A high-speed automatic driving decision system, method and vehicle based on frequency division triggering

Abstract

The application provides a high-speed automatic driving decision system, method and vehicle based on frequency division triggering, the driving decision system comprises: an environment frame module for processing collected external data and converting the external data into first data; a decision module for generating reasonable scene decisions according to the first data; a scene decision arbitration module for arbitrating whether the scene decisions are feasible and generating an arbitration result; a scene state module for determining a target scene according to the scene decisions and the arbitration result and switching a current scene to the target scene; and a scene corresponding planner for selecting a suitable planner to perform trajectory planning according to the target scene; the trigger conditions of the decision module and the scene decision arbitration module are different, and when the trigger conditions are met, the corresponding modules are triggered; and the trigger conditions are measured by a planning period and a number of arbitration failure frames. The application solves the technical problem that the same planning period is adopted by the planning module and the decision module, which may cause the stability of the planned trajectory to be reduced.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of automatic driving, in particular to a high-speed automatic driving decision system based on frequency division triggering, a high-speed automatic driving decision method based on frequency division triggering and a vehicle. BACKGROUND

[0002] Most of the existing automatic driving systems are synchronous in the same planning cycle, that is, in one planning cycle, the decision module provides the corresponding external information processing results, including but not limited to the driving style of the driver, environmental information, scene information, etc., and then the planning module makes corresponding path planning. The advantage of this is that the corresponding environmental information can be accurately obtained in each planning cycle, and timely judgment can be made. This triggering method is more in line with human driving habits in the city, that is, real-time acquisition of environmental information and corresponding processing, and then conversion into throttle, brake, steering wheel, turn signal action.

[0003] However, in the actual implementation process, there is a problem that the interaction between the decision module and the planning module in the existing decision system adopts a serial mode, and uses time-triggered control, that is, a predetermined publishing frequency is maintained. In some special cases, such as the other vehicle driving on the line, the self-vehicle decision may fluctuate, resulting in reduced stability of the planning trajectory. SUMMARY

[0004] The present application solves the technical problem that the same planning cycle of the planning module and the decision module may cause the stability of the planning trajectory to be reduced.

[0005] To solve the above problems, the present application provides a high-speed automatic driving decision system based on frequency division triggering, which comprises: an environment frame module, the environment frame module is used for processing the collected external data and converting the external data into first data; a decision module, the decision module generates reasonable scene decisions according to the first data; a scene decision arbitration module, the scene decision arbitration module is used for arbitrating whether the scene decisions are feasible to generate arbitration results, wherein the arbitration results include: arbitration pass and arbitration failure; a scene state module, the scene state module is used for determining a target scene according to the scene decisions and the arbitration results, and switching the current scene to the target scene; a scene corresponding planner, the scene corresponding planner is used for selecting a suitable planner for trajectory planning according to the target scene; wherein the triggering conditions of the decision module, the scene decision arbitration module and the scene corresponding planner are different, when the triggering conditions are met, the corresponding modules are triggered; the triggering conditions of the decision module and the scene decision arbitration module are measured by the planning cycle and the number of frames of arbitration failure; the triggering condition of the scene corresponding planner is every 1 planning cycle.

[0006] Compared with the prior art, the technical effects achieved by adopting the technical scheme are as follows: the triggering conditions of the decision module, the scene decision arbitration module and the scene corresponding planner are different, the scene corresponding planner is triggered once every 1 planning period, and the triggering conditions of the decision module and the scene decision arbitration module are measured by the planning period and the number of arbitration failure frames. In the application, different triggering conditions are adopted for the decision module, the scene decision arbitration module and the scene corresponding planner, so as to reduce the calculation pressure on the vehicle-mounted computing power platform; at the same time, by giving different frequencies to different modules, the low-frequency decision and high-frequency planning are started from the user driving behavior, so that the stability can be effectively increased while the performance is ensured. In order to increase the response to special situations, an event triggering method (that is, triggering by measuring the number of arbitration failure frames) is introduced, that is, in the specified dangerous situation, the scene decision is directly reselected, so as to reduce the harm caused by the non-standard driving of other vehicles to the vehicle.

[0007] In an example of the application, the decision module comprises: a scene decision module, the scene decision module being configured to generate a reasonable first decision according to the first data in the case that the ego vehicle does not comply with the lane keeping rule; wherein the lane keeping rule is a rule under which the ego vehicle can continue to perform lane keeping; the first decision comprises: a lever lane change, a free lane change, ramp merging and ramp merging out; the triggering condition of the scene decision module comprises a first time triggering condition and a first event triggering condition; the first time triggering condition is: every a planning period; the first event triggering condition is: the arbitration result is arbitration failure.

[0008] Compared with the prior art, the technical effects achieved by adopting the technical scheme are as follows: the triggering mode of the scene decision module comprises a first time triggering and a first event triggering, and the scene decision module is triggered when the first time triggering and / or the first event triggering are met. The first time triggering is triggered once every a planning period, and a is an integer greater than 1; the event triggering is scene decision arbitration failure. The first time triggering frequency is set to be triggered once every multiple planning periods, that is, low-frequency triggering, because in a high-speed scene, high speed brings higher danger to scenes other than straight driving; lane changing and lane borrowing in a low-speed situation can be easily and frequently performed, but in a high-speed situation, a slight action of the steering wheel can cause terrible consequences; therefore, in a high-speed situation, the stability of the scene is more important, and the switching frequency does not need to be so high. The first event triggering frequency is scene decision arbitration failure, which means that the current scene is switched to the target scene, and due to factors such as external static environment, dynamic environment and human factors, the scene switching is dangerous or is cancelled by human, so the decision needs to be recalculated.

[0009] In an example of the present application, the decision module further comprises: a lane keeping decision module, the lane keeping decision module being configured to generate a reasonable second decision according to the first data when the ego vehicle meets a lane keeping rule; wherein the second decision comprises: cruising and following a vehicle; and the triggering condition of the lane keeping decision module comprises a second time triggering condition and a second event triggering condition; the second time triggering condition is: every 1 planning period; and the second event triggering condition is: consecutive The arbitration result is arbitration failure.

[0010] Compared with the prior art, the technical effects achieved by the technical scheme are as follows: the scene decision module is mainly related to the judgment in the multi-lane and congestion scenes. After the environment frame module ends, the lane keeping decision module and the scene decision module are synchronously entered. In high speed, the lane keeping is the basis of all other scenes, and is the most important and most commonly used scene, and accounts for a large part of the time in normal driving. Therefore, the second time triggering condition is to trigger once every 1 planning period, that is, high-frequency triggering, because the right of way of the ego lane exists, and the lane keeping decision module can also be regarded as a bottom module. The second event triggering frequency is consecutive Frame scene decision arbitration failure means that the current scene is switched to the target scene, and the scene switching exists danger or is canceled by a person for consecutive times, so that the decision needs to be recalculated. When the second time triggering condition and / or the second event triggering condition are met, the lane keeping decision module is triggered.

[0011] In an example of the present application, the triggering condition of the scene decision arbitration module comprises a third time triggering condition and a third event triggering condition; the third time triggering condition is: every planning period; and the third event triggering condition is: consecutive The arbitration result is arbitration failure.

[0012] Compared with the prior art, the technical effects achieved by the technical scheme are as follows: after the corresponding scene decision (including the first decision and / or the second decision) is obtained, it is needed to arbitrate whether the scene decision is feasible. When the third time triggering condition and / or the third event triggering condition are met, the scene decision arbitration module is triggered. The third time triggering condition is to trigger once every 1 planning period; and the third event triggering frequency is consecutive Frame scene decision arbitration failure means that the scene switching exists danger or is canceled by a person for consecutive times, so that the scene decision arbitration module needs to be recalculated.

[0013] In an example of the present application, the scene decision arbitration module comprises: a system definition unit, the system definition unit comprising: trigger conditions, upstream information judgment, environment semantic judgment after preliminary analysis, etc.; a scene judgment unit, the scene judgment unit being used for judging whether the scene decision is feasible; a redundancy judgment unit, the redundancy judgment unit being used for judging whether the scene decision is feasible twice; wherein the system definition unit, the scene judgment unit and the redundancy judgment unit are linearly performed.

[0014] Compared with the prior art, the technical effects achieved by adopting the technical scheme are: the system definition unit, the scene judgment unit and the redundancy judgment unit, three sub-units, are linearly performed. The scene judgment unit refers to deepened judgment of the scene decision (mainly the lane change scene decision), which needs to make more in-depth and detailed analysis of the vehicle scene where the ego vehicle is located according to various types of environment information obtained, indicate the feasibility and danger of the scene decision, and make a judgment once; a redundancy judgment unit is set, which only uses a mathematical method unit to make a secondary judgment of the scene decision, to ensure the feasibility and safety of actually implementing the scene decision.

[0015] In an example of the present application, the scene state module comprises: a scene switching state machine, the scene switching state machine being used for determining a target scene according to the scene decision when the arbitration result is arbitration pass; and a scene state machine, the scene state machine being used for executing the step of switching the current scene to the target scene after the target scene is determined.

[0016] Compared with the prior art, the technical effects achieved by adopting the technical scheme are: the scene switching state machine enters the internal state machine of a single scene, the scene decision is locked until failure or completion, and then the scene switching state machine is re-entered to return to the lane keeping scene to perform the next possible scene switching action. When the target scene is determined, switching from the current scene to the target scene is a gradual process.

[0017] In an example of the present application, the first data comprises: lane line information, obstacle information, prediction information, positioning information, map information, chassis information of the line control, driving style of another vehicle and driving intention of another vehicle.

[0018] Compared with the prior art, the technical effects achieved by adopting the technical scheme are: the environment frame module is used for converting external data into the data types required by the decision module, including: lane line information, obstacle information, prediction information, positioning information, map information, chassis information of the line control, etc. At the same time, on this basis, the corresponding another vehicle reference information is calculated through the model that has been pre-trained or the module that is calculated in real time, including the driving style of another vehicle and the driving intention of another vehicle.

[0019] In another aspect, the embodiment of the present application also provides a high-speed automatic driving decision method based on frequency division triggering, which is applied to the high-speed automatic driving decision system based on frequency division triggering as described in the first embodiment. The high-speed automatic driving decision method comprises: collecting external data and converting the external data into first data; generating a reasonable scene decision according to the first data, wherein the scene decision comprises a first decision and a second decision; arbitrating whether the scene decision is feasible to generate an arbitration result, wherein the arbitration result comprises arbitration pass and arbitration failure; determining a target scene according to the scene decision and the arbitration result, and switching the current scene to the target scene; and selecting a suitable planner for trajectory planning according to the target scene; wherein when the arbitration result is arbitration pass, it means that the scene decision is feasible; and when the arbitration result is arbitration failure, it means that the scene decision is not feasible.

[0020] Compared with the prior art, the technical effects achieved by adopting the technical solution are as follows: the high-speed automatic driving decision method in the embodiment is applied to the high-speed automatic driving decision system based on frequency division triggering as described in the first embodiment, and has all the beneficial effects of the high-speed automatic driving decision system based on frequency division triggering as described in the first embodiment, which will not be described here.

[0021] In one example of the present application, determining a target scene according to the scene decision and the arbitration result, and switching the current scene to the target scene comprises: in the case where the number of the first decisions is multiple, defining the first decision with arbitration pass as a feasible scene; and when the number of the feasible scenes is multiple, determining the target scene according to the priority of the feasible scenes themselves.

[0022] Compared with the prior art, the technical effects achieved by adopting the technical solution are as follows: in the case where the ego vehicle does not comply with the lane keeping rule, after the scene decision module ends generating the first decision, in the case where the number of the first decisions is multiple, arbitrating the multiple first decisions to generate multiple arbitration results, the arbitration result and the first decision correspond one by one. The first decision with arbitration pass is screened out and is defined as a feasible scene. In the case where the number of the feasible scenes is multiple, i.e., in the same time, there may be multiple scene decisions that are established, therefore, it is necessary to prioritize all the feasible scenes that pass, and determine the target scene according to the scene priority. In the case where the ego vehicle complies with the lane keeping rule, after the lane keeping decision module ends generating the second decision, the second decision is directly input to the scene decision arbitration module to arbitrate the corresponding scene.

[0023] In another aspect, the present application also provides a vehicle comprising: a high-speed automatic driving decision system based on frequency division triggering.

[0024] Compared with the prior art, the technical effects achieved by adopting the technical scheme are as follows: the vehicle in the embodiment has all the beneficial effects of the high-speed automatic driving decision system based on frequency division triggering of any embodiment of the application, which will not be described here.

[0025] After adopting the technical scheme of the application, the following technical effects can be achieved:

[0026] (1) Different trigger conditions are adopted for the decision module, the scene decision arbitration module and the scene corresponding planner, so as to reduce the calculation pressure on the vehicle-mounted computing platform; at the same time, by giving different frequencies to different modules, the low-frequency decision and high-frequency planning are started from the user driving behavior, so that the stability can be effectively increased while the performance is ensured.

[0027] (2) In order to increase the response to special situations, an event triggering method (that is, triggering by measuring the number of arbitration failure frames) is introduced, that is, in the specified dangerous situation, the scene decision reselection is directly entered, so as to reduce the harm caused by the non-standard driving of other vehicles to the vehicle. BRIEF DESCRIPTION OF DRAWINGS

[0028] Figure 1 A structure diagram of a high-speed automatic driving decision system based on frequency division triggering is provided for the first embodiment of the application.

[0029] Figure 2 A flowchart of a high-speed automatic driving decision method based on frequency division triggering is provided for the second embodiment of the application.

[0030] REFERENCE SIGNS:

[0031] 100, environment frame module; 200, decision module; 210, scene decision module; 220, lane keeping decision module; 300, scene decision arbitration module; 400, scene state module; 410, scene switching state machine; 420, scene state machine; 500, scene corresponding planner. DETAILED DESCRIPTION

[0032] In order to make the above-mentioned objects, features and advantages of the application more obvious and easy to understand, the technical solutions in the embodiments of the application are described clearly and completely. Obviously, the described embodiments are only part of the embodiments of the application, not all the embodiments. Based on the embodiments in the application, all other embodiments obtained by those skilled in the art without creative labor are within the scope of protection of the application.

[0033]

EMBODIMENT ONE

[0034] Reference is made to Figure 1The application provides a high-speed automatic driving decision system based on frequency division triggering, comprising an environment frame module 100, a decision module 200, a scene decision arbitration module 300, a scene state module 400 and a scene corresponding planner 500, the environment frame module 100 is used for processing collected external data and converting the external data into first data; the decision module 200 generates reasonable scene decisions according to the first data; the scene decision arbitration module 300 is used for arbitrating whether the scene decisions are feasible and generating arbitration results, wherein the arbitration results include arbitration pass and arbitration failure; the scene state module 400 is used for determining a target scene according to the scene decisions and the arbitration results and switching a current scene to the target scene; the scene corresponding planner 500 is used for selecting a suitable planner to perform trajectory planning according to the target scene; wherein the triggering conditions of the decision module 200, the scene decision arbitration module 300 and the scene corresponding planner 500 are different, and when the triggering conditions are met, the corresponding modules are triggered; the triggering conditions of the decision module 200 and the scene decision arbitration module 300 are measured by a planning period and the number of frames of arbitration failure; and the triggering condition of the scene corresponding planner 500 is every 1 planning period.

[0035] In a specific embodiment, in the prior art, in each planning period, the first data collected in each planning period is processed by the decision module 200 first, and corresponding path planning is made by the planning module, if all the information is reprocessed in each planning period, it will cause a great calculation burden on the computing platform, and even it may be unable to interact with other modules according to the agreed period, resulting in serious problems such as information loss, delay and affecting the performance of the automatic driving vehicle. Therefore, the application gives different frequencies to different modules, starts from the user driving behavior, makes low-frequency decisions and high-frequency planning, reduces the calculation pressure on the vehicle computing platform while ensuring performance.

[0036] Specifically, the external data is collected by a laser radar and / or a camera, the environment frame module 100 serves as an interface of the external data and internal data, is mainly responsible for processing the external data and converting it into a data type required by the decision module 200 (that is, the first data); after the environment frame module 100 ends, the decision module 200 is entered, the decision module 200 comprises a scene decision module 210 and a lane keeping decision module 220, reasonable scene decisions are generated by the decision module 200, the number of scene decisions is at least one, and the scene decisions include but are not limited to: a lever lane changing decision, a free lane changing decision, a ramp merging decision, a ramp merging-out decision, a cruise decision and a following vehicle decision.

[0037] After obtaining the corresponding scene decision, it is necessary to arbitrate whether the scene decision is feasible, and the scene decision arbitration module 300 judges whether the scene decision is feasible by generating an arbitration result. When the arbitration result is arbitration pass, it means that the scene decision is feasible; when the arbitration result is arbitration failure, it means that the scene decision is not feasible. The scene state module 400 is used to determine the target scene according to the scene decision and the arbitration result, and switch the current scene to the target scene. The switching from the current scene to the target scene is a gradual process. Because the planning end points required by different target scenes are different, the scene corresponding planner 500 is mainly used to select a suitable planner for trajectory planning according to the final target, to ensure the normal operation of the function.

[0038] Among them, the triggering conditions of the decision module 200, the scene decision arbitration module 300 and the scene corresponding planner 500 are different. The scene corresponding planner 500 is triggered once every 1 planning period. The triggering conditions of the decision module 200 and the scene decision arbitration module 300 are measured by the planning period and the number of frames of arbitration failure. In this application, different triggering conditions are adopted for the decision module 200, the scene decision arbitration module 300 and the scene corresponding planner 500, to reduce the computing pressure on the vehicle-mounted computing platform. At the same time, by giving different frequencies to different modules, starting from the user driving behavior, low-frequency decision and high-frequency planning can effectively increase the stability while ensuring the performance. At the same time, in order to increase the response to special situations, an event triggering method (that is, triggered by the number of frames of arbitration failure) is introduced, that is, in the specified dangerous situation, directly enter the re-selection of scene decision, to reduce the harm caused by the non-standard driving of other vehicles to the vehicle.

[0039] Further, the decision module 200 includes a scene decision module 210, which is used to generate a reasonable first decision according to the first data when the vehicle does not comply with the lane keeping rule. The lane keeping rule is a rule that the vehicle can continue to perform lane keeping. The first decision includes lane changing with a lever, free lane changing, ramp merging and ramp merging. The triggering conditions of the scene decision module 210 include a first time triggering condition and a first event triggering condition. The first time triggering condition is every a planning period. The first event triggering condition is that the arbitration result is arbitration failure.

[0040] Specifically, when the ego vehicle does not comply with the lane keeping rule, it means that the ego vehicle cannot keep the current lane to continue running: for example, ramp up and down, the road in front disappears, etc., or the ego vehicle is affected by other vehicles (such as the front vehicle being much lower than the normal driving speed, the ego vehicle cutting in with strong intention, etc.), the normal driving of the ego vehicle will be affected, and the ego vehicle must make appropriate strategies to respond. The environment frame module 100 ends, that is, enters the scene decision module 210, because it involves the acquisition of road right, so the scene decision module 210 is also a module that improves the driving experience and to a certain extent avoids risks. The first decision is a lane change scene decision, including but not limited to: lever lane change, free lane change, ramp merging and ramp exiting, and the scene decision module can be regarded as a lane change decision module.

[0041] The triggering mode of the scene decision module 210 is divided into first time triggering and first event triggering, and the scene decision module 210 is triggered when the first time triggering and / or the first event triggering are met. The first time triggering is triggered once every a planning period, and a is an integer greater than 1; the event triggering is scene decision arbitration failure. The first time triggering frequency is set to trigger once every multiple planning periods, that is, low-frequency triggering, because in high-speed scenes, high speed brings higher danger to scenes other than straight ahead; lane changing and lane borrowing in low-speed situations can be performed more easily and frequently, but at high speed, a slight action of the steering wheel can cause terrible consequences; therefore, at high speed, the stability of the scene is more important, and the switching frequency does not need to be so high. The first event triggering frequency is scene decision arbitration failure, which means that the current scene is switched to the target scene, and due to external static environment, dynamic environment, human factors, etc., the scene switching is dangerous or canceled by human, and the decision needs to be recalculated.

[0042] Preferably, the preferred value of a is 5, that is, the first time triggering is triggered once every 5 planning periods.

[0043] Among them, the event triggering mainly refers to function degradation or even function exit, which is determined by the degree of arbitration failure. Because the lane keeping decision is the most basic and relatively safest function, if the lane keeping decision arbitration fails, it means that the ego vehicle driving is threatened, at this time other functions need to be involved to ensure the safety of the ego vehicle and as far as possible normal driving. If it is a slight impact, such as close-range cutting into the ego lane and the lane keeping cannot cope with this situation, function degradation will be performed; if it is a serious impact, such as the front vehicle suddenly braking, at this time the safety function such as AEB will be triggered.

[0044] Further, the decision module 200 further comprises a lane keeping decision module 220, the lane keeping decision module 220 being configured to generate a reasonable second decision according to the first data in a case that the ego vehicle meets a lane keeping rule; wherein the second decision comprises cruise and following; the trigger condition of the lane keeping decision module 220 comprises a second time trigger condition and a second event trigger condition; the second time trigger condition is that triggering once every 1 planning period; the second event trigger condition is that triggering continuously The arbitration result is arbitration failure.

[0045] Specifically, when the ego vehicle meets the lane keeping rule, it indicates that the ego vehicle needs to keep the current lane to continue running, and the second decision is a lane keeping decision, which comprises but is not limited to cruise and following.

[0046] The scene decision module 210 is mainly related to the judgment in the multi-lane and congestion scenes. After the environment frame module 100 ends, the lane keeping decision module 220 and the scene decision module 210 are entered synchronously, in the high-speed scene, the lane keeping is the basis of all other scenes, and is the most important and most commonly used scene, which occupies the vast majority of time in normal driving, therefore, the second time trigger condition is triggering once every 1 planning period, i.e. high-frequency triggering, because of the right of way of the ego lane, it can also be regarded as a bottom-up module. The second event trigger frequency is triggering continuously The frame scene decision arbitration failure means that the current scene switches to the target scene, and the scene switching exists danger or is canceled by human being continuously for multiple times, so that the decision needs to be recalculated. When the second time trigger and / or the second event trigger are met, the lane keeping decision module is triggered.

[0047] Preferably, The preferred value of the second event trigger is 3, i.e. the second event trigger frequency is 3 frame scene decision arbitration failures continuously.

[0048] Further, the trigger condition of the scene decision arbitration module 300 comprises a third time trigger condition and a third event trigger condition; the third time trigger condition is that triggering once every planning period; the third event trigger condition is that triggering continuously The arbitration result is arbitration failure.

[0049] Specifically, after the corresponding scene decision (comprising the first decision and / or the second decision) is acquired, it is needed to arbitrate whether the scene decision is feasible. When the third time trigger and / or the third event trigger are met, the scene decision arbitration module 300 is triggered. The third time trigger condition is triggering once every 1 planning period; the third event trigger frequency is triggering continuously The frame scene decision arbitration failure means that the scene switching exists danger or is canceled by human being continuously for multiple times, so that the scene decision arbitration module 300 needs to be recalculated.

[0050] Preferably, The preferred value is 3, which means that the third event is triggered when the scene decision arbitration fails for 3 consecutive frames.

[0051] Furthermore, the scenario decision arbitration module 300 includes: a system definition unit, a scenario judgment unit, and a redundancy judgment unit. The system definition unit includes: triggering conditions, upstream information judgment, and environmental semantic judgment after preliminary analysis. The scenario judgment unit is used to make a first judgment on whether the scenario decision is feasible. The redundancy judgment unit is used to make a second judgment on whether the scenario decision is feasible. The system definition unit, scenario judgment unit, and redundancy judgment unit are performed linearly.

[0052] Specifically, scenario decision arbitration is mainly divided into three sub-units: system definition unit, scenario judgment unit, and redundancy judgment unit. These three sub-units operate linearly. The system definition unit, which defines the basic requirements, includes triggering conditions, upstream information judgment, and environmental semantic judgment after preliminary analysis. The scenario judgment unit refers to a deeper judgment of scenario decisions (mainly lane-changing scenario decisions). It requires a more in-depth and detailed analysis of the vehicle's location based on various acquired environmental information, indicating the feasibility and risk of the scenario decision, and making an initial judgment. Because the above two units may deviate from the actual scenario due to errors in external information, in-depth information mining, or limitations in the model's generalization ability, a redundant judgment unit is needed. This unit uses only mathematical methods to perform a secondary judgment on the scenario decision, ensuring the feasibility and safety of the actual implementation of the scenario decision.

[0053] Furthermore, the scene state module 400 includes a scene switching state machine 410 and a scene state machine 420. The scene switching state machine 410 is used to determine the target scene based on the scene decision when the arbitration result is that the arbitration is passed. The scene state machine 420 is used to execute the step of switching the current scene to the target scene after the target scene is determined.

[0054] Specifically, the scenario switching state machine 410 refers to the switching between scenarios (mainly the current scenario and the target scenario), such as: lane keeping → ramp merging, lane keeping → lever lane changing. Scenario switching is relatively simple; as long as the decision itself is safe, i.e., the scenario decision arbitration passes, the scenario switch is considered successful. At the same time, multiple scenario decisions may be successful; therefore, all scenario decisions need to be prioritized. When multiple scenario decisions are safe and can provide scenario switching, the target scenario is selected according to the priority of the scenario decisions themselves.

[0055] The scene switching state machine 410 passes into the internal scene state machine 420 of a single scene, the scene decision is locked until failure or completion, and then returns to the lane keeping scene through the scene switching state machine to perform the next possible scene switching action. When the target scene has been determined, switching from the current scene to the target scene is a gradual process. Preferably, taking the lane change with a dial lever as an example, there are triggering the dial lever, reminding the ego vehicle that the dial lever is about to change lanes based on regulations, continuing the original scene, starting the lane change, entering the target lane, ending the lane change, and in addition to the main branch, returning to the ego lane in case of lane change failure, corresponding secondary lane change waiting after returning to the ego lane, and returning to the ego lane after canceling the dial lever to ensure normal driving and lane change. The state process of the normal state process is added to handle additional situations.

[0056] Further, the first data includes: lane line information, obstacle information, prediction information, positioning information, map information, chassis information, other vehicle driving style and other vehicle driving intention.

[0057] Specifically, the environment frame module 100 is used to convert external data into the data type required by the decision module 200, including: lane line information, obstacle information, prediction information, positioning information, map information, chassis information, etc. At the same time, on this basis, further through the model that has been pre-trained or the module that calculates in real time, the corresponding other vehicle reference information is calculated, including other vehicle driving style, other vehicle driving intention, etc.

[0058] [Embodiment Two]

[0059] Referring to Figure 2 It is a flowchart of a high-speed automatic driving decision method based on frequency division triggering provided by the second embodiment of the application, and the high-speed automatic driving decision method is applied to the high-speed automatic driving decision system based on frequency division triggering of the first embodiment above. The high-speed automatic driving decision method comprises:

[0060] Step S10: collecting external data and converting the external data into first data;

[0061] Step S20: generating a reasonable scene decision according to the first data, wherein the scene decision includes a first decision and a second decision;

[0062] Step S30: arbitrating whether the scene decision is feasible to generate an arbitration result, and the arbitration result includes: arbitration pass and arbitration failure;

[0063] Step S40: determining a target scene according to the scene decision and the arbitration result, and switching the current scene to the target scene;

[0064] Step S50: selecting a suitable planner according to the target scene to perform trajectory planning;

[0065] When the arbitration result is arbitration pass, it indicates that the scene decision is feasible; when the arbitration result is arbitration failure, it indicates that the scene decision is infeasible.

[0066] In one specific embodiment, the high-speed automatic driving decision method in the embodiment is applied to the high-speed automatic driving decision system based on frequency division triggering in the first embodiment, and has all the beneficial effects of the high-speed automatic driving decision system based on frequency division triggering in the first embodiment, which will not be repeated here.

[0067] Further, the step S40 comprises:

[0068] In the case where the number of first decisions is multiple, the first decision with arbitration pass is defined as a feasible scene;

[0069] When the number of feasible scenes is multiple, the target scene is determined according to the priority of the feasible scene itself.

[0070] Specifically, in the case where the ego vehicle does not conform to the lane keeping rule, after the scene decision module ends generating the first decision, in the case where the number of first decisions is multiple, arbitration is made on the multiple first decisions to generate multiple arbitration results, and the arbitration result corresponds to the first decision one by one. The first decision with arbitration pass is screened out and defined as a feasible scene. In the case where the number of feasible scenes is multiple, i.e., in the same time, there may be multiple scene decisions that are established, therefore, it is necessary to prioritize all the feasible scenes with pass, and determine the target scene according to the scene priority.

[0071] In the case where the ego vehicle conforms to the lane keeping rule, after the lane keeping decision module ends generating the second decision, the second decision is directly input to the scene decision arbitration module for arbitration on the corresponding scene.

[0072]

Embodiment Three

[0073] The embodiment also provides a vehicle comprising the high-speed automatic driving decision system based on frequency division triggering in the first embodiment.

[0074] Specifically, the embodiment can achieve the technical effects corresponding to any of the technical solutions in the first embodiment, which will not be repeated here.

[0075] Finally, it should be noted that: the above embodiments are only used to illustrate the technical solutions of the present application, but not to limit them; although the present application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that: it can still modify the technical solutions recorded in the foregoing embodiments, or make equivalent replacement to part of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A high-speed autonomous driving decision system based on frequency division trigger, characterized in that, The high-speed automatic driving decision system comprises: an environment frame module (100) configured to process collected external data and convert the external data into first data; a decision module (200) configured to generate a reasonable scene decision according to the first data; a scene decision arbitration module (300) configured to arbitrate whether the scene decision is feasible and generate an arbitration result, wherein the arbitration result comprises arbitration pass and arbitration failure; a scene state module (400) configured to determine a target scene according to the scene decision and the arbitration result and switch a current scene to the target scene; a scene corresponding planner (500) configured to select a suitable planner for trajectory planning according to the target scene; wherein trigger conditions of the decision module (200), the scene decision arbitration module (300) and the scene corresponding planner (500) are different, and when the trigger conditions are met, the corresponding modules are triggered; the trigger conditions of the decision module (200) and the scene decision arbitration module (300) are measured by a planning period and a frame number of arbitration failure; and the trigger condition of the scene corresponding planner (500) is every 1 planning period.

2. The high-speed autonomous driving decision system of claim 1, wherein, The decision module (200) comprises: a scene decision module (210) configured to generate a reasonable first decision according to the first data when a vehicle does not meet a lane keeping rule; wherein the lane keeping rule is a rule under which the vehicle can continue to perform lane keeping; the first decision comprises a lever lane change, a free lane change, ramp merging and ramp exiting; and trigger conditions of the scene decision module (210) are divided into a first time trigger condition and a first event trigger condition; the first time trigger condition is every a planning period; and the first event trigger condition is that the arbitration result is arbitration failure.

3. The high-speed autonomous driving decision system of claim 2, wherein, The decision module (200) further comprises: a lane keeping decision module (220) configured to generate a reasonable second decision according to the first data when the vehicle meets the lane keeping rule; The second decision includes: cruising and following a vehicle; the trigger condition of the lane keeping decision module (220) is divided into a second time trigger condition and a second event trigger condition; the second time trigger condition is: every 1 planning period; the second event trigger condition is: consecutive The arbitration result is arbitration failure.

4. The high-speed automatic driving decision system according to claim 1, wherein trigger conditions of the scene decision arbitration module (300) are divided into a third time trigger condition and a third event trigger condition; the third time trigger condition is every planning period; The third event trigger condition is that consecutive The arbitration result is arbitration failure.

5. The high-speed autonomous driving decision system of claim 4, wherein, the scene decision arbitration module (300) comprises: a system definition unit comprising trigger conditions, upstream information judgment, environment semantic judgment after preliminary analysis and the like; a scene judgment unit configured to make a first judgment on whether the scene decision is feasible; a redundancy judgment unit configured to make a second judgment on whether the scene decision is feasible; The system definition unit, the scene judgment unit, and the redundancy judgment unit are linear.

6. The high-speed autonomous driving decision system of claim 1, wherein, The scene state module (400) includes: A scene switching state machine (410) configured to determine the target scene according to the scene decision when the arbitration result is arbitration pass; A scene state machine (420) configured to perform a step of switching the current scene to the target scene after determining the target scene.

7. The high-speed automatic driving decision system of claim 1, wherein: The first data includes lane line information, obstacle information, prediction information, positioning information, map information, chassis-by-wire information, other vehicle driving style, and other vehicle driving intention.

8. A high-speed autonomous driving decision method based on frequency division trigger, the high-speed autonomous driving decision method is applied to the high-speed autonomous driving decision system based on frequency division trigger as claimed in any one of claims 1-7, characterized in that, The high-speed automatic driving decision method includes: Collecting external data and converting the external data into first data; Generating a reasonable scene decision according to the first data, wherein the scene decision includes a first decision and a second decision; Arbitrating whether the scene decision is feasible to generate an arbitration result, wherein the arbitration result includes arbitration pass and arbitration failure; Determining a target scene according to the scene decision and the arbitration result, and switching a current scene to the target scene; Selecting a suitable planner according to the target scene to perform trajectory planning; When the arbitration result is arbitration pass, the scene decision is feasible; and when the arbitration result is arbitration failure, the scene decision is not feasible.

9. The high-speed automatic driving decision method of claim 8, wherein: Determining a target scene according to the scene decision and the arbitration result, and switching a current scene to the target scene includes: When the number of the first decisions is multiple, defining the first decision with arbitration pass as a feasible scene; When the number of the feasible scenes is multiple, determining the target scene according to the priority of the feasible scenes themselves.

10. A vehicle characterized by comprising: A high-speed automatic driving decision system based on frequency division triggering. Any one of claims 1-7.

Images