A method, device, electronic device, medium and product for autonomous driving to escape from trouble

Abstract

The present application provides an autonomous driving escape method, device, electronic device, medium and product. According to an example of the present application, the method may include: obtaining the perception information of the vehicle during the autonomous driving process; inputting the perception information into an escape decision model to obtain an escape strategy output by the escape decision model, wherein the escape decision model is used to determine the type of trapped scene to which the perception information belongs, and selecting the highest priority escape strategy from the escape strategies of different priorities corresponding to the trapped scene type; executing the escape strategy to control the vehicle to escape from the trapped scene.

Description

Technical Field

[0001] The present application relates to the field of autonomous driving technology, and in particular to an autonomous driving escape method, device, electronic equipment, medium and product. Background Art

[0002] With the rapid development of autonomous driving technology, its application scenarios are becoming increasingly extensive, from closed test areas to open roads, and then to commercial operations in specific environments. Autonomous driving vehicles need to face a variety of complex and changeable road and traffic conditions. In these cases, autonomous driving vehicles may encounter various emergencies and difficulties, causing the vehicle to fall into driving difficulties. At this time, the vehicle's ability to escape from difficulties becomes particularly important.

[0003] For example, during the automatic driving process of the vehicle, if it encounters a plastic bag, it dares not move forward and keeps stopping and waiting. The vehicle does not have the ability to automatically escape from the predicament and cannot automatically escape from the predicament. Summary of the invention

[0004] In order to overcome the problems existing in the related technologies, the present application provides an autonomous driving escape method, device, electronic equipment, medium and product.

[0005] According to a first aspect of any embodiment of the present application, a method for escaping from an obstacle in an automatic driving system is provided, the method comprising:

[0006] Obtain the vehicle's perception information during autonomous driving;

[0007] Inputting the perception information into an escape decision model to obtain an escape strategy output by the escape decision model; wherein the escape decision model is used to determine the type of trapped scene to which the perception information belongs, and selecting the highest priority escape strategy from the escape strategies of different priorities corresponding to the trapped scene type;

[0008] Execute the escape strategy to control the vehicle to escape from the trapped scene.

[0009] According to a second aspect of any embodiment of the present application, there is provided an automatic driving escape device, the device comprising:

[0010] An information acquisition module is used to obtain the perception information of the vehicle during the autonomous driving process;

[0011] A strategy acquisition module, used to input the perception information into an escape decision model to obtain an escape strategy output by the escape decision model; wherein the escape decision model is used to determine the type of trapped scene to which the perception information belongs, and select the highest priority escape strategy from the escape strategies of different priorities corresponding to the trapped scene type;

[0012] The strategy execution module is used to execute the escape strategy and control the vehicle to escape from the trapped scene.

[0013] According to a third aspect of any embodiment of the present application, there is provided an electronic device, including:

[0014] processor;

[0015] a memory for storing processor-executable instructions;

[0016] The processor implements the method described in any embodiment of the present application by running the executable instructions.

[0017] According to a fourth aspect of any embodiment of the present application, a computer-readable storage medium is provided, on which computer instructions are stored. When the instructions are executed by a processor, the method described in any embodiment of the present application is implemented.

[0018] According to a fifth aspect of any embodiment of the present application, a computer program product is provided, on which a computer program / instruction is stored. When the computer program / instruction is executed by a processor, the method described in any embodiment of the present application is implemented.

[0019] The technical solution provided by this application may have the following beneficial effects:

[0020] According to the above embodiments, by acquiring the perception information of the vehicle during the automatic driving process, the perception information is input into the escape decision model to obtain the escape strategy output by the escape decision model, and according to the escape strategy, the vehicle is controlled to escape from the trapped scene, and the escape decision model is used to determine the type of trapped scene to which the perception information belongs, and the highest priority escape strategy is selected from the escape strategies of different priorities corresponding to the trapped scene type, so that the vehicle can identify and classify different trapped scenes, and adopt corresponding escape strategies for different scenes, thereby improving the flexibility and coping ability of the vehicle in various trapped environments.

[0021] It should be understood that the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the present application. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] The accompanying drawings, which are incorporated in the specification and constitute a part of this application, illustrate embodiments consistent with the application and, together with the description, serve to explain the principles of the application.

[0023] Figure 1 is a structural schematic diagram of an automatic driving system according to an exemplary embodiment of the present application;

[0024] Figure 2 is a flow chart of an automatic driving escape method according to an exemplary embodiment of the present application;

[0025] Figure 3 is a flow chart of another method for getting out of trouble in an automatic driving system according to an exemplary embodiment of the present application;

[0026] Figure 4 is a structural schematic diagram of an electronic device according to an exemplary embodiment of the present application;

[0027] Figure 5 It is a block diagram of an automatic driving escape device according to an exemplary embodiment of the present application. DETAILED DESCRIPTION

[0028] Exemplary embodiments will be described in detail herein, examples of which are shown in the accompanying drawings. When the following description refers to the drawings, the same numbers in different drawings represent the same or similar elements unless otherwise indicated. The implementations described in the following exemplary embodiments do not represent all implementations consistent with the present application. Instead, they are merely examples of devices and methods consistent with some aspects of the present application as detailed in the appended claims.

[0029] The terms used in this application are for the purpose of describing specific embodiments only and are not intended to limit this application. The singular forms of "a", "said" and "the" used in this application and the appended claims are also intended to include plural forms unless the context clearly indicates other meanings. It should also be understood that the term "and / or" used herein refers to and includes any or all possible combinations of one or more associated listed items.

[0030] It should be understood that although the terms first, second, third, etc. may be used in the present application to describe various information, these information should not be limited to these terms. These terms are only used to distinguish the same type of information from each other. For example, without departing from the scope of the present application, the first information may also be referred to as the second information, and similarly, the second information may also be referred to as the first information. Depending on the context, the word "if" as used herein may be interpreted as "at the time of" or "when" or "in response to determining".

[0031] At present, since the vehicle does not have the ability to automatically escape from trouble, it cannot automatically escape from trouble when encountering extreme weather conditions, unknown obstacles and other driving difficulties during the autonomous driving process.

[0032] In order to solve the above problems, the present application proposes an automatic driving escape method. To further illustrate the present application, the following embodiments are provided:

[0033] See also Figure 1 , Figure 11 is a schematic diagram of the structure of an automatic driving system according to an exemplary embodiment of the present application. The automatic driving escape method can be executed by the automatic driving system 10, and the automatic driving system 10 can be applied to a vehicle or a cloud service end. The automatic driving system 10 may include: a perception sensor 101, a controller 102, an interaction module 103 and an actuator 104.

[0034] The perception sensor 101 is used to convert the collected environmental information into digital signals to achieve all-round perception of the vehicle's surrounding environment. The controller 102 is used to make escape decisions and path planning for the vehicle.

[0035] The interaction module 103 is used to exchange information and interact with passengers and the external environment. The actuator 104 is used to receive control instructions from the controller 102 and convert the control instructions into actual physical actions of the vehicle.

[0036] See also Figure 2 , Figure 2 It is a flow chart of an autonomous driving escape method according to an exemplary embodiment of the present application.

[0037] by Figure 1 Taking the automatic driving system 10 shown in the figure as an example, the automatic driving escape method may include the following steps:

[0038] Step 202: Acquire the perception information of the vehicle during the autonomous driving process.

[0039] In this step, during the vehicle's autonomous driving process, multiple perception sensors 101 such as lidar, cameras and radars can work together to collect environmental information such as roads and traffic flows around the vehicle, convert the environmental information into digital signals, and send the digital signals to the controller 102.

[0040] The controller 102 fuses the digital signals sent by the multiple perception sensors 101 of the vehicle to obtain comprehensive and accurate perception information.

[0041] Among them, the perception information is the perception fusion information of the vehicle's surrounding environment, which is used to identify the environmental characteristics of the vehicle, such as vehicles, pedestrians, obstacles, traffic signs, etc.

[0042] Step 204: input the perception information into the escape decision model to obtain the escape strategy output by the escape decision model; wherein the escape decision model is used to determine the type of trapped scene to which the perception information belongs, and select the highest priority escape strategy from the escape strategies of different priorities corresponding to the trapped scene type.

[0043] In this step, the controller 102 inputs the perception information into the escape decision model in the controller 102 to obtain the escape strategy output by the escape decision model.

[0044] The escape decision model includes a trapped scene collector and an escape strategy collector. The trapped scene collector stores a plurality of trapped scene types collected in advance, and the escape strategy collector stores a plurality of escape strategies of different priorities corresponding to each trapped scene type collected in advance.

[0045] The escape decision model can determine whether the trapped scene type to which the perception information belongs exists in the trapped scene collector. If the trapped scene type to which the perception information belongs exists in the trapped scene collector, it means that the vehicle is in a trapped scene, the trapped scene type to which the perception information belongs is determined, and the highest priority escape strategy is selected from the escape strategies of different priorities corresponding to the trapped scene type.

[0046] If the trapped scene type to which the perception information belongs does not exist in the trapped scene collector, it means that the vehicle is not in a trapped scene, and the vehicle is controlled to travel normally according to the planned path.

[0047] The trapped scene type is the type of trapped scene the vehicle is in, and is used to determine the escape strategy corresponding to the perception information. The trapped scene type may be a dead-end road or a pothole road that results in a roadless passage, or an unknown obstacle in front of the vehicle.

[0048] The escape strategy is a solution to the problem of a vehicle being trapped, such as the identification logic of static objects and movable objects, the countermeasure processing of self-switching paths, the strategy logic of automatic escape, the path of courtesy, etc.

[0049] The priority of the escape strategy is the execution order of multiple escape strategies corresponding to the trapped scenario type, which is determined based on factors such as the urgency of the escape strategy, the success rate, and the impact on the safety of the vehicle or occupants.

[0050] The escape decision model is used to assist decision-making in vehicle escape scenarios. Based on the pre-collected trapped scenario types and their corresponding escape strategies, the escape plan is quickly determined by real-time analysis and processing of perception information. The escape decision model can be a neural network model, a deep learning model, a decision tree, etc.

[0051] The trapped scene type, escape strategy, escape strategy priority, and the corresponding relationship between the trapped scene type and the escape strategy can be set and adjusted according to the actual autonomous driving scene, and the escape decision model can be reloaded. The reloaded escape decision model is used to analyze and make decisions on the perceived information.

[0052] In one embodiment, the trapped scene type includes at least one of the following: no road to pass, obstacles blocking passage, traffic signs blocking passage, and road conditions blocking passage.

[0053] Among them, no-road passage means that there is no road ahead for the vehicle to pass. No-road passage may include: there are oncoming vehicles, but both sides have no path to pass; dead-end roads or bumpy roads result in no-road passage and crossing the stop line at a city intersection but there is no road ahead.

[0054] Obstacle blocking the passage is an unknown object on the road that blocks the normal passage of vehicles. Obstacle blocking the passage may include: there is a foreign object on the road blocking the passage, but there is no car ahead and there is a foreign object on the road blocking the passage, but there is a car ahead, etc.

[0055] Traffic signs blocking traffic means setting clear traffic signs or markings on the road to indicate that vehicles are not allowed to continue moving forward. Traffic signs blocking traffic may include: reverse arrows on the road and no-travel signs on the road.

[0056] Road conditions that prevent passage are those that make it impossible for vehicles or pedestrians to pass safely due to abnormal conditions of the road itself. Road conditions that prevent passage may include: the slope value is too large and exceeds the normal slope, and the bumpy road features are clearly not conducive to passage, etc.

[0057] As described above, by setting trapped scenario types such as no road to pass, obstacles blocking passage, traffic signs blocking passage, and road conditions blocking passage, it is possible to more accurately identify more types of complex scenarios that the vehicle is currently facing, thereby formulating more customized and targeted escape strategies.

[0058] Step 206: Execute the escape strategy to control the vehicle to escape from the trapped scene.

[0059] In this step, the controller 102 sends the escape strategy output by the escape decision model to the actuator 104, and the actuator 104 executes the escape strategy to control the vehicle to escape from the trapped scene.

[0060] Among them, the trapped scenario is a situation where the vehicle cannot continue to move forward due to various factors during the autonomous driving process.

[0061] The automatic driving escape method of the present embodiment obtains perception information of the vehicle during the automatic driving process, inputs the perception information into an escape decision model, obtains an escape strategy output by the escape decision model, controls the vehicle to escape from the trapped scene according to the escape strategy, uses the escape decision model to determine the type of trapped scene to which the perception information belongs, and selects the highest priority escape strategy from the escape strategies of different priorities corresponding to the trapped scene type, so that the vehicle can identify and classify different trapped scenes, and adopt corresponding escape strategies for different scenes, thereby improving the flexibility and coping ability of the vehicle in various trapped environments.

[0062] In addition, adopting a priority sorting mechanism for escape strategies can help optimize the resource allocation of escape processing, while increasing the success rate of escape, reducing the time of trying ineffective methods, and improving the efficiency of escape.

[0063] In the above-mentioned embodiment, it is introduced that the optimal escape strategy corresponding to the trapped scene type to which the perception information belongs is selected by using the escape decision model, and the trapped scene is escaped by executing the escape strategy. In the following embodiment, a detailed description will be given for checking whether the vehicle has escaped successfully, and it can be applied to any of the above embodiments.

[0064] In one embodiment, the plurality of perception sensors 101 may collect environmental information around the vehicle again in a timed or periodic manner, convert the environmental information into digital signals, and send the digital signals to the controller 102 .

[0065] The controller 102 again fuses the received multiple digital signals to obtain the next perception information of the vehicle during the automatic driving process.

[0066] The controller 102 inputs the next perception information into the escape decision model. The escape decision model determines whether the trapped scene type to which the perception information belongs and the trapped scene type to which the next perception information belongs have changed. If the trapped scene type to which the perception information belongs and the trapped scene type to which the next perception information belongs are inconsistent, it indicates that the vehicle has been successfully escaped, and the trapped features of the vehicle's positioning point are recorded.

[0067] If the trapped scene type to which the next perception information belongs does not exist in the trapped scene collector, it can also be determined that the trapped scene type to which the perception information belongs is inconsistent with the trapped scene type to which the next perception information belongs, indicating that the vehicle has successfully escaped.

[0068] During the subsequent autonomous driving process of the vehicle, the path planning will avoid detouring around this positioning point to avoid being trapped again.

[0069] The next perception information is the perception information acquired next time and adjacent to the perception information, and is used to reflect whether the vehicle has escaped from the trapped scene.

[0070] Trapped features are attribute information of trapped scenes, which are used to avoid positioning points during path planning. For example: road feature attributes, traffic restriction attributes, traffic flow attributes, obstacle occurrence frequency, and road abnormality characteristics.

[0071] After recording the trapped features of the positioning point, if it is detected that there is a situation that violates the trapped features at the positioning point, the trapped features of the positioning point are deleted.

[0072] For example, if the characteristic attribute of the road where the vehicle's positioning point is recorded as a dead-end road, the road where the positioning point is located will be avoided in subsequent path planning to avoid being trapped in a dead-end road again.

[0073] If it is detected that other vehicles can pass through the location point normally, it means that the road may support normal vehicle passage due to reasons such as completion of road construction. The road characteristic attribute of the location point as a dead-end road can be deleted, thereby generating a more reasonable and faster driving route based on timely updated road information, avoiding unnecessary detours and delays, and improving overall travel efficiency.

[0074] As described above, by acquiring the next perception information of the vehicle during the autonomous driving process again, when the trapped scene type to which the perception information belongs is inconsistent with the trapped scene type to which the next perception information belongs, the trapped features of the vehicle's positioning point are recorded to ensure that the positioning point is avoided during subsequent path planning, thereby preventing the vehicle from falling into the trapped scene of the positioning point again during driving, and realizing self-optimization of the vehicle's path planning during the autonomous driving process.

[0075] In one embodiment, when the trapped scene type to which the perception information belongs is consistent with the trapped scene type to which the next perception information belongs, it indicates that the highest priority escape strategy has not successfully enabled the vehicle to escape from the trapped scene.

[0076] The escape decision model determines that the type of trapped scene to which the next perception information belongs has not changed, and the escape decision model outputs the next priority escape strategy corresponding to the trapped scene type. The controller 102 sends the next priority escape strategy to the actuator 104.

[0077] The actuator 104 executes the next priority escape strategy to control the vehicle to escape from the trapped scene. The controller 102 obtains and determines whether the trapped scene type to which the perception information belongs has changed after executing the next priority escape strategy. If the trapped scene type has not changed, the next priority escape strategy is selected and executed until the trapped scene type changes and the vehicle successfully escapes from the trapped scene.

[0078] The next priority is the priority of the next escape strategy to be executed after the previous escape strategy is executed, determined in the order of priority. The next priority escape strategy is the escape strategy with the highest priority among the remaining escape strategies after excluding the escape strategies that have been executed.

[0079] As described above, when the type of trapped scene to which the perception information belongs is consistent with the type of trapped scene to which the next perception information belongs, the next perception information is input into the escape decision model, the next priority escape strategy output by the escape decision model is obtained, the next priority escape strategy is executed, the vehicle is controlled to escape from the trapped scene, and escape strategies of different priorities are tried step by step, which can avoid wasting time and resources on inefficient or ineffective escape methods, reduce the time the vehicle is trapped, and improve escape efficiency.

[0080] In one embodiment, the escape strategies of different priorities corresponding to each trapped scene type may include: an ultimate strategy of the lowest priority and escape strategies of other priorities.

[0081] When the escape strategies with other priorities cannot escape the trapped scene, the controller 102 uses the ultimate strategy with the lowest priority, and the interactive module 103 can report the perception information to the background service. The personnel or systems with professional knowledge and experience in the background service can formulate customized escape instructions based on more extensive geographic information, traffic conditions and real-time data.

[0082] The interaction module 103 can receive the escape instruction provided by the background service, and send the escape instruction to the controller 102. The controller 102 parses the specific escape plan or action requirements in the escape instruction, and the actuator 104 controls the vehicle to escape from the trapped scene.

[0083] Based on the reported perception information and the escape instructions fed back by the background service, the types of trapped scenarios and corresponding escape strategies can be added to the escape decision model, and the escape decision model can be continuously optimized and iterated, thereby improving the vehicle's automatic escape capability and adapting to the ever-changing environment and complex trapped scenarios.

[0084] Among them, the ultimate strategy is to seek help from the background service remotely when other high-priority escape strategies corresponding to the trapped scene type are invalid. The escape instruction is an escape plan provided by the background service remote assistance.

[0085] The escape strategies of other priorities are escape strategies with a higher priority than the ultimate strategy among the escape strategies of different priorities corresponding to the trapped scene types.

[0086] As described above, when other priority escape strategies are unable to escape from the trapped scene, the ultimate strategy with the lowest priority is used to report the perception information to the background service, and the vehicle is controlled to escape from the trapped scene based on the escape instructions received from the background service. By introducing remote assistance from the background service, it is ensured that even in the face of extreme or complex trapped scenes, the escape instructions provided by the background service can be used to reduce the risk of accidents and improve the success rate and reliability of autonomous driving.

[0087] In one embodiment, after obtaining the escape strategy output by the escape decision model, the interaction module 103 can prompt the escape strategy to the assisting party in the trapped scene based on preset interaction methods such as voice, speaker, touch screen, gesture, etc. Based on the assisting party's indication information on the escape strategy, the vehicle is controlled to escape from the trapped scene.

[0088] Among them, assisting parties include at least one of the following: passengers in the vehicle and other vehicles trapped in the scene.

[0089] The instruction information is specific guidance or feedback provided by the assisting party to the autonomous driving system on how to execute the escape strategy, such as confirmation information on the escape strategy, suggestions for adjusting the escape strategy, etc.

[0090] Exemplarily, the interaction module 103 can perform voice interaction with other vehicles through external voice, establishing voice communication between the vehicle and other vehicles in the trapped scene, etc. The interaction module 103 can also interact with passengers in the vehicle through speakers, touch screens, etc.

[0091] As described above, by using a preset interactive method, the escape strategy is prompted to at least one assisting party among other vehicles and passengers in the trapped scene, and based on the assisting party's indication of the escape strategy, the vehicle is controlled to escape from the trapped scene, so that other vehicles and passengers in the vehicle can directly participate in the escape process, such as opening up channels or indicating safe paths, etc., thereby improving the efficiency and success rate of automatic escape.

[0092] To further illustrate the process of autonomous driving escape, Figure 3 A flow chart of another method for getting out of trouble by automatic driving is shown, and the method for getting out of trouble by automatic driving may include the following steps:

[0093] Step 302: Acquire the perception information of the vehicle during the autonomous driving process.

[0094] In this step, the controller 102 fuses the digital signals sent by the multiple perception sensors 101 of the vehicle to obtain the perception information of the vehicle's surrounding environment during the automatic driving process.

[0095] Exemplarily, the perception information acquired by the controller 102 includes: a dead-end road with no way to pass and a no-passing sign.

[0096] Step 304: Determine whether there is a trapped scene type to which the perception information belongs.

[0097] In this step, the controller 102 inputs the perception information into the escape decision model, and the escape decision model determines whether there is a trapped scene type to which the perception information belongs in the trapped scene collector.

[0098] Exemplarily, the controller 102 inputs the perception information of the dead-end road with no way through and the no-passing sign into the escape decision model, and the escape decision model determines whether there is a trapped scene type to which the dead-end road with no way through and the no-passing sign belongs in the trapped scene collector.

[0099] If there is no trapped scene type to which the perception information belongs, continue to execute step 306;

[0100] If there is a trapped scene type to which the perception information belongs, step 308 is executed.

[0101] Step 306: Control the vehicle to travel normally according to the planned path.

[0102] In this step, when there is no trapped scene type to which the perception information belongs, it means that the vehicle is not in a trapped scene, and the automatic driving system 10 controls the vehicle to drive normally according to the planned path.

[0103] Step 308: Selecting the highest priority escape strategy from the escape strategies of different priorities corresponding to the trapped scene type.

[0104] In this step, when there is a trapped scene type to which the perception information belongs, it means that the vehicle is in a trapped scene. The escape decision model determines the trapped scene type to which the perception information belongs, and selects the highest priority escape strategy from the escape strategies of different priorities corresponding to the trapped scene type.

[0105] Exemplarily, the escape decision model determines that the trapped scene types belonging to the dead-end road with no way to pass and the no-passing sign are no way to pass and traffic signs blocking passage, and selects the highest priority escape strategy from the escape strategies of different priorities corresponding to the trapped scene types. The highest priority escape strategies corresponding to no way to pass and traffic signs blocking passage are to control the vehicle to turn around at the original position.

[0106] Step 310: Execute the escape strategy to control the vehicle to escape from the trapped scene.

[0107] In this step, the actuator 104 executes the escape strategy output by the escape decision model to control the vehicle to escape from the trapped scene.

[0108] Exemplarily, the actuator 104 controls the vehicle to turn around at the original position and try to escape from the trapped scene.

[0109] Step 312: Obtain the next perception information of the vehicle during the autonomous driving process again.

[0110] In this step, the controller 102 again obtains the next perception information of the vehicle during the automatic driving process.

[0111] Step 314: Determine whether the type of trapped scene to which the next perception information belongs has changed.

[0112] In this step, the controller 102 inputs the next perception information into the escape decision model to determine whether the type of trapped scene to which the next perception information belongs has changed.

[0113] If the type of trapped scene to which the next perception information belongs changes, continue to execute step 316;

[0114] If the type of trapped scene to which the next perception information belongs has not changed, step 318 is executed.

[0115] Step 316: Record the trapped features of the vehicle's positioning point.

[0116] In this step, when the trapped scene type to which the perception information belongs is inconsistent with the trapped scene type to which the next perception information belongs, it is determined that the trapped scene type to which the next perception information belongs has changed, indicating that the executed escape strategy has enabled the vehicle to successfully escape from the trapped scene, and the controller 102 records the trapped characteristics of the vehicle's positioning point.

[0117] For example, if the U-turn position is sufficient and the vehicle turns successfully, the trapped scene type to which the next perception information belongs changes, indicating that the vehicle has escaped from the trapped scene, and the trapped feature of the vehicle's positioning point is recorded as a road feature attribute of a dead-end road.

[0118] Step 318: Obtain the next priority escape strategy output by the escape decision model.

[0119] In this step, when the trapped scene type to which the perception information belongs is consistent with the trapped scene type to which the next perception information belongs, it is determined that the trapped scene type to which the next perception information belongs has not changed, indicating that the executed escape strategy has not successfully enabled the vehicle to escape from the trapped scene. The controller 102 obtains the next priority escape strategy output by the escape decision model.

[0120] Exemplarily, if the U-turn position is insufficient and the vehicle fails to turn around successfully, the controller 102 obtains the next priority escape strategy output by the escape decision model, which is to slowly reverse along the original route until there is an intersection where the vehicle can change lanes.

[0121] Step 320: Execute the next priority escape strategy to control the vehicle to escape from the trapped scene.

[0122] In this step, the actuator 104 executes the next priority escape strategy to control the vehicle to escape from the trapped scene. Steps 312 and 314 are executed again to determine whether the trapped scene type to which the sensed information belongs changes after the next priority escape strategy is executed, and to determine whether the vehicle is successfully escaped.

[0123] For example, the actuator 104 controls the vehicle to reverse slowly along the original route until there is an intersection where the vehicle can change lanes, in an attempt to escape from the trapped scene. After the lane change is successful, the trapped scene type to which the perception information belongs changes, and the trapped feature of the vehicle's positioning point is recorded as a road feature attribute of a dead-end road.

[0124] Step 322: When other escape strategies with different priorities cannot escape from the trapped scene, the perception information is reported to the background service.

[0125] In this step, the escape strategy is selected and executed step by step according to the priority level, and the vehicle is controlled to try to escape from the trapped scene. In the case that the escape strategies of other priorities cannot escape from the trapped scene, the ultimate strategy with the lowest priority is used, and the interaction module 103 reports the perception information to the background service.

[0126] Step 324: Control the vehicle to escape from the trapped scene according to the escape instruction provided by the background service.

[0127] In this step, the interaction module 103 sends the escape instruction received from the background service to the controller 102, the controller 102 parses the escape instruction, and the actuator 104 controls the vehicle to escape from the trapped scene.

[0128] Figure 4 is a schematic diagram of the structure of an electronic device according to an exemplary embodiment of the present application. Figure 4 At the hardware level, the electronic device includes a processor 402, an internal bus 404, a network interface 406, a memory 408, and a non-volatile memory 410, and may also include hardware required for other services. The processor 402 reads the corresponding computer program from the non-volatile memory 410 into the memory 408 and then runs it, forming an automatic driving escape device at the logical level. Of course, in addition to software implementations, this application does not exclude other implementations, such as logic devices or a combination of software and hardware, etc., that is to say, the execution subject of the following processing flow is not limited to each logic unit, but can also be hardware or logic devices.

[0129] Figure 5 is a block diagram of an automatic driving escape device according to an exemplary embodiment of the present application. Figure 5 , the device may include: an information acquisition module 502, a policy acquisition module 504 and a policy execution module 506, wherein:

[0130] The information acquisition module 502 is used to acquire the perception information of the vehicle during the automatic driving process;

[0131] The strategy acquisition module 504 is used to input the perception information into the escape decision model to obtain the escape strategy output by the escape decision model; wherein the escape decision model is used to determine the type of trapped scene to which the perception information belongs, and select the highest priority escape strategy from the escape strategies of different priorities corresponding to the trapped scene type;

[0132] The strategy execution module 506 is used to execute the escape strategy and control the vehicle to escape from the trapped scene.

[0133] In one example, the information acquisition module 502 is further used to reacquire the next perception information of the vehicle during the automatic driving process; wherein the next perception information is used to reflect whether the vehicle has escaped from the trapped scene; the strategy execution module 506 is further used to record the trapped features of the vehicle's positioning point when the trapped scene type to which the perception information belongs is inconsistent with the trapped scene type to which the next perception information belongs; the trapped features are used to avoid the positioning point during path planning.

[0134] In one example, the strategy acquisition module 504 is further used to input the next perception information into the escape decision model to obtain the next priority escape strategy output by the escape decision model when the type of trapped scene to which the perception information belongs is consistent with the type of trapped scene to which the next perception information belongs; the strategy execution module 506 is further used to execute the next priority escape strategy to control the vehicle to escape from the trapped scene.

[0135] In one example, the escape strategies of different priorities include: an ultimate strategy with the lowest priority and escape strategies of other priorities; the strategy execution module 506, when used to execute the escape strategies and control the vehicle to escape from the trapped scene, includes: when the escape strategies of other priorities cannot escape from the trapped scene, using the ultimate strategy with the lowest priority to report the perception information to the background service; and controlling the vehicle to escape from the trapped scene according to the escape instructions provided by the background service.

[0136] In one example, the strategy execution module 506, when used to execute the escape strategy and control the vehicle to escape from the trapped scene, includes: based on a preset interaction method, prompting the escape strategy to an assisting party in the trapped scene; wherein the assisting party includes at least one of the following: passengers in the vehicle, other vehicles in the trapped scene; based on the assisting party's indication information on the escape strategy, controlling the vehicle to escape from the trapped scene.

[0137] In one example, the trapped scenario type includes at least one of the following: no road to pass, obstacles blocking passage, traffic signs blocking passage, and road conditions blocking passage.

[0138] The implementation process of the functions and effects of each unit in the above-mentioned device is specifically described in the implementation process of the corresponding steps in the above-mentioned method, and will not be repeated here.

[0139] For the device embodiment, since it basically corresponds to the method embodiment, the relevant parts can refer to the partial description of the method embodiment. The device embodiment described above is only schematic, wherein the modules described as separate components may or may not be physically separated, and the components displayed as modules may or may not be physical modules, that is, they may be located in one place, or they may be distributed on multiple network modules. Some or all of the modules may be selected according to actual needs to achieve the purpose of the present application scheme. A person of ordinary skill in the art can understand and implement it without paying creative labor.

[0140] In an exemplary embodiment, a non-temporary computer-readable storage medium including instructions is also provided, such as a memory including instructions. The above instructions can be executed by a processor of an autonomous driving escape device to implement a method as described in any of the above embodiments.

[0141] The non-temporary computer-readable storage medium may be a ROM, a random access memory (RAM), a CD-ROM, a magnetic tape, a floppy disk, an optical data storage device, etc., and the present application does not limit this.

[0142] In an exemplary embodiment, a computer program product including a computer program / instruction is also provided. The computer program / instruction can be executed by a processor of an autonomous driving escape device to implement a method as described in any of the above embodiments.

[0143] The above describes specific embodiments of the present application. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recorded in the claims can be performed in an order different from that in the embodiments and still achieve the desired results. In addition, the processes depicted in the accompanying drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0144] Those skilled in the art will readily appreciate other embodiments of the present application after considering the specification and practicing the inventions claimed herein. The present application is not limited to the precise structures described above and shown in the accompanying drawings, and various modifications and changes may be made without departing from the scope thereof. The scope of the present application is limited only by the appended claims.

[0145] The above description is only a preferred embodiment of the present application and is not intended to limit the present application. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present application shall be included in the scope of protection of the present application.

Claims

1. A method for escaping from an obstacle in an automatic driving system, characterized in that: The method comprises: Obtain the vehicle's perception information during autonomous driving; Inputting the perception information into an escape decision model to obtain an escape strategy output by the escape decision model; wherein the escape decision model is used to determine the type of trapped scene to which the perception information belongs, and selecting the highest priority escape strategy from the escape strategies of different priorities corresponding to the trapped scene type; Executing the escape strategy to control the vehicle to escape from the trapped scene; The method further comprises: When the trapped scene type to which the perception information belongs is consistent with the trapped scene type to which the next perception information belongs, inputting the next perception information into the escape decision model to obtain the next priority escape strategy output by the escape decision model; Execute the next priority escape strategy to control the vehicle to escape from the trapped scene.

2. The method according to claim 1, characterized in that The method further comprises: Acquiring again the next perception information of the vehicle during the automatic driving process; wherein the next perception information is used to reflect whether the vehicle has escaped from the trapped scene; When the trapped scene type to which the perception information belongs is inconsistent with the trapped scene type to which the next perception information belongs, the trapped features of the positioning point of the vehicle are recorded; the trapped features are used to avoid the positioning point during path planning.

3. The method according to claim 1, characterized in that The escape strategies of different priorities include: the ultimate strategy of the lowest priority and escape strategies of other priorities; The executing the escape strategy to control the vehicle to escape from the trapped scene includes: When the escape strategies of other priorities cannot escape the trapped scene, using the ultimate strategy of the lowest priority to report the perception information to the background service; According to the escape instruction provided by the background service, the vehicle is controlled to escape from the trapped scene.

4. The method according to claim 1, characterized in that: The executing the escape strategy to control the vehicle to escape from the trapped scene includes: Based on a preset interaction method, the escape strategy is prompted to an assisting party in the trapped scene; wherein the assisting party includes at least one of the following: a passenger in the vehicle, and other vehicles in the trapped scene; Based on the instruction information of the assisting party on the escape strategy, the vehicle is controlled to escape from the trapped scene.

5. The method according to claim 1, characterized in that The trapped scenario type includes at least one of the following: no road to pass, obstacles blocking passage, traffic signs blocking passage, and road conditions blocking passage.

6. An automatic driving escape device, characterized in that: The device comprises: An information acquisition module is used to obtain the perception information of the vehicle during the autonomous driving process; A strategy acquisition module, used to input the perception information into an escape decision model to obtain an escape strategy output by the escape decision model; wherein the escape decision model is used to determine the type of trapped scene to which the perception information belongs, and select the highest priority escape strategy from the escape strategies of different priorities corresponding to the trapped scene type; A strategy execution module, used to execute the escape strategy and control the vehicle to escape from the trapped scene; The strategy acquisition module is further used to input the next perception information into the escape decision model to obtain the next priority escape strategy output by the escape decision model when the type of trapped scene to which the perception information belongs is consistent with the type of trapped scene to which the next perception information belongs; The strategy execution module is also used to execute the escape strategy of the next priority level to control the vehicle to escape from the trapped scene.

7. An electronic device, characterized in that: include: processor; a memory for storing processor-executable instructions; The processor implements the method according to any one of claims 1 to 5 by running the executable instructions.

8. A computer-readable storage medium having computer instructions stored thereon, characterized in that: When the instruction is executed by a processor, the method according to any one of claims 1 to 5 is implemented.

9. A computer program product having a computer program / instructions stored thereon, characterized in that: When the computer program / instructions are executed by a processor, the method according to any one of claims 1 to 5 is implemented.

Images