Fault processing method of vehicle automatic driving system, electronic equipment and storage medium

Abstract

The invention relates to a fault processing method for an automatic driving system of a vehicle, electronic equipment and a storage medium, and the method comprises the steps: obtaining the fault duration of the automatic driving system in the vehicle and a fault set at each moment in the fault duration, and carrying out the fault processing of the automatic driving system according to the vehicle state of the vehicle for the fault set at each moment; determining a predicted driving distance after the vehicle executes each behavior combination, and determining a reference driving behavior corresponding to the fault set and a reference recovery behavior of each fault according to the predicted driving distance of each behavior combination, therefore, the reference driving behavior corresponding to each fault set in the fault duration and the reference recovery behavior of each fault are obtained, the fault processing of the automatic driving system of the vehicle is realized, the fault processing of the automatic driving system is realized, and the fault recovery efficiency of the automatic driving system is improved on the premise of ensuring the safety of the vehicle. Decisions between driving processing and recovery processing of the vehicle are optimized, and the overall performance of fault processing of the vehicle is improved.

Description

Technical Field

[0001] The present disclosure relates to the field of autonomous driving technology, and in particular to a fault handling method, electronic device, and storage medium for an autonomous driving system of a vehicle. Background Art

[0002] The autonomous driving system usually consists of multiple modules such as the control module, perception module, positioning module, chassis module, etc. working together to realize the vehicle's autonomous driving function. During the process of vehicle autonomous driving, various modules of the autonomous driving system may fail.

[0003] The existing fault handling method is usually to control the vehicle to pull over and prompt the driver to make repairs in time through the dashboard. However, some faults can actually recover by themselves. If the vehicle is directly controlled to pull over, it may affect the operating efficiency of the autonomous vehicle, and directly controlling the vehicle to stop will also affect traffic safety. Summary of the invention

[0004] In order to solve the above technical problems or at least partially solve the above technical problems, the embodiments of the present disclosure provide a fault handling method, an electronic device and a storage medium for a vehicle automatic driving system to solve the problems of low operating efficiency and impact on traffic safety in the prior art fault handling.

[0005] In a first aspect, an embodiment of the present disclosure provides a method for handling a fault of an automatic driving system of a vehicle, the method comprising:

[0006] Obtaining a fault duration of an automatic driving system in a vehicle and a set of faults at each moment within the fault duration;

[0007] For each fault set at a moment, according to the vehicle state of the vehicle, determining the predicted driving distance after the vehicle executes each behavior combination, wherein the behavior combination includes the vehicle driving behavior and the fault recovery behavior corresponding to the fault in the fault set;

[0008] According to the predicted driving distance of each behavior combination, the reference driving behavior corresponding to the fault set and the reference recovery behavior of each fault are determined, and the reference driving behavior corresponding to each fault set within the fault duration and the reference recovery behavior of each fault are obtained.

[0009] In a second aspect, an embodiment of the present disclosure further provides a fault handling device for a vehicle automatic driving system, the device comprising:

[0010] A fault acquisition module, used to acquire a fault duration of the automatic driving system in the vehicle and a set of faults at each moment within the fault duration;

[0011] A distance prediction module, for determining, for each fault set at a moment, a predicted driving distance after the vehicle executes each behavior combination according to the vehicle state of the vehicle, wherein the behavior combination includes the vehicle driving behavior and the fault recovery behavior corresponding to the fault in the fault set;

[0012] The behavior determination module is used to determine the reference driving behavior corresponding to the fault set and the reference recovery behavior of each fault according to the predicted driving distance of each behavior combination, and obtain the reference driving behavior corresponding to each fault set within the fault duration and the reference recovery behavior of each fault.

[0013] In the third aspect, an embodiment of the present disclosure also provides an electronic device, comprising: one or more processors; a storage device for storing one or more programs; when the one or more programs are executed by the one or more processors, the one or more processors implement the fault handling method of the vehicle automatic driving system as described above.

[0014] In a fourth aspect, an embodiment of the present disclosure further provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the fault handling method for the vehicle automatic driving system as described above.

[0015] A method for handling a fault in an automatic driving system of a vehicle provided by an embodiment of the present disclosure obtains a fault duration of the automatic driving system in the vehicle and a set of faults at each moment within the fault duration, and then for each fault set at each moment, determines a predicted driving distance after the vehicle executes each behavior combination according to the vehicle state, and determines a reference driving behavior corresponding to the fault set and a reference recovery behavior for each fault according to the predicted driving distance of each behavior combination, thereby obtaining reference driving behaviors corresponding to each fault set within the fault duration and a reference recovery behavior for each fault, thereby handling a fault of the automatic driving system of the vehicle. By determining the predicted driving distance of each behavior combination, the method aims to ensure safety and maintain good maneuverability when the vehicle copes with a fault, and determines the reference driving behavior of the vehicle and the reference recovery behavior of each fault from each behavior combination of the fault handling, thereby optimizing the decision between the vehicle's driving handling and the recovery handling on the premise of ensuring vehicle safety, improving the overall performance of the vehicle in a fault condition, and realizing a safer and more efficient automatic driving fault handling function. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] The above and other features, advantages and aspects of the embodiments of the present disclosure will become more apparent with reference to the following detailed description in conjunction with the accompanying drawings. Throughout the accompanying drawings, the same or similar reference numerals represent the same or similar elements. It should be understood that the drawings are schematic and the originals and elements are not necessarily drawn to scale.

[0017] Figure 1 is a flow chart of a method for troubleshooting a vehicle automatic driving system in an embodiment of the present disclosure;

[0018] Figure 2 is a flowchart of a fault handling method of a vehicle automatic driving system in another embodiment of the present disclosure;

[0019] Figure 3 is a flowchart of a fault handling method of a vehicle automatic driving system in another embodiment of the present disclosure;

[0020] Figure 4 A schematic diagram of a game tree in an embodiment of the present disclosure;

[0021] Figure 5 A schematic diagram of updating the revenue of a game tree in an embodiment of the present disclosure;

[0022] Figure 6 A schematic diagram of selecting a combination sequence in an embodiment of the present disclosure;

[0023] Figure 7 It is a structural schematic diagram of a fault handling device of a vehicle automatic driving system in an embodiment of the present disclosure;

[0024] Figure 8 It is a structural schematic diagram of an electronic device in an embodiment of the present disclosure. DETAILED DESCRIPTION

[0025] Embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. Although certain embodiments of the present disclosure are shown in the accompanying drawings, it should be understood that the present disclosure can be implemented in various forms and should not be construed as being limited to the embodiments described herein, which are instead provided for a more thorough and complete understanding of the present disclosure. It should be understood that the drawings and embodiments of the present disclosure are only for exemplary purposes and are not intended to limit the scope of protection of the present disclosure.

[0026] It should be noted that the concepts such as "first" and "second" mentioned in the present disclosure are only used to distinguish different devices, modules or units, and are not used to limit the order or interdependence of the functions performed by these devices, modules or units.

[0027] The names of the messages or information exchanged between multiple devices in the embodiments of the present disclosure are only used for illustrative purposes and are not used to limit the scope of these messages or information.

[0028] Figure 11 is a flow chart of a method for troubleshooting a vehicle automatic driving system in an embodiment of the present disclosure. The method can be executed by a fault handling device for a vehicle automatic driving system, which can be implemented in software and / or hardware, and can be configured in an electronic device. Figure 1 As shown, the method may specifically include the following steps:

[0029] S110, obtaining a fault duration of the automatic driving system in the vehicle and a set of faults at each moment within the fault duration.

[0030] In the disclosed embodiment, the autonomous driving system may be composed of multiple modules, such as a regulation and control module, a perception module, a positioning module, and a chassis module. A failure in the autonomous driving system may specifically refer to a failure of a module in the autonomous driving system.

[0031] Considering that the autonomous driving system may face different fault conditions at different times, specifically, the status information of each module can be obtained, and whether each module has a fault can be determined based on the status information, so as to obtain the fault duration and the fault set at each time within the duration. At the same time, when a fault is detected, the vehicle status, fault information (such as the number of occurrences of the fault) and surrounding environment information of the vehicle can also be obtained.

[0032] The fault duration may refer to the total duration of the fault of the autonomous driving system; the fault set at each moment may be composed of all faults detected at that moment. It should be noted that the fault set at a moment may include faults occurring in multiple modules, or multiple faults occurring in one module.

[0033] For example, the fault duration can be represented by T, and the set of all faults under the fault duration can be represented as: Γ = {N i |i=1,2,...,T},N i Represents the set of faults at time i within the fault duration.

[0034] In the disclosed embodiment, the fault handling behavior of the autonomous driving system can be determined by a fault handling game model. The fault handling game model may involve concepts such as players, game strategies, game benefits, and game constraints. Among them, the player is the object of participating in the behavior selection. In the disclosed embodiment, the fault is the player, and the fault set can be understood as the player set.

[0035] S120. For each fault set at a moment, determine the predicted driving distance after the vehicle executes each behavior combination according to the vehicle state, wherein the behavior combination includes the vehicle driving behavior and fault recovery behavior corresponding to the fault in the fault set.

[0036] In the disclosed embodiment, the vehicle driving behavior and the fault recovery behavior may be strategies for handling faults, that is, they may be understood as player strategies.

[0037] The vehicle driving behavior may be a driving control strategy for the vehicle in response to a fault. For example, the vehicle driving behavior may include no braking, low speed (such as driving at a speed of 1 m / s), slow braking (such as driving at a speed of -1 m / s), and braking slowly. 2 acceleration brake), emergency brake (such as -2m / s 2 acceleration braking).

[0038] Among them, the fault recovery behavior can be the recovery control strategy of the autonomous driving system for the fault. Exemplarily, the fault recovery behavior can include non-automatic recovery, automatic recovery, and limited automatic recovery. Limited automatic recovery can be automatic recovery when the number of occurrences of the fault does not exceed the set number, and non-automatic recovery when the number of occurrences of the fault exceeds the set number. For non-automatic recovery, the specific performance of the vehicle may be parking and unable to resume autonomous driving; for automatic recovery, the specific performance of the vehicle may be continued driving, or parking first and then resuming autonomous driving.

[0039] In the disclosed embodiment, a behavior combination can be understood as a player strategy combination, and a behavior combination is composed of a vehicle driving behavior for a fault and a fault recovery behavior; a vehicle executes a behavior combination, specifically, first executes the vehicle driving behavior in the behavior combination, and then executes the fault recovery behavior in the behavior combination.

[0040] In addition, the predicted driving distance can be the distance that the vehicle can travel after executing a combination of behaviors. Considering that in the game process, the distance traveled by the vehicle can reflect the pros and cons of the strategy to a certain extent, the longer the driving distance, the more likely the strategy can ensure safety and maintain good mobility when dealing with failures, thereby obtaining relatively high benefits. Therefore, the predicted driving distance can be used as the benefit in the game process, and the quality of each strategy can be evaluated by the predicted driving distance, so as to select the optimal strategy.

[0041] Specifically, for each fault set at a moment, each behavior combination for the fault can be determined according to each fault in the fault set. Further, according to the vehicle state of the vehicle, the state of the vehicle after executing each behavior combination is predicted, and then the predicted driving distance of the vehicle is determined according to the predicted state. The vehicle state can be the actual driving state corresponding to each moment within the duration of the fault, including the vehicle position, vehicle speed, and vehicle acceleration.

[0042] Exemplarily, taking four vehicle driving behaviors and three fault recovery behaviors as an example, assuming that the number of faults in the fault set is one, 4×3 behavior combinations can be determined for this fault, and the predicted driving distances of 12 behavior combinations can be obtained in combination with the vehicle status.

[0043] S130. Determine the reference driving behavior corresponding to the fault set and the reference recovery behavior of each fault according to the predicted driving distance of each behavior combination, and obtain the reference driving behavior corresponding to each fault set within the fault duration, and the reference recovery behavior of each fault.

[0044] Specifically, for each fault set at each moment, after obtaining the predicted driving distance of each behavior combination, we can take the maximum predicted driving distance as the goal, that is, maximize the benefit at that moment, determine the optimal vehicle driving behavior for the entire fault set as the reference driving behavior, and determine the optimal fault recovery behavior for each fault in the entire fault set as the reference recovery behavior for each fault.

[0045] Exemplarily, following the above example, the behavior combination with the largest predicted driving distance can be screened out based on the predicted driving distances of the 12 behavior combinations, and the vehicle driving behavior and fault recovery behavior in the behavior combination can be respectively determined as the reference driving behavior and reference recovery behavior.

[0046] It should be noted that in the embodiments of the present disclosure, the fault handling strategy is determined by means of game theory, and specifically two methods can be used, namely static game theory and dynamic game theory. Among them, static game theory can be a one-time selection of the optimal vehicle driving behavior and fault recovery behavior to maximize the benefits at that moment, which is suitable for scenarios where the fault situation is relatively simple, does not change much, and it is difficult to accurately predict the future development of the fault. Dynamic game theory can be to continuously predict and evaluate the benefits under different combinations of behaviors in the future, and select a sequence of behavior combinations that can maximize the predicted driving distance during the entire game process. It is suitable for scenarios where the fault situation is relatively complex and there may be multiple changes and mutual influences.

[0047] Specifically, according to the number of faults in the fault set, such as first determining whether the number of faults in the fault set is single, static game or dynamic game can be selected to determine the final fault handling strategy. The specific process of static game and dynamic game can be seen in the examples below and will not be repeated here.

[0048] In the disclosed embodiment, the above method can be used at each moment during the fault duration to obtain the reference driving behavior and reference recovery behavior at each moment, and output them to the decision module. The decision module can control the vehicle based on the reference driving behavior and reference recovery behavior.

[0049] The fault handling method of the vehicle automatic driving system provided by the embodiment of the present disclosure obtains the fault duration of the automatic driving system in the vehicle and the fault set at each moment within the fault duration, and then for each fault set at each moment, determines the predicted driving distance of the vehicle after executing each behavior combination according to the vehicle state, and determines the reference driving behavior corresponding to the fault set and the reference recovery behavior of each fault according to the predicted driving distance of each behavior combination, so as to obtain the reference driving behavior corresponding to each fault set within the fault duration and the reference recovery behavior of each fault, so as to realize the fault handling of the vehicle automatic driving system. By determining the predicted driving distance of each behavior combination, the method aims to ensure safety and maintain good maneuverability when the vehicle copes with a fault, and determines the reference driving behavior of the vehicle and the reference recovery behavior of each fault from each behavior combination of the fault handling, so as to optimize the decision between the vehicle's driving processing and the recovery processing on the premise of ensuring the safety of the vehicle, improve the overall performance of the vehicle in the case of a fault, and realize a safer and more efficient automatic driving fault handling function.

[0050] Figure 2 FIG. 1 is a flow chart of a method for handling a fault in an automatic driving system of a vehicle in another embodiment of the present disclosure. The method describes the use of static game theory to determine a handling strategy for a fault. Figure 2 As shown, the fault handling method of the vehicle automatic driving system provided by the embodiment of the present disclosure includes the following steps:

[0051] S210, obtaining a fault duration of the automatic driving system in the vehicle and a set of faults at each moment within the fault duration.

[0052] S220 . For each fault set at a moment, when the number of faults in the fault set is single, combine all vehicle driving behaviors and all fault recovery behaviors to obtain all behavior combinations.

[0053] In the disclosed embodiment, static game theory can be applied to the situation where a module in the autonomous driving system fails and no further failures occur. In this case, static game theory can be used to make quick decisions to deal with the current emergency situation.

[0054] Specifically, for a set of faults at a certain moment, we can first determine whether the number of faults in the set is single. If so, it means that only one module of the autonomous driving system has a fault at that moment, and the optimal solution can be determined through a one-time game. Among them, the static game first needs to determine all possible strategy combinations.

[0055] For example, any vehicle driving behavior and any fault recovery behavior can be selected and combined, and this operation can be repeated to obtain all behavior combinations. For example, taking four vehicle driving behaviors and three fault recovery behaviors as an example, 4×3 behavior combinations can be obtained.

[0056] S230. For each behavior combination, determine a predicted driving distance after the vehicle executes the behavior combination according to the vehicle state.

[0057] Specifically, after obtaining each behavior combination, the state of the vehicle after executing each behavior combination can be predicted according to the vehicle state, so as to determine the predicted driving distance of the vehicle after executing the behavior combination according to the predicted state.

[0058] In the disclosed embodiment, in order to ensure the safety of the vehicle in handling faults, the behavior combinations may be screened before the predicted driving distance of each behavior combination is determined to filter out behavior combinations with low safety after the vehicle executes them.

[0059] In a specific implementation, before determining, for each behavior combination, a predicted driving distance after the vehicle executes the behavior combination according to the vehicle state, the method further includes:

[0060] For each behavior combination, the predicted state of the vehicle after executing the behavior combination is determined according to the vehicle state; whether the predicted state meets the preset safety constraints is determined, and if not, the behavior combination is eliminated.

[0061] Specifically, the predicted state after the vehicle executes the combined behavior can be determined based on the vehicle state, and then it can be determined whether the predicted state meets the preset safety constraints. The preset safety constraints can be pre-set constraints to ensure the driving safety of the vehicle, such as no collision risk with obstacles, no collision risk with road network boundaries, etc.

[0062] In one example, the preset safety constraints include a preset road network constraint and a preset obstacle distance constraint; wherein, the preset road network constraint is that the distance between the vehicle and the road network tangent is greater than a preset first distance, and the preset obstacle distance constraint is that the distance between the vehicle and the obstacle is greater than a preset second distance.

[0063] Specifically, the distance between the vehicle and the road network tangent can be determined according to the predicted state, and then it can be determined whether the distance is greater than a preset first distance. If so, it is determined that the predicted state satisfies the preset road network constraint. The preset first distance can be determined according to the width of the vehicle body, for example, the preset first distance can be half of the width of the vehicle body.

[0064] Furthermore, the distance between the vehicle and each obstacle can be determined according to the predicted state, and then it can be determined whether the distance between the vehicle and each obstacle is greater than a preset second distance. If so, it is determined that the predicted state satisfies the preset obstacle distance constraint. The preset second distance can be a preset minimum safety distance constant, and different types of obstacles can correspond to different preset second distances, for example, the preset second distance corresponding to pedestrians is greater than the preset second distance corresponding to motor vehicles.

[0065] Through the above-mentioned preset road network constraints, when screening behavior combinations, behavior combinations that will exceed the road network range after the vehicle executes can be eliminated, thereby avoiding safety accidents and ensuring the safety of fault handling; and through the above-mentioned preset obstacle distance constraints, when screening behavior combinations, behavior combinations that may cause the vehicle to collide with obstacles after execution can be eliminated, thereby avoiding collision risks and ensuring the safety of fault handling.

[0066] Specifically, the behavior combinations whose predicted states satisfy the preset safety constraints can be retained, and the behavior combinations whose predicted states do not satisfy the preset safety constraints can be eliminated to ensure the safety of the vehicle's handling of faults.

[0067] S240. Determine the reference driving behavior corresponding to the fault set and the reference recovery behavior of each fault according to the predicted driving distance of each behavior combination, and obtain the reference driving behavior corresponding to each fault set within the fault duration, and the reference recovery behavior of each fault.

[0068] Specifically, the goal may be to maximize the predicted driving distance, select the behavior combination with the largest predicted driving distance from various behavior combinations, and obtain the reference driving behavior and the reference recovery behavior based on the selected behavior combination. For example, the following formula may be used:

[0069]

[0070] In the formula, S(i,t) represents the predicted driving trajectory at time i, is the behavior combination for fault k at time i, is the behavior combination for other faults except fault k at time i, Divide by time i In the static game process (i.e., the fault set contains only a single fault), the above formula can be understood as selecting the behavior combination with the largest predicted driving trajectory from all the behavior combinations for the fault.

[0071] The fault handling method for the vehicle automatic driving system provided by the embodiment of the present disclosure, when the number of faults in the fault set is single, obtains all behavior combinations by combining all vehicle driving behaviors and all fault recovery behaviors, and then determines the predicted driving trajectory under each behavior combination respectively, and determines the reference driving behavior and the reference recovery behavior according to the size of the predicted driving trajectory, thereby realizing the static game of a single fault. By selecting the optimal vehicle driving behavior and fault recovery behavior at one time, a fault decision can be made quickly to ensure the efficiency of vehicle fault handling.

[0072] Figure 3 FIG. 1 is a flow chart of a method for handling a fault in an automatic driving system of a vehicle in another embodiment of the present disclosure, which describes the use of dynamic game theory to determine a handling strategy for a fault. Figure 3 As shown, the fault handling method of the vehicle automatic driving system provided by the embodiment of the present disclosure includes the following steps:

[0073] S310, obtaining the fault duration of the automatic driving system in the vehicle and the fault set at each moment within the fault duration.

[0074] S320. For each fault set at each moment, when there are multiple faults in the fault set, the vehicle state is used as the root node of the game tree. Starting from the root node, according to the order of occurrence of each fault in the fault set, nodes are extended in turn according to all behavior combinations of each fault to obtain the game tree corresponding to the fault set.

[0075] In the disclosed embodiments, dynamic game theory can be applied to situations where a module in an autonomous driving system sends multiple faults, or multiple modules fail. Considering that multiple faults may be related to or influence each other, the dynamic game theory can be used to comprehensively consider the changes in various factors and make better long-term decisions.

[0076] Specifically, for a set of faults at a certain moment, we can first determine whether the number of faults in the fault set is single. If not, it means that the autonomous driving system has multiple faults at this moment. By continuously predicting and evaluating the benefits under different behavior combinations in the future, we can select the combination sequence that can maximize the predicted driving distance in the entire game process.

[0077] Among them, the dynamic game needs to first follow the order of occurrence of each fault in the fault set, according to all the behavior combinations of each fault, starting from the root node (i.e., the vehicle state), and extending downward in sequence to obtain the game tree for the fault set.

[0078] In a specific implementation, starting from the root node, according to the occurrence order of each fault in the fault set, nodes are extended in turn according to all behavior combinations of each fault to obtain a game tree corresponding to the fault set, including the following steps:

[0079] Step 321: sort the faults according to the order in which the faults in the fault set occur, take the first fault in the fault set as the fault to be processed, and take the root node as the node to be extended;

[0080] Step 322: determine all behavior combinations for the fault to be processed, and determine the predicted state of the vehicle after executing each behavior combination according to the node to be extended, and obtain the child node of the node to be extended;

[0081] Step 323: take the next fault of the pending faults in the fault set as a new pending fault, and take the child node as a new node to be extended, and return to execute the step of determining all behavior combinations for the pending faults until the pending fault is the last fault in the fault set.

[0082] In step 321, all faults in the fault set may be sorted according to the order in which the faults in the fault set occur, and then the first fault therein may be used as a fault to be processed, and the root node may be used as a node to be extended.

[0083] Furthermore, in step 322, all vehicle driving behaviors and all fault recovery behaviors can be combined to obtain all behavior combinations of the faults to be processed, and then according to the node to be extended, the state of the vehicle after executing each behavior combination is predicted to obtain the predicted state of the vehicle, and each predicted state is used as a child node of the node to be extended.

[0084] For example, Figure 4 is a schematic diagram of a game tree in an embodiment of the present disclosure, such as Figure 4 As shown, the root node is node 0, which can be used as the node to be extended. For fault 1, there are behavior combination A, behavior combination B and behavior combination C. Fault 1 is a fault to be processed. The states after executing A, B and C can be predicted respectively according to the node to be extended to obtain child nodes 1-1, 1-2 and 1-3.

[0085] Furthermore, in step 323, the next fault of the pending fault can be taken as a new pending fault, and the child node can be taken as a new node to be extended, so as to continue to determine all behavior combinations of the pending fault, and determine the predicted state of the vehicle after executing each behavior combination of the pending fault based on the node to be extended, and obtain the child node, and repeat this process until the pending fault is the last fault in the fault set.

[0086] See also Figure 4After obtaining nodes 1-1, 1-2, and 1-3, nodes 1-1, 1-2, and 1-3 can be used as new nodes to be extended, and fault 2 can be used as a new fault to be processed, and the child nodes of each node to be extended can be extended downward. Taking node 1-1 as an example, assuming that there are two behavior combinations for fault 2, the state after executing behavior combinations D and E can be predicted based on node 1-1, and the child nodes 2-1 and 2-2 of node 1-1 can be obtained. Referring to this process, all nodes extended for fault 2 can be obtained, and by analogy, all nodes extended for fault 3 can be obtained.

[0087] It should be noted that in the process of extending nodes to construct a game tree, in order to ensure the safety of the vehicle in handling faults, node extension can also be performed in combination with preset safety constraints. In one example, before determining the predicted state of the vehicle after executing each combination of behaviors according to the node to be extended and obtaining the child nodes of the node to be extended, it also includes:

[0088] Determine whether the node to be extended meets the preset safety constraints; if so, execute the step of determining the predicted state of the vehicle after executing each combination of behaviors according to the node to be extended, otherwise, use the node to be extended as the leaf node of the game tree.

[0089] Specifically, before extending the node to be extended to obtain its node, it can be determined whether the node to be extended meets the preset safety constraint. The preset safety constraint may include a preset road network constraint and a preset obstacle distance constraint.

[0090] If the preset safety constraints are met, the predicted state of the vehicle after executing each combination of behaviors can be determined based on the node to be extended to obtain the child nodes of the node to be extended. If the preset safety constraints are not met, the node to be extended can be used as the leaf node of the game tree and will not be extended downward.

[0091] For example, Figure 4 As shown, Figure 4 Nodes 2-2, 2-3, and 2-5 in the game tree may be nodes that do not meet the preset safety constraints. Such nodes are regarded as leaf nodes of the game tree and are not extended downward. In this way, it is possible to avoid extending nodes with safety risks downward. While ensuring the safety of vehicle fault handling, it is also possible to reduce the scale of the game tree, thereby improving the efficiency of dynamic games.

[0092] Through the above steps 321 to 323, for a set of faults at a moment, decisions can be made in turn at that moment in the order in which the faults occur, thereby drawing a game tree to facilitate subsequent evaluation of the benefits under different behavior combinations, and to help find the behavior combination sequence that maximizes the benefits of the entire fault handling process.

[0093] S330. Determine the terminal non-leaf node of the game tree as the node to be played. For each node to be played, determine the optimal behavior combination under the node to be played based on the predicted driving distance of each behavior combination under the node to be played. And based on the predicted driving distance of the optimal behavior combination, update the predicted driving distance of the behavior combination corresponding to the node to be played under the parent node of the node to be played.

[0094] After constructing the game tree corresponding to the fault set, we can further start from the bottom up and update the benefits of its parent node according to the benefits corresponding to each node.

[0095] Specifically, the terminal non-leaf node of the game tree can be started and the terminal non-leaf node can be determined as the node to be played. The terminal non-leaf node can refer to a non-leaf node at the end of the game tree, and a non-leaf node refers to a node without a child node. Figure 4 For example, the terminal non-leaf nodes may include node 2-1 and node 2-4.

[0096] For each node to be gamed, the behavior combination with the largest predicted driving distance can be selected according to the predicted driving distance of each behavior combination under the node to be gamed as the optimal behavior combination under the node to be gamed. Figure 5 This is a schematic diagram of a game tree profit update in an embodiment of the present disclosure. The optimal behavior combination under node 2-1 is I, and the optimal behavior combination under node 2-4 is M.

[0097] Furthermore, the predicted driving distance of the behavior combination related to the node to be played (i.e., the behavior combination that extends the node to be played) under the parent node of the node to be played can be updated according to the driving distance of the optimal behavior combination under the node to be played.

[0098] like Figure 5 As shown, for node 2-1, the optimal behavior combination under it is I. Then, according to the benefit _I of I, the benefit of the behavior combination D under the parent node 1-1 can be updated to benefit _I. For node 2-4, the optimal behavior combination under it is M. Then, according to the benefit _M of M, the benefit of the behavior combination G under the parent node 1-3 can be updated to benefit _M.

[0099] S340, taking the parent node as a new node to be played, returning to execute the step of determining the optimal behavior combination under the node to be played according to the predicted driving distance of each behavior combination under the node to be played, until the parent node of the node to be played is the root node.

[0100] Specifically, after updating the benefits of the behavior combination related to the node to be played under the parent node of the node to be played, the parent node of the node to be played can be re-used as the new node to be played, thereby repeating the above operation, that is, for each node to be played, determining the optimal behavior combination under the node to be played, and then updating the benefits of the behavior combination related to the node to be played under the parent node according to the benefits of the optimal behavior combination, and so on, until the parent node of the node to be played is the root node.

[0101] Continue to use Figure 5 In the example, for node 1-1, the optimal behavior combination under it is D. Then, according to the benefit_I of D, the benefit of the behavior combination A under the parent node 0 can be updated to benefit_I. For node 1-3, the optimal behavior combination under it is G. Then, according to the benefit_M of G, the benefit of the behavior combination C under the parent node 0 can be updated to benefit_M.

[0102] S350. Determine the reference driving behavior corresponding to the fault set and the reference recovery behavior of each fault according to the predicted driving distance of each behavior combination, and obtain the reference driving behavior corresponding to each fault set within the fault duration and the reference recovery behavior of each fault.

[0103] Specifically, after constructing the game tree and updating the benefits (predicted driving distances) of some behavior combinations from bottom to top, the reference driving behavior corresponding to the fault set and the reference recovery behavior of each fault can be determined based on the benefits.

[0104] In a specific implementation, determining a reference driving behavior corresponding to a fault set and a reference recovery behavior for each fault according to the predicted driving distance of each behavior combination includes the following steps:

[0105] Step 351: According to the optimal behavior combination under each node in the game tree, determine the combination sequence composed of each optimal behavior combination;

[0106] Step 352: Determine a reference recovery behavior for each fault in the fault set according to the fault recovery behavior in each combination in the combination sequence;

[0107] Step 353: Determine a reference driving behavior for the fault set according to the vehicle driving behaviors in each combination in the combination sequence and the preset driving behavior priority.

[0108] Among them, in step 351, starting from the root node of the game tree, the optimal behavior combination under the root node can be first placed in the combination sequence, and then the optimal behavior combination under the child node corresponding to the optimal behavior combination can be placed in the combination sequence, and this process is repeated until the optimal behavior combination for the last fault in the fault set is placed in the combination sequence, and a combination sequence composed of multiple optimal behavior combinations is obtained.

[0109] For example, Figure 6 is a schematic diagram of selecting a combination sequence in an embodiment of the present disclosure, such as Figure 6 As shown, starting from the root node of the game tree, the optimal behavior combination under node 0 is A, the optimal behavior combination under the child node 1-1 corresponding to A is D, and the optimal behavior combination under the child node 2-1 corresponding to D is I, then the combination sequence can be composed of A, D, and I.

[0110] Alternatively, in step 351, we can also start from the root node of the game tree, put all the behavior combinations under the root node into different combination sequences, and then select the best behavior combination under each child node of the root node and put it into the corresponding combination sequence, and repeat this process to obtain multiple combination sequences. Figure 6 As shown, combination sequence 1 includes A, D, and I, combination sequence 2 includes B and F, and combination sequence 3 includes C, G, and K.

[0111] Furthermore, in step 352, the fault recovery behaviors in each combination in the combination sequence may be determined as reference behaviors for each fault in the fault set. Figure 6 , the fault recovery behaviors in A, D, and I can be used as the reference recovery behavior for fault 1, the reference recovery behavior for fault 2, and the reference recovery behavior for fault 3, respectively.

[0112] Furthermore, in step 353, the vehicle driving behaviors in each combination in the combination sequence may be extracted first, such as Figure 6 The vehicle driving behaviors of A, D, and I in the above example are considered, and considering that the vehicle usually only adopts one driving behavior for a fault set at a moment, a reference driving behavior can be selected from all the extracted vehicle driving behaviors according to the preset driving behavior priority. The preset driving behavior priority can be determined according to the distance that the vehicle can travel after executing each driving behavior. For example, the driving behavior priority is: no braking > low speed > slow braking > sudden braking.

[0113] For example, among the vehicle driving behaviors of A, D, and I, the vehicle driving behavior of D has the highest priority, so the vehicle driving behavior of D can be used as the reference driving behavior. It should be noted that if there are multiple combination sequences, each combination sequence can determine a reference driving behavior and a reference recovery behavior for each fault.

[0114] Through the above steps 351 to 353, a strategy sequence that can maximize the vehicle driving distance in the entire game process can be selected, and then a reference driving behavior for the fault set and a reference recovery behavior for each fault can be obtained, thereby ensuring the reliability of fault handling under multiple fault conditions.

[0115] In the disclosed embodiment, for each moment within the duration of the fault, the above process can be used to obtain the reference driving behavior for the fault set at that moment and the reference recovery behavior for each fault, and provide them to the decision module. If there are multiple reference strategies (including reference driving behaviors and reference recovery behaviors for each fault) output to the decision module for the fault set at a moment, the decision module can determine the rationality of each reference strategy and then control the vehicle.

[0116] The fault handling method of the vehicle automatic driving system provided by the embodiment of the present disclosure, when there are multiple faults in the fault set, takes the vehicle state as the root node, extends each node downward to obtain a game tree according to the occurrence order of the faults in the fault set, and then updates the predicted driving distances of some behavior combinations in the game tree from bottom to top. Finally, according to the predicted driving distances of each behavior combination, a combination sequence that can maximize the predicted driving distance of the vehicle in the entire game process is selected to make accurate long-term decisions, thereby ensuring the processing safety and reliability of the automatic driving system under multiple fault conditions.

[0117] Figure 7 FIG. 1 is a schematic diagram of a fault handling device for a vehicle automatic driving system in an embodiment of the present disclosure. The device can execute the method provided in any embodiment of the present disclosure, such as Figure 7 As shown, the device includes: a fault acquisition module 710 , a distance prediction module 720 and a behavior determination module 730 .

[0118] A fault acquisition module 710 is used to acquire a fault duration of the automatic driving system in the vehicle and a set of faults at each moment within the fault duration;

[0119] The distance prediction module 720 is used to determine, for each fault set at a moment, a predicted driving distance after the vehicle executes each behavior combination according to the vehicle state of the vehicle, wherein the behavior combination includes the vehicle driving behavior and the fault recovery behavior corresponding to the fault in the fault set;

[0120] The behavior determination module 730 is used to determine the reference driving behavior corresponding to the fault set and the reference recovery behavior of each fault according to the predicted driving distance of each behavior combination, and obtain the reference driving behavior corresponding to each fault set within the fault duration, and the reference recovery behavior of each fault.

[0121] The fault handling device of the vehicle automatic driving system provided in the embodiment of the present disclosure can execute the steps of the fault handling method of the vehicle automatic driving system provided in the embodiment of the method of the present disclosure, and the execution steps and beneficial effects are no longer repeated here.

[0122] Figure 8Schematic diagram of the structure of an electronic device in the embodiment of the present disclosure. Figure 8 , which shows a structural schematic diagram of an electronic device 500 suitable for implementing the embodiments of the present disclosure. Figure 8 The electronic device shown is only an example and should not bring any limitation to the functions and scope of use of the embodiments of the present disclosure.

[0123] like Figure 8 As shown, the electronic device 500 may include a processing device 501, a ROM 502, a RAM 503, a bus 504, an input / output (I / O) interface 505, an input device 506, an output device 507, a storage device 508, and a communication device 509. The processing device (e.g., a central processing unit, a graphics processor, etc.) 501 can perform various appropriate actions and processes to implement the method of the embodiment described in the present disclosure according to the program stored in the read-only memory (ROM) 502 or the program loaded from the storage device 508 to the random access memory (RAM) 503. In the RAM 503, various programs and data required for the operation of the electronic device 500 are also stored. The processing device 501, the ROM 502, and the RAM 503 are connected to each other through the bus 504. The input / output (I / O) interface 505 is also connected to the bus 504.

[0124] In particular, according to an embodiment of the present disclosure, the process described above with reference to the flowchart can be implemented as a computer software program. For example, an embodiment of the present disclosure includes a computer program product, which includes a computer program carried on a non-transitory computer-readable medium, and the computer program contains a program code for executing the method shown in the flowchart, thereby implementing the fault handling method of the vehicle automatic driving system as described above. In such an embodiment, the computer program can be downloaded and installed from the network through the communication device 509, or installed from the storage device 508, or installed from the ROM 502. When the computer program is executed by the processing device 501, the above-mentioned functions defined in the method of the embodiment of the present disclosure are executed.

[0125] It should be noted that the computer-readable medium disclosed above may be a computer-readable signal medium or a computer-readable storage medium or any combination of the above two. The computer-readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or device, or any combination of the above. More specific examples of computer-readable storage media may include, but are not limited to: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In the present disclosure, a computer-readable storage medium may be any tangible medium containing or storing a program that may be used by or in combination with an instruction execution system, device or device. In the present disclosure, a computer-readable signal medium may include a data signal propagated in a baseband or as part of a carrier wave, in which a computer-readable program code is carried. This propagated data signal may take a variety of forms, including but not limited to an electromagnetic signal, an optical signal, or any suitable combination of the above. The computer readable signal medium may also be any computer readable medium other than a computer readable storage medium, which may send, propagate or transmit a program for use by or in conjunction with an instruction execution system, apparatus or device. The program code contained on the computer readable medium may be transmitted using any suitable medium, including but not limited to: wires, optical cables, RF (radio frequency), etc., or any suitable combination of the above.

[0126] The computer-readable medium may be included in the electronic device, or may exist independently without being installed in the electronic device. The computer-readable medium carries one or more programs, and when the one or more programs are executed by the electronic device, the electronic device:

[0127] Obtaining a fault duration of an automatic driving system in a vehicle and a set of faults at each moment within the fault duration;

[0128] For each fault set at a moment, according to the vehicle state of the vehicle, determining the predicted driving distance after the vehicle executes each behavior combination, wherein the behavior combination includes the vehicle driving behavior and the fault recovery behavior corresponding to the fault in the fault set;

[0129] According to the predicted driving distance of each behavior combination, the reference driving behavior corresponding to the fault set and the reference recovery behavior of each fault are determined, and the reference driving behavior corresponding to each fault set within the fault duration and the reference recovery behavior of each fault are obtained.

[0130] Optionally, when the above one or more programs are executed by the electronic device, the electronic device may also execute other steps described in the above embodiments.

[0131] In the context of the present disclosure, a machine-readable medium may be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, device, or equipment. A machine-readable medium may be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium may include, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, device, or equipment, or any suitable combination of the foregoing. A more specific example of a machine-readable storage medium may include an electrical connection based on one or more lines, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0132] The above description is only a preferred embodiment of the present disclosure and an explanation of the technical principles used. Those skilled in the art should understand that the scope of disclosure involved in the present disclosure is not limited to the technical solutions formed by a specific combination of the above technical features, but should also cover other technical solutions formed by any combination of the above technical features or their equivalent features without departing from the above disclosed concept. For example, the above features are replaced with the technical features with similar functions disclosed in the present disclosure (but not limited to) by each other to form a technical solution.

Claims

1. A method for troubleshooting a vehicle automatic driving system, characterized in that: The method comprises: Obtaining a fault duration of the automatic driving system in the vehicle and a set of faults at each moment within the fault duration; For each fault set at a moment, according to the vehicle state of the vehicle, determining the predicted driving distance after the vehicle executes each behavior combination, wherein the behavior combination includes the vehicle driving behavior and the fault recovery behavior corresponding to the fault in the fault set; According to the predicted driving distance of each behavior combination, the reference driving behavior corresponding to the fault set and the reference recovery behavior of each fault are determined, and the reference driving behavior corresponding to each fault set within the fault duration and the reference recovery behavior of each fault are obtained.

2. The method according to claim 1, characterized in that Determining, according to the vehicle state of the vehicle, a predicted driving distance after the vehicle executes each combination of behaviors, including: In the case where the number of faults in the fault set is single, all vehicle driving behaviors and all fault recovery behaviors are combined to obtain all behavior combinations; For each behavior combination, a predicted driving distance of the vehicle after executing the behavior combination is determined according to the vehicle state of the vehicle.

3. The method according to claim 2, characterized in that Before determining, for each behavior combination, a predicted driving distance of the vehicle after executing the behavior combination according to the vehicle state, the method further includes: For each behavior combination, determining a predicted state of the vehicle after executing the behavior combination according to the vehicle state of the vehicle; Determine whether the predicted state meets the preset safety constraint, and if not, eliminate the behavior combination.

4. The method according to claim 1, characterized in that: Determining, according to the vehicle state of the vehicle, a predicted driving distance after the vehicle executes each combination of behaviors, including: In the case where there are multiple faults in the fault set, the vehicle state is used as the root node of the game tree, and starting from the root node, nodes are extended according to all behavior combinations of each fault in the fault set in the order of occurrence, to obtain a game tree corresponding to the fault set; Determine the terminal non-leaf node of the game tree as a node to be played, determine the optimal behavior combination under the node to be played according to the predicted driving distance of each behavior combination under the node to be played, and update the predicted driving distance of the behavior combination corresponding to the node to be played under the parent node of the node to be played according to the predicted driving distance of the optimal behavior combination; The parent node is used as a new node to be played, and the step of determining the optimal behavior combination under the node to be played according to the predicted driving distance of each behavior combination under the node to be played is returned to execute for each node to be played, until the parent node of the node to be played is the root node.

5. The method according to claim 4, characterized in that Starting from the root node, according to the occurrence order of each fault in the fault set, nodes are extended according to all behavior combinations of each fault in turn to obtain a game tree corresponding to the fault set, including: Sorting the faults according to the order in which the faults in the fault set occur, taking the first fault in the fault set as a fault to be processed, and taking the root node as a node to be extended; Determine all behavior combinations for the fault to be processed, and determine the predicted state of the vehicle after executing each behavior combination according to the node to be extended, and obtain the child node of the node to be extended; The next fault of the pending fault in the fault set is taken as a new pending fault, and the child node is taken as a new node to be extended, and the step of determining all behavior combinations for the pending fault is returned to be executed until the pending fault is the last fault in the fault set.

6. The method according to claim 4, characterized in that Determining a reference driving behavior corresponding to the fault set and a reference recovery behavior for each fault according to the predicted driving distance of each behavior combination includes: According to the optimal behavior combination under each node in the game tree, determining a combination sequence composed of each optimal behavior combination; Determining a reference recovery behavior for each fault in the fault set according to the fault recovery behavior in each combination in the combination sequence; A reference driving behavior for the fault set is determined according to the vehicle driving behavior in each combination in the combination sequence and a preset driving behavior priority.

7. The method according to claim 5, characterized in that Before determining the predicted state of the vehicle after executing each behavior combination according to the node to be extended and obtaining the child node of the node to be extended, the method further includes: Determining whether the node to be extended meets the preset safety constraints; If so, the step of determining the predicted state of the vehicle after executing each combination of behaviors according to the node to be extended is performed; otherwise, the node to be extended is used as a leaf node of the game tree.

8. The method according to claim 3 or 7, characterized in that: The preset safety constraints include preset road network constraints and preset obstacle distance constraints; Among them, the preset road network constraint is that the distance between the vehicle and the road network tangent is greater than a preset first distance, and the preset obstacle distance constraint is that the distance between the vehicle and the obstacle is greater than a preset second distance.

9. An electronic device, characterized in that: The electronic device comprises: one or more processors; A storage device for storing one or more programs; When the one or more programs are executed by the one or more processors, the one or more processors implement the method according to any one of claims 1 to 8.

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

Images