaborated on in detail here. In the embodiment of the present invention, the first minimum acceleration is the minimum acceleration required for the interactive obstacle to safely overtake the self-vehicle, that is, when the acceleration of the interactive obstacle reaches the first minimum acceleration, the self-vehicle can be safely overtaken, that is, the behavioral intention of safely overtaking the self-vehicle can be achieved; the first maximum acceleration is the maximum acceleration required for the interactive obstacle to safely give way to the self-vehicle, that is, when the acceleration of the interactive obstacle does not exceed the first maximum acceleration, the self-vehicle can be safely given way, that is, the behavioral intention of safely giving way to the self-vehicle can be achieved. Since the behavior intention probability can characterize the possibility of the interactive obstacle achieving the corresponding behavior intention, and the first minimum acceleration and the first maximum acceleration can also reflect the possibility of the interactive obstacle safely achieving the corresponding behavior intention, therefore, as a possible implementation method, the first minimum acceleration can be used as the expected acceleration to achieve the behavior intention of overtaking the ego vehicle, and the first maximum acceleration can be used as the expected acceleration to achieve the behavior intention of giving way to the ego vehicle, and the behavior intention probability of the interactive obstacle can be determined based on the acceleration of the interactive obstacle at the current moment (also referred to as "current acceleration" in the embodiment of the present invention). Specifically, in operation S111, the ST diagram of the ego vehicle in the obstacle Frenet coordinate system can be constructed based on the predicted trajectory of the interactive obstacle and the planned trajectory of the ego vehicle, and the first minimum acceleration and the first maximum acceleration can be determined based on the ST diagram. Figure 3 A specific implementation of operation S111 is schematically shown, as follows: Figure 3 As shown, operation S111 may be specifically implemented as including:

[0061] Operation S111A, constructing an ST graph of the ego vehicle in the obstacle Frenet coordinate system according to the predicted trajectory of the interactive obstacle and the planned trajectory of the ego vehicle;

[0062] Operation S111B, determining a first minimum acceleration required for the interactive obstacle to safely overtake the vehicle according to the upper boundary of the ST graph and the current speed of the interactive obstacle;

[0063] Operation S111C: determining a first maximum acceleration required for the interactive obstacle to safely yield to the vehicle according to the lower boundary of the ST graph and the current speed of the interactive obstacle.

[0064] In operation S111A, the ST diagram of the ego vehicle in the obstacle Frenet coordinate system is a distance-time diagram between the ego vehicle and the interactive obstacle established in the Frenet coordinate system with the interactive obstacle position as the origin, wherein the horizontal axis is time T and the vertical axis is the longitudinal distance between the ego vehicle and the interactive obstacle at different times T. The ST diagram of the ego vehicle in the obstacle Frenet coordinate system based on the predicted trajectory of the interactive obstacle and the planned trajectory of the ego vehicle can be realized by referring to the relevant existing technologies, which will not be described in detail here. The upper boundary of the ST diagram thus constructed represents the minimum limit to which the interactive obstacle can overtake the ego vehicle, and the lower boundary represents the maximum limit to which the interactive obstacle can give way to the ego vehicle. Taking the straight-ahead conflict scenario of the ego vehicle at the intersection as an example, the ST diagram of the ego vehicle in the obstacle Frenet coordinate system is established as follows: Fig.11 As shown, the lower boundary in the ST graph is the upper boundary of the planning for the obstacle to give way to the vehicle, and the upper boundary in the ST graph is the lower boundary of the planning for the obstacle to overtake the vehicle. Based on this, in operations S111B and S111C, the minimum acceleration value required to reach the upper boundary and the maximum acceleration value required to reach the lower boundary can be determined based on the upper and lower boundaries of the ST graph and the current speed of the interactive obstacle. Figure 4 The specific implementation of operation S111B in one embodiment is schematically shown, as shown in FIG. Figure 4 As shown, it can be implemented as including:

[0065] Operation S111B1, sampling is performed on the upper boundary of the ST graph to obtain a first sampling point set;

[0066] Operation S111B2, determining the average acceleration of the interactive obstacle to each sampling point in the first sampling point set according to the current speed of the interactive obstacle and the s value and t value of each sampling point in the first sampling point set on the ST graph; and

[0067] Operation S111B3: taking the maximum value of the average accelerations of the interactive obstacle reaching each sampling point in the first sampling point set as the first minimum acceleration required for the interactive obstacle to safely overtake the ego vehicle.

[0068] In operation S111B1, the upper boundary may be sampled at a certain time interval, for example, at a time interval of 1 second, to obtain a sample including at least one sampling point (such as Fig.11 A first sampling point set of sampling points L in the ST diagram, wherein each sampling point is represented by a value pair (s, t), s is the value of the sampling point on the vertical axis on the ST diagram, i.e., the s value, and t is the value of the sampling point on the horizontal axis on the ST diagram, i.e., the t value.

[0069] In operation S111B2, the average acceleration of the interactive obstacle to each sampling point in the first sampling point set can be determined based on the displacement formula. For example, the sampling point is represented by a value pair (s, t), and the velocity of the interactive obstacle at the current moment is v. Specifically, the average acceleration a of the interactive obstacle to the sampling point (s, t) can be calculated by the following formula:

[0070]

[0071] In operation S111B3, the maximum value of the average acceleration of the interactive obstacle reaching all the upper boundary sampling points can be determined from the average acceleration of the interactive obstacle reaching each sampling point determined in operation S111B2. When the interactive obstacle reaches the maximum value, it can safely reach the upper boundary. Therefore, when the acceleration of the interactive obstacle is not less than the maximum value, it can safely overtake the self-vehicle. Therefore, the maximum value can be used as the first minimum acceleration required for the interactive obstacle to safely overtake the self-vehicle. Thus, when the interactive obstacle reaches or exceeds the first minimum acceleration, the behavioral intention of safely overtaking the self-vehicle can be achieved.

[0072] Figure 5 The specific implementation of operation S111C in one embodiment is schematically shown, as shown in FIG. Figure 5 As shown, it can be implemented as including:

[0073] Operation S111C1, sampling is performed on the lower boundary of the ST graph to obtain a second sampling point set;

[0074] Operation S111C2, determining the average acceleration of the interactive obstacle to each sampling point in the second sampling point set according to the current speed of the interactive obstacle and the s value and t value of each sampling point in the second sampling point set on the ST graph;

[0075] Operation S111C3: taking the minimum value of the average accelerations of the interactive obstacle reaching the sampling points in the second sampling point set as the first maximum acceleration required for the interactive obstacle to safely yield to the ego vehicle.

[0076] In operation S111C1, the lower boundary may be sampled at a certain time interval, for example, at a time interval of 1 second, to obtain a sample including at least one sampling point (such as Fig.11 A second set of sampling points (sampling points B in the ST diagram), wherein each sampling point is represented by a value pair (s, t), s is the value of the sampling point on the vertical axis on the ST diagram, i.e., the s value, and t is the value of the sampling point on the horizontal axis on the ST diagram, i.e., the t value.

[0077] In operation S111C2, the average acceleration of the interactive obstacle to each sampling point in the second sampling point set can be determined based on the displacement formula. For example, the sampling point is represented by a value pair (s, t), and the velocity of the interactive obstacle at the current moment is v. Specifically, the average acceleration a of the interactive obstacle to the sampling point (s, t) can be calculated by the following formula:

[0078]

[0079] In operation S111C3, the minimum value of the average acceleration of the interactive obstacle reaching all the lower boundary sampling points can be determined from the average acceleration of the interactive obstacle reaching each sampling point determined in operation S111C2. Since the interactive obstacle can safely reach the lower boundary only when it reaches the minimum value, the self-vehicle can be safely yielded when the acceleration of the interactive obstacle is not greater than the minimum value. Therefore, the minimum value can be used as the first maximum acceleration required for the interactive obstacle to safely yield to the self-vehicle. Therefore, when the interactive obstacle does not exceed the first maximum acceleration, the behavioral intention of safely yielding to the self-vehicle can be achieved.

[0080] Afterwards, in operation 112, the second behavior intention probability used to characterize the behavior intention of overtaking the self-vehicle can be determined according to the similarity between the current acceleration of the interactive obstacle and the first minimum acceleration, and the first behavior intention probability used to characterize the behavior intention of yielding to the self-vehicle can be determined according to the similarity between the current acceleration of the interactive obstacle and the first maximum acceleration. It can be understood that the closer the current acceleration is to the first minimum acceleration, the higher the possibility that the behavior intention of the interactive obstacle is to overtake the self-vehicle, and the closer the current acceleration is to the first maximum acceleration, the higher the possibility that the behavior intention of the interactive obstacle is to yield to the self-vehicle, thus, the first behavior intention probability and the second behavior intention probability can be set to be determined by the following formulas respectively:

[0081]

[0082]

[0083] Among them, P in the above formula y represents the probability of the first behavior intention, P o represents the probability of the second behavior intention, a represents the current acceleration of the interactive obstacle, and a + represents the first minimum acceleration, and a- represents the first maximum acceleration. Thus, when the current acceleration a is greater than or equal to a + , the probability of the second behavior intention P o In the calculation formula, The value of is greater than or equal to 1, the interactive obstacle is more likely to overtake the vehicle, and when the current acceleration a is less than or equal to a -When the probability of the first behavior intention is P y In the calculation formula, The value of is greater than or equal to 1, and the interactive obstacle has a higher probability of yielding to the vehicle. Therefore, according to the above formula, the current acceleration a is greater than or equal to a + , the probability of the second behavior intention is higher, indicating that the overtaking intention of the interactive obstacle is 1, the yielding intention is 0, and the current acceleration a is less than or equal to a - When , the value of the probability of the first behavior intention is higher, indicating that the overtaking intention of the interactive obstacle is 0 and the yielding intention is 1. Therefore, the behavior intention of the interactive obstacle can be predicted according to the values ​​of the first behavior intention probability and the second behavior intention probability. When the probability of the first behavior intention is higher, it means that it is more likely to yield to the own vehicle, and when the probability of the second behavior intention is higher, it means that it is more likely to overtake the own vehicle.

[0084] In the embodiment of the present invention, the behavior intention probability reflects the similarity between the acceleration of the interactive obstacle at the current moment and the expected acceleration of the corresponding behavior intention. In the early stage of the interaction between the interactive obstacle and the vehicle, the uncertainty of the prediction of the interactive obstacle's behavior intention is high. However, as the interaction time accumulates, the confidence in the interactive obstacle's behavior intention will gradually increase. Therefore, in the preferred implementation, in order to achieve a more accurate prediction of the interactive obstacle's behavior intention, the time-based confidence can be used to optimize and adjust the behavior intention probability to obtain a behavior intention probability that can more accurately reflect the interactive obstacle's behavior intention. Figure 6 The specific implementation method flow of operation S11 in another embodiment of the present invention is schematically shown. Figure 6 As shown, its implementation includes:

[0085] Operation S111, determining a first minimum acceleration required for the interactive obstacle to safely overtake the ego vehicle and a first maximum acceleration required for safely giving way to the ego vehicle based on the predicted trajectory of the interactive obstacle and the planned trajectory of the ego vehicle; and

[0086] Operation S112, determining a behavior intention probability of the interactive obstacle at a current moment according to the current acceleration, the first minimum acceleration, and the first maximum acceleration of the interactive obstacle;

[0087] Operation S113, determining an observation time for the interactive obstacle based on the vehicle perception data;

[0088] Operation S114, determining a conflict time with the interactive obstacle based on the predicted trajectory of the interactive obstacle and the planned trajectory of the ego vehicle;

[0089] Operation S115, determining the confidence of the behavior intention according to the observation time and the conflict time;

[0090] Operation S116 : updating the behavior intention probability of the interactive obstacle at the current moment according to the confidence level, and using the updated behavior intention probability as the behavior intention probability of the interactive obstacle at the current moment.

[0091] The specific implementation of operation S111 and operation S112 is as follows: Figure 2 The embodiments shown are the same, so reference can be made to the previous description, which will not be repeated here. In operation S113, the observation time specifically refers to the time from the detection of the interactive obstacle to the current moment, which is determined by the moment when the interactive obstacle is detected and the current moment. In a specific implementation, the moment when the interactive obstacle is detected can be obtained through the self-vehicle perception data, and the observation time can be determined in combination with the current moment, wherein the self-vehicle perception data can be obtained through a front module such as a perception module. In operation S114, the conflict time specifically refers to the time from the current moment to the moment when the predicted interactive obstacle and the self-vehicle begin to collide and conflict, which is determined by the current moment and the predicted collision and conflict moment of the interactive obstacle and the self-vehicle, wherein the predicted collision and conflict moment of the interactive obstacle and the self-vehicle can be the time of the predicted collision starting point during the collision between the interactive obstacle and the self-vehicle, which is determined based on the predicted trajectory of the interactive obstacle and the planned trajectory of the self-vehicle, and the determination method can be implemented with reference to the prior art, for example, with reference to the method for determining the time of the collision starting point described in the application number 202210272795.6, which is not limited in the embodiment of the present invention. Since the uncertainty of the prediction of the interactive obstacle behavior intention is high in the early stage of the interaction between the interactive obstacle and the ego vehicle, the confidence of the interactive obstacle behavior intention will gradually increase with the accumulation of interaction time. Therefore, the confidence can be set to be positively correlated with the conflict time and negatively correlated with the observation time to reflect the influence of the accumulation of interaction time on the confidence of the behavior intention. Specifically, in operation S115, the following formula can be used to determine the confidence of the interactive obstacle behavior intention:

[0092]

[0093] Among them, K in the formula represents the confidence of the determined behavioral intention of the interactive obstacle, t1 represents the observation time, and t2 represents the conflict time. The sum of t1 and t2 represents the time from the detection of the interactive obstacle to the actual occurrence of the conflict. When the obstacle prediction result is relatively accurate, the sum of t1 and t2 does not change much (that is, if it is predicted that a conflict will occur after t2=4s when t1=0s, if the prediction is accurate, then when t1=2s, it will be predicted that a conflict will occur after t2=2s, and the sum of t1 and t2 can basically remain stable), so K will gradually converge to 1.

[0094] After the confidence level is determined, in operation S116, the behavior intention probability can be updated based on the confidence level, so that the determined behavior intention probability can reflect the influence of the interaction time length on the accuracy of the predicted behavior intention, thereby improving the accuracy and stability of the prediction results of the behavior intention of the interactive obstacle. Specifically, since the confidence level reflects the interaction time length, and the longer the interaction time length, the higher the confidence level, and the more accurate the prediction results of the behavior intention, the behavior intention probability can be set to be positively correlated with the confidence level. For example, the behavior intention probability calculated by operation S112 can be updated by the following formula:

[0095] P y =max(min(K*P y ,1),P y )

[0096] P o =max(min(K*P o ,1),P o )

[0097] In order to enable the ego vehicle to lead the interaction relationship according to the situation, so as to gain more sufficient interaction space, and to make safer and more reasonable speed decisions and plans based on the prediction of the behavioral intention of the interactive obstacle when the predicted feasible solution is not good, the embodiment of the present invention will be based on the analysis of the behavioral intention of the interactive obstacle, and first update the obstacle ST risk field under the premise of safety and feasibility, and make speed decisions and plans based on the updated obstacle ST risk field, so as to increase the planning space of the ego vehicle and improve the success rate and rationality of the decision and planning results. As a possible implementation method, in the process of the ego vehicle leading the interaction relationship, it can be prioritized to determine whether the behavior of the ego vehicle overtaking the interactive obstacle is feasible. Therefore, the obstacle ST risk field can be updated based on the judgment of the probability of the first behavioral intention of the interactive obstacle, so as to make speed decisions and plans based on the updated obstacle ST risk field. Specifically, Figure 7 A specific implementation method flow of updating the obstacle ST risk field according to the behavior intention information of the interactive obstacle is shown, such as Figure 7 As shown, it can be implemented as including:

[0098] Operation S121, determining a first behavior intention probability of an interactive obstacle, and when the first behavior intention probability satisfies a first preset condition, executing operation S122;

[0099] Operation S122, determining a third acceleration required for the self-vehicle to yield to the interactive obstacle based on the first comfortable acceleration of the self-vehicle, and determining whether the third acceleration satisfies a second preset condition, and if so, executing operation S123;

[0100] Operation S123: updating the obstacle ST risk field based on the third acceleration.

[0101] In operation S121, the first preset condition is specifically a condition for judging whether the yielding behavior intention of the interactive obstacle is suitable for the self-vehicle to take overtaking behavior, that is, a condition for judging whether the overtaking behavior of the self-vehicle is safe. As a possible implementation method, the first preset condition can be specifically a preset minimum probability threshold, so that it can be determined whether the yielding behavior intention of the interactive obstacle is obvious by judging whether the probability of the first behavior intention of the interactive obstacle is greater than the minimum probability threshold, and then it can be determined whether the overtaking behavior of the self-vehicle is safe, that is, whether it is suitable for overtaking. When the probability of the first behavior intention is high, the yielding intention of the interactive obstacle is obvious, and it is safe and appropriate for the self-vehicle to take overtaking behavior at this time, otherwise there is a safety hazard. Therefore, by setting the minimum probability threshold and comparing the probability threshold of the first behavior intention of the interactive obstacle, it is possible to preliminarily judge the possibility of the self-vehicle taking overtaking behavior. When it is preliminarily judged that the self-vehicle has the possibility of taking overtaking behavior, subsequent processing will be performed, otherwise the self-vehicle needs to change the behavior intention of overtaking. Exemplarily, the minimum probability threshold can be set to 0.5.

[0102] In operation S122, when it is determined through operation S121 that the vehicle may take overtaking action, that is, the interactive obstacle may take the yielding action, the feasibility, rationality and safety of the vehicle taking the overtaking action will be further judged to improve the intelligence of the speed decision-making and planning method. As a possible implementation method, the feasibility, rationality and safety of the vehicle taking the overtaking action and the interactive obstacle taking the yielding action can be judged based on the first comfortable acceleration that the vehicle can take. Among them, the first comfortable acceleration is an acceleration that the vehicle can take under normal operating conditions according to experience, which will not make the driver and passengers feel uncomfortable. For example, it can be set to 1m / s 2 . Specifically, the feasibility and rationality of the interactive obstacle giving way to the own vehicle can be further judged when the own vehicle adopts the first comfortable acceleration to overtake the interactive obstacle. Therefore, in operation S122, the acceleration required for the interactive obstacle to give way to the own vehicle (referred to as the third acceleration in the embodiment of the present invention) can be determined based on the first comfortable acceleration of the own vehicle, and the feasibility of the interactive obstacle giving way to the own vehicle can be judged according to the rationality of the third acceleration, that is, the rationality of the prediction result of the interactive obstacle behavior intention and the feasibility of the current decision-making and planning based on the interactive prediction result are determined. Among them, the specific implementation method of determining the third acceleration required for the interactive obstacle to give way to the own vehicle based on the first comfortable acceleration of the own vehicle can be: setting the acceleration of the own vehicle to the first comfortable acceleration (such as 1m / s 2), reconstruct the ST diagram of the ego vehicle in the Frenet coordinate system of the obstacle, and calculate the first maximum acceleration required for the ego vehicle to give way to the interactive obstacle based on the reconstructed ST diagram, and use the recalculated first maximum acceleration as the third acceleration required for the ego vehicle to give way to the interactive obstacle. Among them, the ST diagram of the ego vehicle in the Frenet coordinate system of the obstacle constructed here is the same as the ST diagram described above, and its construction method is also the same, and the method for calculating the first maximum acceleration required for the ego vehicle to give way to the interactive obstacle based on the reconstructed ST diagram is also the same as the above, as can be referred to Figure 5 The method for calculation is used, so it will not be repeated here. After calculating the third acceleration required for the interactive obstacle to give way to the own vehicle, it is also necessary to judge the rationality and feasibility of the third acceleration, which can be specifically achieved by setting the second preset condition. In operation S122, when the third acceleration meets the second preset condition, it means that it is reasonable, feasible and safe for the interactive obstacle to give way to the own vehicle. Therefore, the second preset condition can be set according to the reasonable acceleration range that the interactive obstacle can reach. For example, the second preset condition can be set to not exceed the reasonable acceleration range of the interactive obstacle. For example, the second preset condition can be set to not exceed (-2m / s 2 ~1.5m / s 2), where the range is the acceleration range of the interactive obstacle under the normal driving state. When the third acceleration meets the second preset condition, it means that the interactive obstacle can achieve the yielding behavior to the self-vehicle within a reasonable acceleration driving range, and combined with its yielding behavior intention probability, its behavior intention to yield to the self-vehicle is reasonable and has a high probability, so the speed decision and planning can be made based on the interactive prediction result of the interactive obstacle. In this case, the original obstacle ST risk field is insufficient in rationality and intelligence because it does not consider the interactive prediction result of the behavioral intention of the interactive obstacle. Based on this, when it is judged through operations S121 and S122 that the behavioral intention of the interactive obstacle to yield to the self-vehicle is feasible, the obstacle ST risk field will be updated based on the behavioral intention prediction result, so that the speed decision and planning are based on the updated obstacle ST risk field, that is, the interactive prediction result with the interactive obstacle is considered, thereby improving the rationality and intelligence of the speed decision and planning, making it more in line with the driver's behavioral habits and more in line with the actual scene processing method. Specifically, in operation S123, the obstacle ST risk field is updated based on the third acceleration, that is, when the obstacle ST risk field is updated, the acceleration of the interactive obstacle is determined based on the third acceleration. For example, the third acceleration can be used as the standard value, and the difference between the current acceleration of the interactive obstacle and the third acceleration can be used as the variance, and the acceleration of the interactive obstacle can be set according to the normal distribution. Since the original obstacle ST risk field before the update uses the acceleration of the interactive obstacle as the standard value, and the obstacle ST risk field updated in this way has a certain overlap area with the original obstacle ST risk field, which is close to the original distribution, the obstacle ST risk field is updated in this way. Under the premise of ensuring that the self-vehicle planning result is comfortable and feasible, the impact on the interactive obstacle can be minimized. As for other parts of the update of the obstacle ST risk field, the method of constructing and updating the obstacle ST risk field in accordance with the prior art can be used. For example, the method of constructing the dynamic obstacle ST risk field described in the invention patent application with application number 202210272795.6 can be used to reconstruct and update the obstacle ST risk field of the embodiment of the present invention. Alternatively, the method of updating the dynamic obstacle ST risk field described in the invention patent application with application number 202210272795.6 can be used to update the obstacle ST risk field of the embodiment of the present invention. The only difference is that, in the embodiment of the present invention, the acceleration of the interactive obstacle is calculated using the normal distribution method based on the third acceleration.

[0103] As another possible implementation, in the process of the ego vehicle leading the interaction relationship, it is also possible to determine that the behavior of the ego vehicle overtaking the interactive obstacle is not feasible, and try the possibility and feasibility of the ego vehicle giving way to the interactive obstacle. Figure 7 On the basis of, when it is judged that it is not feasible for the ego vehicle to overtake the interactive obstacle, the obstacle ST risk field is updated based on the judgment of the probability of the second behavior intention of the interactive obstacle, so as to make speed decisions and planning based on the updated obstacle ST risk field. Specifically, Figure 8 Another specific implementation method flow of updating the obstacle ST risk field according to the behavior intention information of the interactive obstacle is shown, such as Figure 8 As shown, it can be implemented as including:

[0104] Operation S121, determining a first behavior intention probability of an interactive obstacle, and when the first behavior intention probability satisfies a first preset condition, executing operation S122, and when the first behavior intention probability does not satisfy the first preset condition, executing operation S124;

[0105] Operation S122, determining a third acceleration required for the self-vehicle to yield to the interactive obstacle based on the first comfortable acceleration of the self-vehicle, and determining whether the third acceleration satisfies a second preset condition, if so, executing operation S123, if not, executing operation S124;

[0106] Operation S123, updating the obstacle ST risk field based on the third acceleration;

[0107] Operation S124, determining a second behavior intention probability of the interactive obstacle, and when the second behavior intention probability satisfies a third preset condition, executing operation S125;

[0108] Operation S125, determining a fourth acceleration required for the interactive obstacle to overtake the self-vehicle based on the second comfortable acceleration of the self-vehicle, and determining whether the fourth acceleration satisfies a fourth preset condition, and if so, executing operation S126;

[0109] Operation S126: updating the obstacle risk field based on the fourth acceleration.

[0110] exist Figure 8In the illustrated embodiment, when it is determined that the probability of the first behavior intention does not satisfy the first preset condition or the third acceleration does not satisfy the second preset condition, it indicates that the way of overtaking by the self-vehicle is not feasible. At this time, the possibility, feasibility and rationality of the way of giving way by the self-vehicle will continue to be judged through operations S124 to S126. Among them, operation S124 determines whether the probability of the second behavior intention satisfies the third preset condition to determine the possibility of the self-vehicle giving way to the interactive obstacle. Since it is possible for the self-vehicle to give way only when the interactive obstacle has the behavior intention to overtake the self-vehicle, the third preset condition can be set with reference to the first preset condition, such as setting it to the probability of the second behavior intention not lower than the minimum probability threshold, and the set minimum probability threshold can be, for example, 0.5. Therefore, only when the probability of the second behavior intention is not lower than the minimum probability threshold, that is, when the interactive obstacle is likely to overtake the self-vehicle, the feasibility and rationality of the self-vehicle giving way behavior will continue to be judged.

[0111] In operation S125, when it is determined through operation S124 that the vehicle may yield, that is, the interactive obstacle may overtake, the feasibility, rationality and safety of the vehicle yielding will be further judged to improve the intelligence of the speed decision and planning method. As a possible implementation method, the feasibility, rationality and safety of the vehicle yielding and the interactive obstacle overtaking can be judged based on the second comfortable acceleration that the vehicle can take. The second comfortable acceleration represents the comfortable acceleration required for comfortable deceleration, so it can be the negative value of the first comfortable acceleration. For example, it can be set to -1m / s 2 . Specifically, the feasibility and rationality of the interactive obstacle overtaking the own vehicle can be further judged when the own vehicle adopts the second comfortable acceleration to give way to the interactive obstacle. Therefore, in operation S125, the acceleration required for the interactive obstacle to overtake the own vehicle (referred to as the fourth acceleration in the embodiment of the present invention) can be determined based on the second comfortable acceleration of the own vehicle, and the feasibility of the interactive obstacle overtaking the own vehicle can be judged according to the rationality of the fourth acceleration, that is, the rationality of the prediction result of the interactive obstacle behavior intention and the feasibility of the current decision and planning based on the interactive prediction result are determined. Among them, the specific implementation method of determining the fourth acceleration required for the interactive obstacle to overtake the own vehicle based on the second comfortable acceleration of the own vehicle can be: setting the acceleration of the own vehicle to the second comfortable acceleration (such as -1m / s 2), reconstruct the ST graph of the ego vehicle in the Frenet coordinate system of the obstacle, and calculate the first minimum acceleration required for the interactive obstacle to overtake the ego vehicle based on the reconstructed ST graph, and use the recalculated first minimum acceleration as the fourth acceleration required for the interactive obstacle to overtake the ego vehicle. The ST graph of the ego vehicle in the Frenet coordinate system of the obstacle constructed here is the same as the ST graph described above, and its construction method is also the same, and the method for calculating the first minimum acceleration required for the interactive obstacle to give way to the ego vehicle based on the reconstructed ST graph is also the same as the above, as can be referred to Figure 4 The method is used for calculation, so it will not be repeated here. After the fourth acceleration required for the interactive obstacle to overtake the self-vehicle is calculated, it is also necessary to determine the rationality and feasibility of the fourth acceleration, which can be specifically achieved by setting the fourth preset condition. In operation S125, when the fourth acceleration meets the fourth preset condition, it means that it is reasonable, feasible and safe for the interactive obstacle to overtake the self-vehicle. Therefore, the fourth preset condition can be set according to the reasonable acceleration range that the interactive obstacle can reach, such as setting the fourth preset condition to not exceed the reasonable acceleration range of the interactive obstacle. For example, the fourth preset condition can be set to not exceed (-2m / s 2 ~1.5m / s 2), wherein the range is the acceleration range of the interactive obstacle under the normal driving state. When the fourth acceleration satisfies the fourth preset condition, it means that the interactive obstacle can achieve overtaking behavior of the self-vehicle within a reasonable acceleration driving range, and combined with its overtaking behavior intention probability, its behavior intention to overtake the self-vehicle is reasonable and has a high probability, so the speed decision and planning can be made based on the interactive prediction result of the interactive obstacle. In this case, the original obstacle ST risk field is insufficient in rationality and intelligence because it does not consider the interactive prediction result of the behavior intention of the conflicting obstacle. Based on this, when it is judged through operation S124 and operation S125 that the behavior intention of the interactive obstacle to overtake the self-vehicle is feasible, the obstacle ST risk field will be updated based on the behavior intention prediction result, so that the speed decision and planning are based on the updated obstacle ST risk field, that is, the interactive prediction result with the interactive obstacle is considered, thereby improving the rationality and intelligence of the speed decision and planning, making it more in line with the driver's behavior habits and more in line with the actual scene processing method. Specifically, in operation S126, the obstacle ST risk field is updated based on the fourth acceleration, that is, when the obstacle ST risk field is updated, the acceleration of the interactive obstacle is determined based on the fourth acceleration. Exemplarily, the fourth acceleration can be used as the standard value, and the difference between the current acceleration of the interactive obstacle and the fourth acceleration can be used as the variance. The acceleration of the interactive obstacle is set according to the normal distribution, thereby minimizing the impact on the interactive obstacle while ensuring that the self-vehicle planning results are comfortable and feasible. As for other parts of the update of the obstacle ST risk field, the method of constructing and updating the obstacle ST risk field in accordance with the prior art can be used. For example, the method of constructing the dynamic obstacle ST risk field described in the invention patent application with application number 202210272795.6 can be used to reconstruct and update the obstacle ST risk field of the embodiment of the present invention. Alternatively, the method of updating the dynamic obstacle ST risk field described in the invention patent application with application number 202210272795.6 can be used to update the obstacle ST risk field of the embodiment of the present invention. The only difference is that the acceleration of the interactive obstacle in the embodiment of the present invention is calculated using the normal distribution method based on the fourth acceleration.

[0112] As another possible implementation, when it is determined that the actions of overtaking and giving way to the interactive obstacle are not feasible, the possibility and feasibility of overtaking or giving way to the interactive obstacle can be achieved by adjusting the first comfort acceleration or the second comfort acceleration of the vehicle. Fig. 9 Another specific implementation method flow for updating the obstacle ST risk field according to the behavior intention information of the interactive obstacle is shown, such as Fig. 9 As shown, it can be implemented as including:

[0113] Operation S121, determining a first behavior intention probability of an interactive obstacle, and when the first behavior intention probability satisfies a first preset condition, executing operation S122, and when the first behavior intention probability does not satisfy the first preset condition, executing operation S124;

[0114] Operation S122, determining a third acceleration required for the self-vehicle to yield to the interactive obstacle based on the first comfortable acceleration of the self-vehicle, and determining whether the third acceleration satisfies a second preset condition, if so, executing operation S123, if not, executing operation S124;

[0115] Operation S123, updating the obstacle ST risk field based on the third acceleration;

[0116] Operation S124, determining a second behavior intention probability of the interactive obstacle, and when the second behavior intention probability satisfies a third preset condition, executing operation S125, and when the second behavior intention probability does not satisfy the third preset condition, executing operation S127;

[0117] Operation S125, determining a fourth acceleration required for the interactive obstacle to overtake the self-vehicle based on the second comfortable acceleration of the self-vehicle, and determining whether the fourth acceleration satisfies a fourth preset condition, if so, executing operation S126, if not, executing operation S127;

[0118] Operation S126, updating the obstacle risk field based on the fourth acceleration;

[0119] Operation S127, adjusting the first comfortable acceleration of the own vehicle, re-determining the third acceleration required for the own vehicle to yield to the interactive obstacle based on the adjusted first comfortable acceleration of the own vehicle, and determining whether the third acceleration satisfies the second preset condition, if so, executing operation S123, if not, continuing to execute operation S128;

[0120] Operation S128, adjusting the second comfortable acceleration of the own vehicle, determining the fourth acceleration required for the interactive obstacle to overtake the own vehicle based on the second comfortable acceleration of the own vehicle, and determining whether the fourth acceleration meets a fourth preset condition, and executing operation S126 if it meets, and continuing to execute operation S127 if it does not meet.

[0121] pass Fig. 9In the embodiment shown, when a feasible decision and planning result cannot be found based on the first comfort acceleration and the second comfort acceleration, the acceleration of the vehicle will be adjusted, so that when the feasible solution is not predicted to be good, the planning space of the vehicle can be increased through the active behavior intention attempt of the vehicle, and the success rate and rationality of the decision and planning results can be improved. Therefore, in operation S127 to operation S128, the first comfort acceleration and the second comfort acceleration will be adjusted in sequence to seek a feasible solution until a feasible solution is found, and the obstacle ST risk field is updated based on the feasible solution found. Taking the preferred behavior intention of overtaking the vehicle as an example, the first comfort acceleration can be adjusted first to find a feasible solution under the attempt of overtaking the vehicle, and if it fails, the second comfort acceleration can be adjusted to find a feasible solution under the solution of giving way to the vehicle. Specifically, in operation S127, the first comfortable acceleration may be adjusted by sequentially increasing the preset value, and in operation S128, the second comfortable acceleration may be adjusted by sequentially decreasing the preset value.

[0122] In other embodiments, in an example of adjusting the first comfort acceleration and the second comfort acceleration, in order to ensure the safety and intelligence of decision-making and planning, the upper limit of the first comfort acceleration can also be constrained, so that during the adjustment process, if the first comfort acceleration reaches the constraint condition, that is, reaches the safety upper limit, the self-vehicle overtaking solution is no longer attempted by adjusting the first comfort acceleration, and the self-vehicle yielding solution is continuously attempted by adjusting the second comfort acceleration. When no feasible solution is found under the self-vehicle overtaking solution, giving way to the self-vehicle is used as the preferred solution, and the obstacle ST risk field is updated based on this, and decisions and planning are made based on the updated obstacle ST risk field.

[0123] In other embodiments, the priority of the vehicle's behavioral intention can be determined based on the right of way before attempting to overtake. For example, when the interactive obstacle has a higher right of way, the vehicle will first try to update the obstacle ST risk field under the self-vehicle yielding scheme, and when it is not feasible for the vehicle to yield, it will try to update the obstacle ST risk field under the self-vehicle overtaking scheme; and when the vehicle has a higher right of way, it can update the obstacle ST risk field in a self-vehicle overtaking priority manner as described above, and when it is not feasible for the vehicle to overtake, it will try to give way to the self-vehicle. In some preferred embodiments, when determining interactive obstacles, only dynamic obstacles with lower right of way than the vehicle can be selected as interactive obstacles for the above processing. In this way, it is possible to ignore the restriction of right of way and directly adopt the interactive method described above with self-vehicle overtaking as the priority to update the obstacle ST risk field.

[0124] In some other implementations, in order to make speed decisions and planning more intelligent and more in line with the driver's driving habits, before the ego vehicle actively overtakes and gives way, it can also be based on the analysis of the behavioral intention information of the interactive obstacle. If it is determined that the ego vehicle and the interactive obstacle will not collide if they maintain the current motion state, the obstacle ST risk field will not be updated, but speed decisions and planning will be made directly based on the already constructed obstacle ST risk field. Based on this, Fig.10 The speed decision and planning method based on interaction intention prediction according to another embodiment of the present invention is schematically shown. Fig.10 As shown, its implementation includes:

[0125] Operation S21, obtaining an obstacle ST risk field, and determining an interactive obstacle according to the obstacle ST risk field;

[0126] Operation S22, analyzing the behavior intention of the interactive obstacle to determine the behavior intention information of the conflicting obstacle;

[0127] Operation S23, determining the behavior intention acceleration of the vehicle under different behavior intentions;

[0128] Operation S24 , performing speed decision and planning directly based on the obstacle ST risk field or based on the updated obstacle ST risk field according to the behavior intention information of the interactive obstacle and the behavior intention acceleration of the vehicle.

[0129] Among them, the specific implementation of operation S21 and operation S22 can refer to the above description, and will not be repeated here. Different from the implementation shown in the above, in this implementation, the behavior intention analysis of the self-vehicle is also performed to determine the behavior intention acceleration of the self-vehicle, and the possibility of conflict between the self-vehicle and the interactive obstacle in the existing motion state is determined according to the behavior intention information of the interactive obstacle and the behavior intention acceleration of the self-vehicle, that is, the degree of conflict between the two. If the degree of conflict between the two is weak, in this case, the obstacle ST risk field is not updated, and it is also feasible to directly make speed decisions and plans based on the acquired obstacle ST risk field. Therefore, the obstacle ST risk field will be updated only when the degree of conflict between the self-vehicle and the interactive obstacle is high. Specifically, in operation S23, the behavior intention of the self-vehicle can be set to include overtaking the interactive obstacle and giving way to the interactive obstacle. The analysis of the behavior intention of the self-vehicle can be specifically implemented as determining the second minimum acceleration required for the self-vehicle to safely overtake the interactive obstacle and the second maximum acceleration required for safely giving way to the interactive obstacle, that is, the determined behavior intention acceleration includes the second minimum acceleration and the second maximum acceleration. Among them, the method for determining the second minimum acceleration required for the ego-vehicle to safely overtake the interactive obstacle and the second maximum acceleration required for safely giving way to the interactive obstacle is similar to the method for determining the first minimum acceleration and the first maximum acceleration of the interactive obstacle in the previous article. The only difference is that when determining the second minimum acceleration and the second maximum acceleration of the ego-vehicle, the constructed ST diagram is the ST diagram of the obstacle in the Frenet coordinate system of the ego-vehicle, that is, the meaning of the value of the vertical axis, i.e., the S-axis, in the ST diagram is different. The S-axis of the ST diagram is constructed based on the direction of the planned trajectory of the ego-vehicle, while the vertical axis of the ST diagram of the obstacle in the previous article is constructed based on the method of predicting the obstacle trajectory.

[0130] As a possible implementation, in operation S24, the possibility of a conflict between the vehicle and the interactive obstacle, that is, the degree of conflict between the two, can be determined based on the behavioral intention acceleration of the vehicle. Exemplarily, the degree of conflict between the vehicle and the interactive obstacle can be determined by judging whether the behavioral intention acceleration of the vehicle is in a preset comfort range. Specifically, in operation S24, it can be first judged whether the second minimum acceleration or the second maximum acceleration meets the fifth preset condition. When the second minimum acceleration or the second maximum acceleration both meet the fifth preset condition, the speed decision and planning are directly based on the obstacle ST risk field. When the second minimum acceleration and the second maximum acceleration do not meet the fifth preset condition, the obstacle ST risk field is updated according to the behavioral intention information of the interactive obstacle, and the speed decision and planning are based on the updated obstacle ST risk field. Among them, the fifth preset condition can be set to be in a preset comfort range, and the comfort range can preferably be set based on the existing acceleration or driving state of the vehicle. Exemplarily, it can be set to (-0.6m / s2 ~0.6m / s 2 ), when it is judged that the second minimum acceleration or the second maximum acceleration is in the preset comfort range, it means that the ego vehicle can easily achieve safe overtaking of the interactive obstacle or safe yielding of the interactive obstacle in the current motion state, which means that the conflict between the two in the current motion state is relatively low. In this case, it is safe and feasible to make speed decisions and plans directly based on the obstacle ST risk field. Therefore, it is not necessary for the ego vehicle to predict the behavior intention of the interactive obstacle, nor is it necessary to update the obstacle ST risk field in combination with the behavior intention of the interactive obstacle. If the second minimum acceleration or the second maximum acceleration is not in the preset comfort range, it means that it is difficult for the ego vehicle to safely overtake the interactive obstacle or safely yield to the interactive obstacle, that is, the possibility of conflict between the two is high. At this time, it is necessary to update the obstacle ST risk field in combination with the analysis of the behavior intention of the interactive obstacle, so as to achieve more intelligent, reasonable and safe speed decisions and planning based on the interactive prediction results. Among them, it should be noted that the specific implementation method of speed decision and planning based on the obstacle ST risk field can be implemented with reference to the existing technology, such as referring to the technical solution content disclosed in the invention patent application with application number 202210272793.7, and the embodiments of the present invention are not limited to this.

[0131] Therefore, through the method of the embodiment of the present invention, it is possible to make decisions and plans based on the existing obstacle ST risk field when it is predicted that the planning result of the self-vehicle under the existing operating state is feasible. When it is predicted that the conflict degree of the existing planning results is high, the behavior intentions of the interactive obstacles are continuously analyzed, and the obstacle risk field is updated based on the behavior intentions of the interactive obstacles to increase the decision-making planning space of the self-vehicle, so as to lead the decision-making in a direction that is beneficial to the self-vehicle based on the interaction. This method effectively improves the comfort and success rate of speed decision-making and speed planning, and makes the decision results more intelligent and reasonable.

[0132] Fig.12 The speed decision and planning device based on interaction intention prediction according to one embodiment of the present invention is schematically shown. Fig.12 As shown, the device comprises:

[0133] An interactive obstacle screening module 100 is used to obtain an obstacle ST risk field and determine interactive obstacles according to the obstacle ST risk field;

[0134] The obstacle intention determination module 200 is used to analyze the behavior intention of the interactive obstacle and determine the behavior intention information of the interactive obstacle;

[0135] The speed decision and planning module 400 is used to make speed decisions and plans based on the updated obstacle ST risk field according to the behavior intention information of the interactive obstacle.

[0136] Fig.13 The speed decision and planning device based on interaction intention prediction according to another embodiment of the present invention is schematically shown. Fig.13 As shown, the device comprises:

[0137] An interactive obstacle screening module 100 is used to obtain an obstacle ST risk field and determine interactive obstacles according to the obstacle ST risk field;

[0138] The obstacle intention determination module 200 is used to analyze the behavior intention of the interactive obstacle and determine the behavior intention information of the interactive obstacle;

[0139] The ego vehicle intention determination module 300 is used to determine the behavior intention acceleration of the ego vehicle under different behavior intentions;

[0140] The speed decision and planning module 400 is used to make speed decisions and plans directly based on the obstacle ST risk field or based on the updated obstacle ST risk field according to the behavior intention information of the interactive obstacle and the behavior intention acceleration of the vehicle.

[0141] In an embodiment of the present invention, the obstacle ST risk field may refer to a pre-constructed obstacle ST risk field obtained from a front-end module, or may refer to an obstacle ST risk field updated by the method of an embodiment of the present invention. The front-end module may, for example, be a controller connected to a planner on an autonomous driving vehicle, or a sub-module or processor of the planner for constructing an obstacle ST risk field.

[0142] In the embodiment of the present invention, interactive obstacles refer to those that can observe the state of the ego vehicle and adjust their behavior intentions according to the ego vehicle's behavior, and are conflicting obstacles after removing those that are not affected by the ego vehicle and those with high road rights from the conflicting obstacles. The dynamic obstacles here refer to traffic participants other than the ego vehicle that may move, which may be pedestrians, animals, or other vehicles.

[0143] In some possible implementations, the behavior intention information may be implemented as a behavior intention probability including an interactive obstacle. In the embodiment of the present invention, the behavior intention probability is used to characterize the similarity between the current acceleration of the interactive obstacle and the expected acceleration for achieving the corresponding behavior intention. The higher the behavior intention probability, the closer its current acceleration is to the expected acceleration required to achieve the corresponding behavior intention, which means that it is more likely to achieve the corresponding behavior intention. In some preferred implementations, the behavior intention of the interactive obstacle can be set to include giving way to the own vehicle and overtaking the own vehicle. Taking this as an example, the behavior intention probability specifically includes a first behavior intention probability for characterizing the behavior intention of giving way to the own vehicle and a second behavior intention probability for characterizing the behavior intention of overtaking the own vehicle. The first behavior intention probability characterizes the similarity between the acceleration of the interactive obstacle at the current moment and the expected acceleration required to realize the behavior intention of giving way to the own vehicle, and reflects the possibility that the interactive obstacle can achieve the behavior intention of giving way to the own vehicle based on the acceleration at the current moment. The higher the value, the greater the possibility that the interactive obstacle will execute the behavior intention of giving way to the own vehicle. The second behavior intention probability characterizes the similarity between the acceleration of the interactive obstacle at the current moment and the expected acceleration required to realize the behavior intention of overtaking the own vehicle, and reflects the possibility that the interactive obstacle can achieve the behavior intention of overtaking the own vehicle based on the acceleration at the current moment. The higher the value, the greater the possibility that the interactive obstacle will execute the behavior intention of overtaking the own vehicle.

[0144] In a preferred embodiment, in order to achieve a more accurate prediction of the behavior intention of the interactive obstacle, the time-based confidence can be used to optimize and adjust the behavior intention probability to obtain a behavior intention probability that more accurately reflects the behavior intention of the interactive obstacle. Preferably, the confidence can be determined based on the observation time and the conflict time.

[0145] As a possible implementation, the behavior intention of the vehicle can be set to include overtaking the conflicting obstacle and giving way to the conflicting obstacle. The analysis of the behavior intention of the vehicle can be specifically implemented by determining the second minimum acceleration required for the vehicle to safely overtake the conflicting obstacle and the second maximum acceleration required for the vehicle to safely give way to the conflicting obstacle, that is, the determined behavior intention acceleration includes the second minimum acceleration and the second maximum acceleration.

[0146] Among them, it should be noted that the specific implementation process of each module of the speed decision-making and planning device based on interaction intention prediction in the embodiment of the present invention, for example, the implementation process of determining the interactive obstacle according to the obstacle ST risk field, the implementation process of analyzing the behavioral intention of the interactive obstacle, the implementation process of updating the obstacle ST risk field according to the behavioral intention information of the interactive obstacle, the implementation process of determining the behavioral intention acceleration of the vehicle, etc., can all refer to the description of the method part above, so they will not be repeated here.

[0147] Fig.14A speed decision and planning device based on interaction intention prediction according to another embodiment of the present invention is schematically shown. As shown in the figure, the device includes:

[0148] A memory 60, for storing executable instructions; and

[0149] The processor 61 is used to execute executable instructions stored in the memory, and when the executable instructions are executed by the processor, the steps of the speed decision and planning method based on interaction intention prediction described in any of the aforementioned embodiments are implemented.

[0150] In specific practice, the above-mentioned speed decision and planning device based on interaction intention prediction can be exemplarily applied to mobile tools such as autonomous driving vehicles, unmanned cleaners, unmanned sweepers, robots, etc., to realize speed planning of these mobile tools and improve the comfort of unmanned driving. Specifically, the above-mentioned speed decision and planning device based on interaction intention prediction can be implemented as a controller or planner on mobile tools such as autonomous driving vehicles.

[0151] Fig.15 A control device according to an embodiment of the present invention is schematically shown. Fig.15 As shown, the control device includes:

[0152] A planner 70, configured to perform speed decision and planning according to the method of any of the above embodiments; and

[0153] The controller 71 is used to control the vehicle according to the speed decision and planning result determined by the planner 70.

[0154] In other implementation examples, the planner and controller in the control device can also be integrated into one control module. Optionally, in practical applications, the control device can also include a perception and recognition module and other planning and control modules, such as a path planning controller, a bottom layer controller, etc., which is not limited in the embodiments of the present invention.

[0155] In other embodiments, the present invention Fig.14 The speed decision and planning device based on interaction intention prediction shown or Fig.15 The control device shown can also be applied to a mobile tool, such as an autonomous driving vehicle, so that the mobile tool can have the above functions described in the embodiment of the present invention.

[0156] In some embodiments, an embodiment of the present invention provides a non-volatile computer-readable storage medium, which stores one or more programs including execution instructions, and the execution instructions can be read and executed by an electronic device (including but not limited to a computer, a server, or a network device, etc.) to execute the speed decision-making and planning method based on interaction intention prediction of any of the above embodiments of the present invention.

[0157] In some embodiments, an embodiment of the present invention further provides a computer program product, which includes a computer program stored on a non-volatile computer-readable storage medium, and the computer program includes program instructions. When the program instructions are executed by a computer, the computer executes the speed decision-making and planning method based on interaction intention prediction of any one of the above embodiments.

[0158] In some embodiments, an embodiment of the present invention further provides an electronic device, comprising: at least one processor, and a memory communicatively connected to the at least one processor, wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the speed decision and planning method based on interaction intention prediction of any of the above embodiments.

[0159] In some embodiments, an embodiment of the present invention further provides a storage medium having a computer program stored thereon, which, when executed by a processor, implements the speed decision and planning method based on interaction intention prediction of any of the above embodiments.

[0160] Fig.16 is a schematic diagram of the hardware structure of a speed decision and planning device based on interaction intention prediction provided by another embodiment of the present application. The speed decision and planning device based on interaction intention prediction can be implemented with the structure shown in the figure, such as Fig.16 As shown, the speed decision and planning device based on interaction intention prediction includes:

[0161] One or more processors 610 and memory 620, Fig.16 A processor 610 is taken as an example.

[0162] The speed decision and planning device based on interaction intention prediction may further include: an input device 630 and an output device 640 .

[0163] The processor 610, the memory 620, the input device 630 and the output device 640 may be connected via a bus or other means. Fig.16 The example of connecting through bus is taken in the following.

[0164] The memory 620, as a non-volatile computer-readable storage medium, can be used to store non-volatile software programs, non-volatile computer executable programs and modules, such as program instructions / modules corresponding to the speed decision and planning method based on interaction intention prediction in the embodiment of the present application. The processor 610 executes various functional applications and data processing of the server by running the non-volatile software programs, instructions and modules stored in the memory 620, that is, the speed decision and planning method based on interaction intention prediction in the above method embodiment is implemented.

[0165] The memory 620 may include a program storage area and a data storage area, wherein the program storage area may store an operating system and applications required for at least one function; the data storage area may store data created based on the use of speed decisions and planning methods predicted based on interaction intentions, etc. In addition, the memory 620 may include a high-speed random access memory, and may also include a non-volatile memory, such as at least one disk storage device, a flash memory device, or other non-volatile solid-state storage device. In some embodiments, the memory 620 may optionally include a memory remotely arranged relative to the processor 610, and these remote memories may be connected to the electronic device via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0166] The input device 630 may receive input digital or character information and generate signals related to user settings and function control of the image processing device. The output device 640 may include a display device such as a display screen.

[0167] The one or more modules are stored in the memory 620, and when executed by the one or more processors 610, perform the speed decision and planning method based on interaction intention prediction in any of the above method embodiments.

[0168] The above-mentioned product can execute the method provided in the embodiment of the present application, and has the functional modules and beneficial effects corresponding to the execution method. For technical details not fully described in this embodiment, please refer to the method provided in the embodiment of the present application.

[0169] The electronic devices of the embodiments of the present application exist in various forms, including but not limited to:

[0170] (1) Mobile communication devices: These devices are characterized by their mobile communication functions and their main purpose is to provide voice and data communications. These terminals include: smart phones (such as iPhone), multimedia phones, functional phones, and low-end phones.

[0171] (2) Ultra-mobile personal computer devices: These devices fall into the category of personal computers, have computing and processing capabilities, and generally also have mobile Internet access features. These terminals include: PDA, MID and UMPC devices, such as iPad.

[0172] (3) Portable entertainment devices: These devices can display and play multimedia content. They include audio and video players (such as iPods), handheld game consoles, e-books, as well as smart toys and portable car navigation devices.

[0173] (4) Server: A device that provides computing services. The server consists of a processor, hard disk, memory, system bus, etc. The server is similar to the general computer architecture, but because it needs to provide highly reliable services, it has higher requirements in terms of processing power, stability, reliability, security, scalability, manageability, etc.

[0174] (5) Other electronic devices with data interaction functions.

[0175] The device embodiments described above are merely illustrative, wherein the units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0176] Through the description of the above implementation methods, those skilled in the art can clearly understand that each implementation method can be implemented by means of software plus a general hardware platform, and of course, by hardware. Based on this understanding, the above technical solution is essentially or the part that contributes to the relevant technology can be embodied in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a disk, an optical disk, etc., including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.

[0177] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit it. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A speed decision and planning method based on interaction intention prediction, characterized in that: include: Obtaining an obstacle ST risk field, and determining an interactive obstacle according to the obstacle ST risk field; Analyze the behavioral intention of the interactive obstacle and determine the behavioral intention information of the interactive obstacle; According to the behavior intention information of the interactive obstacle, the obstacle ST risk field is updated, and speed decision and planning are performed based on the updated obstacle ST risk field.

2. The method according to claim 1, characterized in that The behavior intention of the interactive obstacle includes yielding to the own vehicle and overtaking the own vehicle. The behavior intention information includes the behavior intention probability of the interactive obstacle at the current moment. The behavior intention of the interactive obstacle is analyzed to determine the behavior intention information of the interactive obstacle, including: Determine, based on the predicted trajectory of the interactive obstacle and the planned trajectory of the ego vehicle, a first minimum acceleration required for the interactive obstacle to safely overtake the ego vehicle and a first maximum acceleration required for safely giving way to the ego vehicle; The behavior intention probability of the interactive obstacle at the current moment is determined according to the current acceleration, the first minimum acceleration and the first maximum acceleration of the interactive obstacle, wherein the behavior intention probability includes a first behavior intention probability for characterizing the behavior intention of yielding to the own vehicle and a second behavior intention probability for characterizing the behavior intention of overtaking the own vehicle.

3. The method according to claim 2, characterized in that The analyzing the behavior intention of the interactive obstacle to determine the behavior intention information of the interactive obstacle also includes: Determine the observation time of interactive obstacles based on the ego-vehicle perception data; Determine the collision time with the interactive obstacle based on the predicted trajectory of the interactive obstacle and the planned trajectory of the ego vehicle; Determine the confidence level of behavioral intention based on observation time and conflict time; The behavior intention probability of the interactive obstacle at the current moment is updated according to the confidence level, and the updated behavior intention probability is used as the behavior intention probability of the interactive obstacle at the current moment.

4. The method according to claim 2 or 3, characterized in that: The step of determining, based on the predicted trajectory of the interactive obstacle and the planned trajectory of the ego vehicle, a first minimum acceleration required for the interactive obstacle to safely overtake the ego vehicle and a first maximum acceleration required for safely yielding to the ego vehicle comprises: Construct the ST graph of the ego vehicle in the obstacle Frenet coordinate system based on the predicted trajectory of the interactive obstacle and the planned trajectory of the ego vehicle; According to the upper boundary of the ST graph and the current speed of the interactive obstacle, the first minimum acceleration required for the interactive obstacle to safely overtake the ego vehicle is determined; The first maximum acceleration required for the vehicle to safely yield to the interactive obstacle is determined based on the lower boundary of the ST graph and the current speed of the interactive obstacle.

5. The method according to claim 2 or 3, wherein updating the obstacle ST risk field according to the behavior intention information of the interactive obstacle comprises: Determine the probability of the first behavior intention of the interactive obstacle. When the first behavior intention probability meets the first preset condition, determine the third acceleration required for the interactive obstacle to give way to the own vehicle based on the first comfortable acceleration of the own vehicle, and determine whether the third acceleration meets the second preset condition. If so, update the obstacle ST risk field based on the third acceleration.

6. The method according to claim 5, characterized in that The updating of the obstacle ST risk field according to the behavior intention information of the interactive obstacle also includes: When the first behavior intention probability does not meet the first preset condition or the third acceleration does not meet the second preset condition, the second behavior intention probability of the interactive obstacle is judged. When the second behavior intention probability meets the third preset condition, the fourth acceleration required for the interactive obstacle to overtake the own vehicle is determined based on the second comfortable acceleration of the own vehicle, and it is judged whether the fourth acceleration meets the fourth preset condition. If so, the obstacle risk field is updated based on the fourth acceleration.

7. The method according to claim 6, wherein updating the obstacle ST risk field according to the behavior intention information of the interactive obstacle further comprises: When the second behavior intention probability does not meet the third preset condition or the fourth acceleration does not meet the fourth preset condition, adjust the first comfort acceleration of the own vehicle, and redetermine the third acceleration required for the interactive obstacle to give way to the own vehicle based on the adjusted first comfort acceleration of the own vehicle, and judge whether the third acceleration meets the second preset condition. If so, update the obstacle ST risk field based on the third acceleration; if not, adjust the second comfort acceleration of the own vehicle, determine the fourth acceleration required for the interactive obstacle to overtake the own vehicle based on the second comfort acceleration of the own vehicle, and judge whether the fourth acceleration meets the fourth preset condition. If so, update the obstacle risk field based on the fourth acceleration.

8. A speed decision and planning method based on interaction intention prediction, characterized in that: include: Obtaining an obstacle ST risk field, and determining an interactive obstacle according to the obstacle ST risk field; Analyze the behavioral intention of the interactive obstacle and determine the behavioral intention information of the interactive obstacle; Determine the behavior intention acceleration of the ego vehicle under different behavior intentions; According to the behavior intention information of the interactive obstacle and the behavior intention acceleration of the ego vehicle, speed decision and planning are performed directly based on the obstacle ST risk field or based on the updated obstacle ST risk field.

9. The method according to claim 8, characterized in that The behavior intention of the ego vehicle includes overtaking the interactive obstacle and giving way to the interactive obstacle. The behavior intention acceleration includes the second minimum acceleration required for the ego vehicle to safely overtake the interactive obstacle and the second maximum acceleration required for safely giving way to the interactive obstacle. The behavior intention acceleration is determined based on the predicted trajectory of the interactive obstacle and the planned trajectory of the ego vehicle.

10. The method according to claim 9, characterized in that The speed decision and planning is performed directly based on the obstacle ST risk field or based on the updated obstacle ST risk field according to the behavior intention information of the interactive obstacle and the behavior intention acceleration of the vehicle, including: Determine whether the second minimum acceleration or the second maximum acceleration satisfies a fifth preset condition, and when both the second minimum acceleration or the second maximum acceleration satisfy the fifth preset condition, directly perform speed decision and planning based on the obstacle ST risk field; When neither the second minimum acceleration nor the second maximum acceleration satisfies the fifth preset condition, the obstacle ST risk field is updated according to the behavior intention information of the interactive obstacle, and speed decision and planning are performed based on the updated obstacle ST risk field.

11. A speed decision and planning device based on interaction intention prediction, characterized in that: The device comprises: An interactive obstacle screening module, used to obtain an obstacle ST risk field and determine interactive obstacles according to the obstacle ST risk field; The obstacle intention determination module is used to analyze the behavior intention of the interactive obstacle and determine the behavior intention information of the interactive obstacle; The ego vehicle intention determination module is used to determine the behavior intention acceleration of the ego vehicle under different behavior intentions; The speed decision and planning module is used to make speed decisions and plans directly based on the obstacle ST risk field or based on the updated obstacle ST risk field according to the behavior intention information of the interactive obstacle and the behavior intention acceleration of the vehicle.

12. A speed decision and planning device based on interaction intention prediction, characterized in that: include: A memory for storing executable instructions; as well as A processor, configured to execute executable instructions stored in a memory, wherein the executable instructions implement the steps of the method according to any one of claims 1 to 10 when executed by the processor.

13. A mobile tool, characterized in that: Comprising a speed planning device as claimed in claim 12.

14. A storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the steps of the method described in any one of claims 1 to 10 are implemented.

Images