An automatic driving method, device, intelligent driving automobile and driving control system

Abstract

The application discloses an automatic driving method and device, a smart driving automobile and a driving control system, comprising: acquiring real-time driving data of the automobile and collecting real-time environment data around the automobile; wherein the real-time driving data comprises a position of a current section where the automobile is located; the real-time environment data comprises real-time motion behavior data of a target group in a motion state and traffic indication data; downloading a basic driving decision corresponding to the current section from a cloud server; wherein the basic driving decision is obtained by learning historical driving decisions of all vehicles passing through the current section by the cloud server; predicting an expected behavior of the target group in the motion state according to the real-time motion behavior data of the target group in the motion state; predicting an expected driving behavior of the automobile according to the expected behavior of the target group in the motion state, the traffic indication data and the real-time driving data; coupling the expected driving behavior of the automobile with the basic driving decision to obtain a final driving decision for controlling driving, and automatically driving based on the final driving decision.

Description

Technical Field

[0001] This application relates to the field of intelligent driving technology, and in particular to an autonomous driving method, device, intelligent driving vehicle and driving control system. Background Technology

[0002] With the development of intelligent and electric vehicles, autonomous driving has become a standard feature. Autonomous driving collects data about the surrounding environment through various sensors, analyzes and processes this data using complex algorithms, and then automatically controls the vehicle's power, steering, and braking systems. This enables the vehicle to make autonomous decisions about its driving behavior, laying a solid foundation for the construction of future intelligent transportation systems.

[0003] However, traditional autonomous driving relies mainly on sensor data inside the car and environmental perception of the actual vehicle condition to make driving decisions, but it cannot predict expected changes in the vehicle condition. Therefore, it has limitations in dealing with complex and ever-changing traffic scenarios, resulting in low stability of autonomous driving and even traffic accidents. Summary of the Invention

[0004] To address or partially address the limitations of existing autonomous driving technologies in complex and ever-changing traffic scenarios, this invention provides an autonomous driving method, device, intelligent driving vehicle, and driving control system. By referencing the actual traffic environment to predict the vehicle's expected driving behavior and combining it with basic driving decisions from a cloud server, the resulting driving strategy can effectively cope with real-time and ever-changing complex traffic environments, enabling high-precision intelligent driving control of the vehicle and enhancing the stability of autonomous driving.

[0005] To address the aforementioned technical problems, a first aspect of the present invention discloses an autonomous driving method, the method comprising:

[0006] The system acquires real-time driving data of the vehicle and collects real-time environmental data around the vehicle. The real-time driving data includes the vehicle's current location on the road segment. The real-time environmental data includes real-time motion behavior data of a target group in motion and traffic indication data.

[0007] Download the basic driving decisions corresponding to the current road segment from the cloud server; wherein, the basic driving decisions are obtained by the cloud server by learning the historical driving decisions made by all vehicles when passing through the current road segment;

[0008] The expected behavior of the target group in motion is predicted based on the real-time motion behavior data of the target group in motion.

[0009] predicting a target group in motion behavior according to the real-time motion behavior data of the target group in motion;

[0010] coupling the basic driving decision with the target group in motion behavior to obtain a final driving decision for controlling driving, and automatically driving based on the final driving decision.

[0011] Optionally, the target group in motion is a pedestrian or a non-motor vehicle, and the target group in motion behavior is predicted according to the real-time motion behavior data of the pedestrian or the non-motor vehicle, specifically including:

[0012] analyzing the real-time motion behavior data of the pedestrian or the non-motor vehicle to obtain a current behavior intention of the pedestrian or the non-motor vehicle;

[0013] searching for a historical behavior intention of the pedestrian or the non-motor vehicle in a current road section;

[0014] coupling the historical behavior intention with the current behavior intention to obtain the target group in motion behavior and an execution probability thereof.

[0015] Optionally, the target group in motion is a car located around the vehicle; and the target group in motion behavior is predicted according to the real-time motion behavior data of the car, specifically including:

[0016] extracting car navigation key data from the real-time motion behavior data of the car, wherein the car navigation key data at least includes a navigation destination, a travel purpose, a lane where the car is located, and a driver gender;

[0017] predicting a plurality of car driving behavior intentions according to the lane where the car is located;

[0018] predicting execution probabilities of the plurality of car driving behavior intentions with reference to the navigation destination, the travel purpose, and the driver gender.

[0019] Optionally, the target group in motion behavior is predicted according to the real-time motion behavior data of the target group in motion, the traffic indication data, and the real-time driving data, specifically including:

[0020] predicting the target group in motion behavior according to the target group in motion behavior and the execution probability thereof, the plurality of car driving behavior intentions and the execution probabilities thereof, the traffic indication data, and the real-time driving data.

[0021] Optionally, the expected driving behavior of the host vehicle is predicted according to the expected behavior of the pedestrian or non-motor vehicle and its execution probability, the execution probability of the several automobile driving behavior intentions, the traffic indication data and the real-time driving data, and specifically includes:

[0022] The lane in which the host vehicle is located is extracted from the real-time driving data, and the several host vehicle driving behavior intentions are predicted according to the lane in which the host vehicle is located.

[0023] A first influence coefficient on the driving of the host vehicle is determined according to the expected behavior of the pedestrian or non-motor vehicle and its execution probability.

[0024] A second influence coefficient on the driving of the host vehicle is determined according to the several automobile driving behavior intentions and their execution probabilities.

[0025] The third influence coefficient on the driving of the host vehicle is determined by conditionally coupling the driving data of the host vehicle with reference to the traffic indication data and / or the automobile driving data.

[0026] The target execution probability of each of the several host vehicle driving behavior intentions is determined according to the first influence factor, the second influence coefficient and the third influence coefficient.

[0027] Optionally, the expected driving behavior of the host vehicle is coupled with the basic driving decision to obtain a final driving decision for controlling driving, and specifically includes:

[0028] The basic execution probability of each of the several host vehicle driving behavior intentions is extracted from the basic driving decision.

[0029] The basic execution probability and the target execution probability of each of the several host vehicle driving behavior intentions are fused to obtain the final execution probability of each of the several host vehicle driving behavior intentions.

[0030] The behavior intention with the highest probability is selected from the final execution probability of each of the several host vehicle driving behavior intentions as the final driving decision.

[0031] Optionally, after the automatic driving based on the final driving decision, the method further includes:

[0032] The driving behavior corresponding to the final driving decision is verified, and a punishment and reward value of a punishment and reward mechanism is given to the final driving decision according to the verification result.

[0033] When the punishment and reward value of the final driving decision reaches a set threshold, the final driving decision is uploaded to the cloud server for learning to update the basic driving decision.

[0034] The second aspect of the present application discloses an automatic driving device, comprising:

[0035] An acquisition module is configured to acquire real-time driving data of the vehicle and collect real-time environment data around the vehicle; wherein the real-time driving data comprises a position of a current road segment where the vehicle is located; the real-time environment data comprises real-time motion behavior data of a target group in a motion state and traffic indication data;

[0036] A communication module is configured to download a basic driving decision corresponding to the current road segment from a cloud server; wherein the basic driving decision is obtained by learning historical driving decisions of all vehicles passing through the current road segment by the cloud server;

[0037] A first prediction module is configured to predict an expected behavior of the target group in the motion state according to the real-time motion behavior data of the target group in the motion state;

[0038] A second prediction module is configured to predict an expected driving behavior of the vehicle according to the expected behavior of the target group in the motion state, the traffic indication data and the real-time driving data;

[0039] A coupling module is configured to couple the expected driving behavior of the vehicle with the basic driving decision to obtain a final driving decision for controlling driving, and automatically drive based on the final driving decision.

[0040] Optionally, the target group in the motion state is a pedestrian or a non-motor vehicle, and the first prediction module is specifically configured to:

[0041] analyze the real-time motion behavior data of the pedestrian or the non-motor vehicle to obtain a current behavior intention of the pedestrian or the non-motor vehicle;

[0042] find a historical behavior intention of the pedestrian or the non-motor vehicle in the current road segment;

[0043] couple the current behavior intention with the historical behavior intention to obtain an expected behavior of the pedestrian or the non-motor vehicle and an execution probability thereof.

[0044] Optionally, the target group in the motion state is a car located around the vehicle; and the first prediction module is specifically configured to:

[0045] extract car navigation key data from the real-time motion behavior data of the car; wherein the car navigation key data at least comprises a navigation destination, a travel purpose, a lane where the car is located, and a driver gender;

[0046] predict a plurality of car driving behavior intentions according to the lane where the car is located;

[0047] Reference the navigation destination, travel purpose, driver gender, and predict execution probability of the several automobile driving behavior intentions.

[0048] Optionally, the second prediction module is specifically configured to predict the expected driving behavior of the ego vehicle according to the expected behavior of the pedestrian or non-motor vehicle and the execution probability thereof, the several automobile driving behavior intentions and the execution probability thereof, the traffic indication data, and the real-time driving data.

[0049] Optionally, the second prediction module is specifically configured to:

[0050] extract the lane in which the ego vehicle is located from the real-time driving data, and predict several ego vehicle driving behavior intentions according to the lane in which the ego vehicle is located;

[0051] determine a first influence coefficient on ego vehicle driving according to the expected behavior of the pedestrian or non-motor vehicle and the execution probability thereof;

[0052] determine a second influence coefficient on ego vehicle driving according to the several automobile driving behavior intentions and the execution probability thereof;

[0053] reference the traffic indication data and / or the automobile driving data to conditionally couple the ego vehicle driving data, and determine a third influence coefficient on ego vehicle driving;

[0054] determine a target execution probability of each of the several ego vehicle driving behavior intentions according to the first influence factor, the second influence coefficient, and the third influence coefficient.

[0055] Optionally, the coupling module is specifically configured to:

[0056] extract a basic execution probability of each of the several ego vehicle driving behavior intentions from the basic driving decision;

[0057] fuse the basic execution probability of each of the several ego vehicle driving behavior intentions and the target execution probability thereof to obtain a final execution probability of each of the several ego vehicle driving behavior intentions;

[0058] select a behavior intention with the highest probability from the final execution probability of each of the several ego vehicle driving behavior intentions as the final driving decision.

[0059] Optionally, the apparatus further comprises a feedback module configured to:

[0060] verify the driving behavior corresponding to the final driving decision, and assign a punishment or reward value of a punishment or reward mechanism to the final driving decision according to a verification result;

[0061] When the reward value of the final driving decision reaches a set threshold, the final driving decision is uploaded to the cloud server for learning to update the basic driving decision.

[0062] In a third aspect, the application discloses an intelligent driving automobile, which comprises a memory, an intelligent driving domain controller, and a computer program stored in the memory and capable of running on the intelligent driving domain controller, and the intelligent driving domain controller implements the steps of the method in the first aspect when executing the program.

[0063] In a fourth aspect, the application discloses a driving control system, which comprises a cloud server and an intelligent driving automobile as described in the third aspect.

[0064] The application has the following advantages or benefits:

[0065] The application provides an automatic driving method and device, an intelligent driving automobile and a driving control system, which can predict the expected driving behaviors of pedestrians, non-motor vehicles and cars around according to real-time environmental data, and obtain the expected driving behavior of the car by combining real-time driving data, so that the expected driving behavior of the car can respond to the actual traffic environment. In addition, the basic driving decision is downloaded from the cloud server by combining the intelligent driving domain controller and the cloud server, and the basic driving decision is obtained by learning the historical driving decisions of all vehicles on the current road section by the cloud server, so that the basic driving decision can guarantee the reference accuracy of the current road section. The basic driving decision is coupled with the expected driving behavior of the car, and the final driving strategy can well respond to the real-time and complex traffic environment, and can accurately control the intelligent driving of the car, thereby enhancing the stability of automatic driving.

[0066] The above description is only a summary of the technical solutions of the application, and the technical solutions can be implemented according to the content of the specification, and in order to make the above and other purposes, characteristics and advantages of the application more obvious and easy to understand, the following specific embodiments of the application are described. BRIEF DESCRIPTION OF DRAWINGS

[0067] Various other advantages and benefits will become apparent to those of ordinary skill in the art upon reading the following detailed description of the preferred embodiments. The accompanying drawings are included to provide a description of the preferred embodiments and are not meant to limit the present application. Furthermore, the same reference numerals are used throughout the several drawings to denote the same or similar parts. In the drawings:

[0068] Figure 1 Fig. 1 shows a structural schematic diagram of a driving control system according to an embodiment of the application;

[0069] Figure 2A flowchart of an automatic driving method according to an embodiment of the present application is shown.

[0070] Figure 3 A flowchart of predicting an expected driving behavior of the vehicle according to an embodiment of the present application is shown.

[0071] Figure 4 An interaction logic diagram of the vehicle and the cloud server according to an embodiment of the present application is shown.

[0072] Figure 5 A schematic diagram of the intelligent driving domain controller according to an embodiment of the present application is shown. DETAILED DESCRIPTION

[0073] Exemplary embodiments of the present application will be described hereinafter with reference to the accompanying drawings. While exemplary embodiments of the present application are shown in the drawings, it is understood that the present application can be embodied in various forms and should not be limited by the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the present application to those skilled in the art.

[0074] Exemplary embodiments of the present application will be described hereinafter with reference to the accompanying drawings. While exemplary embodiments of the present application are shown in the drawings, it is understood that the present application can be embodied in various forms and should not be limited by the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the present application to those skilled in the art.

[0075] To solve the technical problem that the automatic driving in the prior art has limitations when facing complex and variable traffic scenes, the present application provides an automatic driving method, device, intelligent driving vehicle and driving control system.

[0076] In a first aspect, in order to facilitate the description and understanding of the present application, the driving control system of the present application is introduced as follows. As shown in Figure 1 The driving control system includes a cloud server and an intelligent driving vehicle.

[0077] The cloud server is loaded with a cloud AI (Artificial Intelligence) large model.

[0078] The intelligent driving vehicle includes a collector, a TBOX (Telematics Box, vehicle communication terminal) and an intelligent driving domain controller. The intelligent driving vehicle refers to all vehicles with automatic driving function, and the number thereof is not limited.

[0079] The intelligent driving vehicle collects road and vehicle condition information of a certain intersection by using the collector (such as sensors, cameras, radars, etc.).

[0080] The smart driving car communicates with the cloud server through the TBOX. Specifically, the smart driving car uploads the road condition and vehicle condition information of a certain intersection and the driving decisions that have been made at the intersection to the cloud server through the TBOX for training and learning of the cloud AI large model, to obtain the basic driving decisions about the intersection.

[0081] The cloud server issues the basic driving decisions to the smart driving car through the TBOX for correcting the smart driving decisions of the smart driving car. If the smart driving car is facing a new road, it first drives according to the basic driving decisions issued by the cloud server and makes corrections in combination with the actual situation.

[0082] Inside the smart car, a local AI large model is loaded to predict and analyze the road condition and vehicle condition information collected at the intersection to predict the expected behaviors of the pedestrians, non-motor vehicles and cars around the car, and further predict the smart driving behavior of the car. In addition, the road condition and vehicle condition information collected this time, the expected behaviors of the pedestrians, non-motor vehicles and cars around the car, and the smart driving behavior of the car are transmitted to the cloud server, which is trained and learned by the cloud AI large model in the cloud server. The cloud AI large model contains a plurality of behavior prediction models, such as a pedestrian behavior prediction model and a car behavior prediction model, which can learn more accurate and universal basic driving decisions to improve the accuracy of the traffic prediction at the intersection and optimize the smart driving control of each road segment. The control logic of the car smart driving is described below in the second aspect of the disclosure, which will not be described here.

[0083] In a second aspect, an embodiment of the present application provides an automatic driving method, which is mainly applied to a smart driving domain controller.

[0084] Referring to Figure 2 The method at least includes the following steps:

[0085] S201, acquiring real-time driving data of the car and collecting real-time environmental data around the car.

[0086] The real-time driving data of the car includes but is not limited to: the current speed of the car, the current time, the position of the current road segment where the car is located, the lane where the car is located, the destination of the car, the working time, the type of the destination of the car (such as company, home, shopping mall, hospital, etc.), and the remaining distance to the destination of the car. The real-time driving data of the car is collected by various sensors loaded on the car.

[0087] The real-time environmental data includes real-time motion behavior data of a target group in a motion state and traffic indication data. The target group in a motion state is a pedestrian, a non-motor vehicle, or a car around the car.

[0088] For example, the real-time motion behavior data of a pedestrian includes: static data of the pedestrian (appearance, gender, clothing, company location, distance from the company, etc.), dynamic data of the pedestrian (walking posture, walking speed, walking direction, etc.).

[0089] For example, the real-time motion behavior data of a non-motor vehicle includes: static data of the non-motor vehicle (appearance, gender, clothing, license plate number, company location, distance from the company, etc.), dynamic data of the non-motor vehicle (driving speed, driving direction, etc.).

[0090] Of course, the real-time motion behavior data of all pedestrians around the vehicle (especially in front of the vehicle) and the real-time motion behavior data of non-motor vehicles need to be collected. Among them, the real-time motion behavior data of the target group in motion state is collected by the collection devices such as cameras, radars, communication terminals, etc. loaded on the vehicle. For example, the vehicle uses the camera to collect data such as appearance, gender, clothing of the pedestrian, and finds the corresponding company location, distance from the company, etc. of the pedestrian from the local database or uploaded network. For another example, the vehicle uses the radar to collect and analyze the dynamic data (driving speed, direction, etc.) of the non-motor vehicle.

[0091] The real-time motion behavior data of the vehicle at least includes vehicle navigation key data and vehicle condition data. Among them, the vehicle navigation key data at least includes: navigation destination, travel purpose (such as company, home, shopping mall, hospital, etc.), vehicle purpose (work, medical treatment, will, online car, travel, home, etc.), gender of the vehicle driver, and vehicle lane. The vehicle condition data at least includes: current vehicle speed. Of course, the real-time motion behavior data of all vehicles around the vehicle needs to be collected. Among them, the real-time motion behavior data of the vehicle is collected by the collection devices such as cameras, radars, communication terminals, etc. loaded on the vehicle. For example, the current destination type, purpose, etc. of the vehicle are remotely obtained by using the vehicle-mounted communication terminal.

[0092] The traffic indication data at least includes: current road speed limit, distance from the vehicle to the traffic light intersection, and traffic light intersection switching time. Among them, the traffic indication data is collected by the collection devices such as cameras, radars, communication terminals, etc. loaded on the vehicle.

[0093] S202, downloading the basic driving decision corresponding to the current road section from the cloud server.

[0094] Among them, the basic driving decision is obtained by the cloud server learning the historical driving decisions of the vehicle or all vehicles including the vehicle passing through the current road section.

[0095] Specifically, the cloud server collects historical environment information and historical driving decisions of the vehicle on the current road segment to learn and obtain a basic driving decision for the current road segment. For example, the current road segment is a road segment that the vehicle must pass to go to work, and the work time and route are relatively fixed. The cloud server learns from the historical environment information and historical driving decisions of the vehicle to obtain a basic driving decision that is more suitable for the driving habits of the vehicle. In order to make the basic driving decision more widely applicable, the cloud server learns from the historical environment information and historical driving decisions of all vehicles (including the vehicle) to obtain a basic driving decision that is more adaptive.

[0096] For the vehicle, the TBOX uploads the location of the current road segment where the vehicle is located to the cloud server, and downloads the corresponding basic driving decision from the cloud server. Of course, the vehicle lane can also be uploaded to make the downloaded basic driving decision more accurate for the intelligent driving of the vehicle lane. The basic driving decision includes the basic execution probability of several driving behavior intentions of the vehicle lane. Taking a 3-lane as an example, assuming that the vehicle is in the middle lane, the driving behavior intentions of the vehicle include straight driving at constant speed, straight driving with acceleration, straight driving with deceleration, left lane change, and right lane change. The basic driving decision includes the corresponding basic execution probability. For example, straight driving at constant speed is 0.5, straight driving with acceleration is 0.2, straight driving with deceleration is 0.2, left lane change is 0.1, and right lane change is 0.1. The basic driving decision changes according to the change of the location.

[0097] S203, predicting the expected behavior of the target group in the motion state according to the real-time motion behavior data of the target group in the motion state.

[0098] In the prediction process of the expected behavior of the target group in the motion state, different target groups have different analysis strategies.

[0099] If the target group in motion is a pedestrian or a non-motor vehicle, the real-time motion behavior data of the pedestrian or non-motor vehicle is analyzed to obtain the current behavior intention of the pedestrian or non-motor vehicle. Specifically, the real-time motion behavior data of the target group in motion includes static data and dynamic data. The current behavior intention is obtained by combining the static data and dynamic data analysis. Taking pedestrian A as an example, the current behavior intention and its execution probability are obtained by analyzing the pedestrian's appearance, gender, clothing, company location, distance from the company, walking posture, walking speed, and walking direction. In this embodiment, a local AI large model built-in the vehicle can be used for behavior prediction, or a mapping relationship obtained through long-term learning can be used for behavior prediction. For example, the current behavior intention of pedestrian A is to cross the road in front, and its execution probability is different according to the purpose of pedestrian A. Assuming that the purpose of crossing the road in front is work, the execution probability is A1; assuming that the purpose of crossing the road in front is work, the execution probability is A2; assuming that the purpose of crossing the road in front is work, the execution probability is A3; the first occurrence (unknown): the execution probability is A4.

[0100] Further, the historical behavior intention of the pedestrian or non-motor vehicle on the current road segment is found. For example, the historical behavior intention of pedestrian A on the current road segment is to run a red light, and the execution probability is A5. The current behavior intention is coupled with the historical behavior intention to obtain the expected behavior of the pedestrian or non-motor vehicle and its execution probability. In the coupling process, the average value of the historical behavior intention and the current behavior intention is calculated to obtain the expected behavior of the target group in motion and its execution probability. For example, the current behavior intention of the pedestrian is to cross the road in front, and the execution probability is A3, then the execution probability of the expected behavior of the pedestrian is: A = (A1 + A3) / 2.

[0101] Table 1 is the expected behavior of the target group in motion (pedestrian, non-motor vehicle) and its execution probability.

[0102] Table 1

[0103] Expected behavior Probability of execution Pedestrian A Crosses road ahead 65％ Pedestrian B Crosses road ahead 55％ Non-motorized vehicle C Crosses road ahead 55％ Non-motorized vehicle D Crosses road ahead 50％

[0104] In Table 1, the expected behavior of pedestrian A, pedestrian B, non-motor vehicle C, and non-motor vehicle D will affect the driving of the vehicle, and if the pedestrian is straight along the road and does not affect the driving of the vehicle, it can be ignored. Further, it is judged whether the execution probability of the expected behavior of the pedestrian or non-motor vehicle is greater than a threshold value, if greater than the threshold value, such as greater than 50%, it will affect the driving of the vehicle and be included in the calculation of the influence coefficient. If it is less than the threshold value, it is not considered.

[0105] If the target group in motion is a car located around the host vehicle, the car navigation key data is extracted from the real-time motion behavior data of the car. Among them, the car navigation key data at least includes: navigation destination, travel purpose, car lane, driver gender. According to the car lane, the driving behavior intention of several cars is predicted. In this embodiment, the built-in local AI large model in the car can be used for behavior prediction, or the mapping relationship obtained by long-term learning can be used for behavior prediction. Among them, the driving behavior intention of several cars specifically includes: straight acceleration, straight deceleration, straight uniform speed, left lane change, right lane change. Referring to the navigation destination, the travel purpose, the driver gender, the execution probability of the driving behavior intention of several cars is predicted. For example, the straight acceleration probability is E1, the straight deceleration probability is E2, the straight uniform speed probability is E3, the left lane change probability is E4, and the right lane change probability is E5. Among them, all the cars around the host vehicle refer to the above execution logic, for example, the car in front of the host vehicle, the car behind the host vehicle, the car in front of the left of the host vehicle, the car in front of the right of the host vehicle, the car behind the left of the host vehicle, and the car behind the right of the host vehicle, all get their expected behavior and execution probability.

[0106] S204, according to the expected behavior of the target group in motion, the traffic indication data and the real-time driving data, the expected driving behavior of the host vehicle is predicted.

[0107] Among them, according to the expected behavior of the pedestrian or non-motor vehicle and its execution probability, the driving behavior intention of several cars and its execution probability, the traffic indication data and the real-time driving data, the expected driving behavior of the host vehicle is predicted.

[0108] Referring to Figure 3 , is a flowchart for predicting the expected driving behavior of the host vehicle, which specifically performs the following steps:

[0109] S2041, the lane where the host vehicle is located is extracted from the real-time driving data, and the driving behavior intention of several host vehicles is predicted according to the lane where the host vehicle is located.

[0110] Specifically, the number and type of driving behavior intention corresponding to the non-passable lane are different. For example, if the host vehicle is in the middle lane, its driving behavior intention includes: straight uniform speed, straight acceleration, straight deceleration, left lane change, right lane change.

[0111] S2042, the first influence coefficient of the host vehicle driving is determined according to the expected behavior of the pedestrian or non-motor vehicle and its execution probability.

[0112] Among them, each host vehicle driving behavior intention has its own first influence coefficient.

[0113] Specifically, in the process of determining the first influence coefficient, the expected behavior of the pedestrian or non-motor vehicle is determined to have a first influence relationship on each driving behavior intention of the ego vehicle; wherein the first influence relationship is a positive influence relationship or a negative influence relationship. The expected behavior of the pedestrian or non-motor vehicle has different first influence relationships on different driving behavior intentions of the ego vehicle. According to the execution probability of the expected behavior of the pedestrian or non-motor vehicle, a first influence degree value is determined for each driving behavior intention of the ego vehicle. Wherein the execution probability of the expected behavior of the pedestrian or non-motor vehicle and the first influence degree value of the driving behavior intention of the ego vehicle have a mapping relationship, and the first influence degree value can be obtained according to the mapping relationship. According to the first influence relationship and the first influence degree value of each driving behavior intention of the ego vehicle, a first influence coefficient of each driving behavior intention of the ego vehicle is obtained.

[0114] Taking the expected behavior of pedestrian A: crossing the road in front as an example, if the driving intention of the ego vehicle is to straighten and accelerate, the expected behavior of pedestrian A will have a negative influence relationship on the straightening and accelerating operation of the ego vehicle, and the first influence coefficient of the ego vehicle will be negative. The influence degree is reflected in the specific value of the first influence coefficient. If the driving behavior intention of the ego vehicle is to straighten and decelerate, the expected behavior of pedestrian A will have a positive influence relationship on the straightening and decelerating operation of the ego vehicle, and the first influence coefficient of the ego vehicle will be positive. If the driving intention of the ego vehicle is to change lanes to the left, the expected behavior of pedestrian A will not have an influence on the ego vehicle changing lanes to the left, and the first influence coefficient of the ego vehicle will be 0.

[0115] Referring to Table 2 below, the first influence coefficient of the target group (taking pedestrians and non-motor vehicles as examples) in motion on the driving of the ego vehicle is determined.

[0116] Table 2

[0117] Vehicle straight ahead accelerating Vehicle straight ahead constant speed Vehicle straight ahead decelerating Vehicle left lane change Vehicle right lane change Pedestrian A -0.2 0.1 0.1 0 0 Pedestrian B -0.1 0.1 0.1 0 0 Non-motorized vehicle C -0.3 0.1 0.1 0 0 Non-motorized vehicle D -0.2 0.1 0.1 0 0

[0118] S2043, according to the several automobile driving behavior intentions and their execution probabilities, a second influence coefficient of each on the driving of the ego vehicle is determined.

[0119] Wherein each driving behavior intention of the ego vehicle has a respective second influence coefficient.

[0120] The second influence coefficient is determined by the orientation of the ego vehicle, the several automobile driving behavior intentions and their execution probabilities corresponding to the ego vehicle, and the several driving behavior intentions of the ego vehicle.

[0121] Specifically, in the process of determining the second influence coefficient, according to the position of the car relative to the ego car and each car driving behavior intention, the second influence relationship of each car driving behavior intention to each ego car driving behavior intention is determined; wherein different positions of the car and different car driving behavior intentions will have different second influence relationships to each ego car driving behavior intention, and the second influence relationship includes positive influence relationship and negative influence relationship. Different each car driving behavior intention has different second influence relationship to different ego car driving behavior intention. According to the execution probability of each car driving behavior intention, the second influence degree value to each ego car driving behavior intention is determined. Wherein the execution probability of each car driving behavior intention and the second influence degree value of the ego car driving behavior intention have a mapping relationship, and the second influence degree value can be obtained according to the mapping relationship. According to the second influence relationship and the second influence degree value of each ego car driving behavior intention, the second influence coefficient of each ego car driving behavior intention is obtained.

[0122] Taking the car in front of the ego car as an example, if the driving behavior intention of the ego car is straight acceleration, the driving behavior intentions of the car in front of the ego car, straight acceleration, left lane change and right lane change, have positive influence relationship to the straight acceleration of the ego car, and the second influence coefficient of the ego car driving to the car in front of the ego car is positive, and the influence degree is reflected in the specific value of the second influence coefficient. The driving behavior intention of the car in front of the ego car, straight deceleration, has negative influence relationship to the straight acceleration of the ego car, and the second influence coefficient of the ego car driving to the car in front of the ego car is negative, and the influence degree is reflected in the specific value of the second influence coefficient. The driving behavior intention of the car in front of the ego car, straight constant speed, has no influence to the straight acceleration of the ego car, and the second influence coefficient of the ego car driving to the car in front of the ego car is 0. For another example, if the driving behavior intention of the ego car is left lane change, the driving behavior intentions of the car in front of the ego car, straight acceleration and right lane change, have positive influence relationship to the left lane change of the ego car, and the second influence coefficient of the ego car driving to the car in front of the ego car is positive, and the influence degree is reflected in the specific value of the second influence coefficient. The driving behavior intentions of the car in front of the ego car, straight deceleration and left lane change, have negative influence relationship to the left lane change of the ego car, and the second influence coefficient of the ego car driving to the car in front of the ego car is negative, and the influence degree is reflected in the specific value of the second influence coefficient. The driving behavior intention of the car in front of the ego car, straight constant speed, has no influence to the left lane change of the ego car, and the second influence coefficient of the ego car driving to the car in front of the ego car is 0.

[0123] The following refers to Table 3, which is a specific example of the numerical value of the second influence coefficient of different car driving behavior intentions to the ego car driving.

[0124] Table 3

[0125]

[0126]

[0127] S2044, it is worth noting that the size of the second influence coefficient is also affected by the size of the execution probability corresponding to the driving behavior intention of the car.

[0128] The driving data of the car is conditionally coupled with reference to the traffic indication data and / or the car driving data, and a third influence coefficient affecting the driving of the car is determined.

[0129] The traffic indication data includes at least the current road speed limit, the distance from the car to the traffic light intersection, and the traffic light intersection switching time. The driving data of the car includes but is not limited to the current speed of the car, the current time, the lane where the car is located, the destination of the car, the type of the destination of the car (such as company, home, shopping mall, hospital, etc.), the working time, and the remaining distance to the destination of the car. The car driving data includes but is not limited to the current speed of the car.

[0130] Based on the foregoing disclosed data, at least the following conditional coupling is performed:

[0131] (1) According to the minimum speed limit of the car and the current speed of the car, the third influence coefficient is determined according to the coupling result. The minimum speed limit of the car is obtained by dividing the remaining distance to the destination of the car by the time difference between the current distance and the working time.

[0132] (2) According to the minimum speed limit of the car and the current road speed limit, the third influence coefficient is determined according to the coupling result.

[0133] (3) According to the time when the car arrives at the traffic light intersection and the traffic light intersection switching time, the third influence coefficient is determined according to the coupling result. The time when the car arrives at the traffic light intersection is obtained by dividing the distance from the car to the traffic light intersection by the current speed of the car.

[0134] (4) According to whether there is a car around the car, the fourth influence coefficient is determined according to the coupling result.

[0135] (5) According to the current speed of the car, the current speed of the car is conditionally coupled, and / or according to the first ratio of the current speed difference between the car and the surrounding cars and the traffic light intersection switching time, and the second ratio of the current speed of the car and the traffic light intersection switching time, the fourth influence coefficient is determined by combining the two coupling results.

[0136] Different driving behavior intentions of the car have different feedbacks on the coupling result, which is reflected in the third influence coefficient.

[0137] S2045, according to the first influence factor, the second influence coefficient, the third influence coefficient, the target execution probability of each driving behavior intention of the car is determined.

[0138] Specifically, for each vehicle driving behavior intention, the first influence factor, the second influence coefficient and the third influence coefficient are added, and the corresponding target execution probability is obtained.

[0139] It is worth noting that since the vehicle driving data and the real-time environment data around the vehicle change at any time, the target execution probability of each vehicle driving behavior intention can be calculated every set time (for example, 10 ms).

[0140] S205, coupling the expected driving behavior of the vehicle with the basic driving decision to obtain the final driving decision for controlling driving, and automatically driving based on the final driving decision.

[0141] In the coupling process, the basic execution probability of each vehicle driving behavior intention is extracted from the basic driving decision; the basic execution probability and the target execution probability of each vehicle driving behavior intention are fused to obtain the final execution probability of each vehicle driving behavior intention. Among them, the basic execution probability and the target execution probability of each behavior intention can be fused by adding to obtain the final execution probability of each vehicle driving behavior intention. The behavior intention with the highest probability is selected from the final execution probability of each vehicle driving behavior intention as the final driving decision.

[0142] It is worth noting that since the target execution probability is calculated every set time (for example, 10 ms), the basic execution probability and the target execution probability of each vehicle driving behavior intention are fused every set time (for example, 10 ms) to obtain the final execution probability of each vehicle driving behavior intention. Further, the behavior intention meeting the set number and the set probability threshold is selected from the final execution probability as the final driving decision. For example, the final execution probability is calculated every 10 ms. If the final execution probability of straight acceleration reaches 0.5 or more for 100 consecutive times, straight acceleration is selected as the final execution decision to control the driving of the vehicle.

[0143] After automatic driving based on the final driving decision, the driving behavior corresponding to the final driving decision is verified, and a punishment and reward value is given to the final driving decision according to the verification result; when the punishment and reward value of the final driving decision reaches a set threshold, the final driving decision is uploaded to a cloud server for learning to update the basic driving decision. For example, the driving behavior corresponding to the final driving decision is verified according to the actual driving result, if the driving is successfully completed, a positive reward is given, the reward value +1. If the driving is not successfully completed, such as being stopped by a vehicle or a pedestrian, the reward value -1. And adaptively change the corresponding influence coefficient according to the failure reason. When the reward number reaches 50, it is recognized as an effective judgment, and the final driving decision is uploaded to the cloud server for the cloud AI big model to learn, so as to improve the intelligent control of the road segment.

[0144] Referring to Figure 4 , is a schematic diagram of the interaction logic between the vehicle and the cloud server. In the vehicle, the expected driving behavior of the surrounding pedestrians, non-motor vehicles and cars is predicted according to the real-time environmental data around the vehicle, and the expected driving behavior of the vehicle that can cope with the actual traffic environment is obtained by combining the real-time driving data of the vehicle. In addition, by combining the intelligent driving domain controller and the cloud server, the basic driving decision is downloaded from the cloud server. Since the basic driving decision is obtained by learning the historical driving decisions of all vehicles passing through the current road segment by the cloud server, the reference accuracy of the basic driving decision for the current road segment can be guaranteed. The final driving strategy obtained by coupling the basic driving decision with the expected driving behavior of the vehicle can well cope with the real-time and complex traffic environment, and can intelligently control the vehicle with high precision, thereby enhancing the stability of automatic driving.

[0145] By the technical scheme of the present application, the massive data and computing power pressure are given to the cloud server, which not only reduces the intelligent driving cost of the single vehicle, but also greatly improves the learning effect and iteration ability of the intelligent driving big model due to the relatively concentrated learning data of the cloud server, thereby enhancing the intelligent control of each road segment.

[0146] In a third aspect, based on the same inventive concept as the automatic driving method provided in the first aspect, the present application also provides an automatic driving device loaded in an intelligent driving domain controller chip, referring to Figure 5 , comprising:

[0147] The acquisition module 501 is configured to acquire real-time driving data of the vehicle and collect real-time environmental data around the vehicle; wherein the real-time driving data includes the position of the current road segment where the vehicle is located; and the real-time environmental data includes real-time motion behavior data of a target group in a motion state and traffic indication data.

[0148] The communication module 502 is configured to download a basic driving decision corresponding to the current road section from a cloud server, wherein the basic driving decision is obtained by learning historical driving decisions of all vehicles passing through the current road section by the cloud server;

[0149] The first prediction module 503 is configured to predict an expected behavior of the target group in the motion state according to real-time motion behavior data of the target group in the motion state.

[0150] The second prediction module 504 is configured to predict an expected driving behavior of the ego vehicle according to the expected behavior of the target group in the motion state, the traffic indication data and the real-time driving data.

[0151] The coupling module 505 is configured to couple the expected driving behavior of the ego vehicle with the basic driving decision to obtain a final driving decision for controlling driving, and automatically drive based on the final driving decision.

[0152] It should be noted that the specific manner in which each module performs operations is described in detail in the method embodiments provided in the first aspect above, and the specific implementation process can be referred to the method embodiments provided in the first aspect above, which will not be described in detail here.

[0153] In a fourth aspect, based on the same inventive concept as the method provided in the first aspect, the embodiments of the present application also disclose an intelligent driving car, comprising a memory, an intelligent driving domain controller, and a computer program stored in the memory and capable of running on the intelligent driving domain controller, characterized in that the intelligent driving domain controller implements the steps of the method of any of the preceding aspects when executing the program.

[0154] Although the preferred embodiments of the present application have been described, those skilled in the art can make further changes and modifications to these embodiments once they know the basic inventive concept. Therefore, the appended claims are intended to be interpreted as including all changes and modifications falling within the scope of the present application.

[0155] Obviously, those skilled in the art can make various modifications and variations to the present application without departing from the spirit and scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the claims of the present application and their equivalent technologies, the present application also intends to include these modifications and variations.

Claims

1. An automatic driving method, characterized by, The method comprises: acquiring real-time driving data of the vehicle and collecting real-time environment data around the vehicle; wherein the real-time driving data comprises a position of a current road section where the vehicle is located; the real-time environment data comprises real-time motion behavior data of a target group in a motion state and traffic indication data; downloading a basic driving decision corresponding to the current road section from a cloud server; wherein the basic driving decision is obtained by learning historical driving decisions of all vehicles passing through the current road section by the cloud server; predicting an expected behavior of the target group in the motion state according to the real-time motion behavior data of the target group in the motion state, specifically comprising: if the target group in the motion state is a pedestrian or a non-motor vehicle, analyzing the real-time motion behavior data of the pedestrian or the non-motor vehicle to obtain a current behavior intention of the pedestrian or the non-motor vehicle; searching for a historical behavior intention of the pedestrian or the non-motor vehicle on the current road section; coupling the historical behavior intention with the current behavior intention to obtain an expected behavior of the pedestrian or the non-motor vehicle and an execution probability thereof; if the target group in the motion state is a car located around the vehicle, extracting car navigation key data from the real-time motion behavior data of the car; wherein the car navigation key data at least comprises: a navigation destination, a travel purpose, a lane where the car is located, and a driver gender; predicting a plurality of car driving behavior intentions according to the lane where the car is located; predicting execution probabilities of the plurality of car driving behavior intentions by referring to the navigation destination, the travel purpose, and the driver gender; predicting an expected driving behavior of the vehicle according to the expected behavior of the target group in the motion state, the traffic indication data, and the real-time driving data, specifically comprising: predicting the expected driving behavior of the vehicle according to the expected behavior of the pedestrian or the non-motor vehicle and the execution probability thereof, the execution probabilities of the plurality of car driving behavior intentions, the traffic indication data, and the real-time driving data; coupling the expected driving behavior of the vehicle with the basic driving decision to obtain a final driving decision for controlling driving, and automatically driving based on the final driving decision.

2. The method of claim 1, wherein, The method comprises: extracting a lane where the vehicle is located from the real-time driving data, predicting a plurality of vehicle driving behavior intentions according to the lane where the vehicle is located; determining a first influence coefficient on driving of the vehicle according to the expected behavior of the pedestrian or the non-motor vehicle and the execution probability thereof; determining a second influence coefficient on driving of the vehicle according to the plurality of car driving behavior intentions and the execution probabilities thereof; performing conditional coupling on the vehicle driving data by referring to the traffic indication data and / or the car driving data to determine a third influence coefficient on driving of the vehicle; According to the first influence coefficient, the second influence coefficient, and the third influence coefficient, a target execution probability of each of the several driving behavior intentions of the ego vehicle is determined.

3. The method of claim 2, wherein, The coupling of the basic driving decision and the expected driving behavior of the ego vehicle to obtain a final driving decision for controlling driving specifically includes: extracting a basic execution probability of each of the several driving behavior intentions of the ego vehicle from the basic driving decision; fusing the basic execution probability and the target execution probability of each of the several driving behavior intentions of the ego vehicle to obtain a final execution probability of each of the several driving behavior intentions of the ego vehicle; selecting a behavior intention with the highest probability from the final execution probability of each of the several driving behavior intentions of the ego vehicle as the final driving decision.

4. The method of claim 1, wherein, After the ego vehicle is automatically driven based on the final driving decision, the method further includes: verifying a driving behavior corresponding to the final driving decision, and giving a punishment and reward value of a punishment and reward mechanism to the final driving decision according to a verification result; when the punishment and reward value of the final driving decision reaches a set threshold, uploading the final driving decision to the cloud server for learning to update the basic driving decision.

5. An automatic driving device characterized by comprising: includes: an acquisition module configured to acquire real-time driving data of the ego vehicle and collect real-time environment data around the ego vehicle; wherein the real-time driving data includes a position of a current road segment where the ego vehicle is located; and the real-time environment data includes real-time motion behavior data of a target group in a motion state and traffic indication data; a communication module configured to download a basic driving decision corresponding to the current road segment from a cloud server; wherein the basic driving decision is obtained by learning historical driving decisions of all vehicles passing through the current road segment by the cloud server; a first prediction module configured to predict an expected behavior of the target group in the motion state according to the real-time motion behavior data of the target group in the motion state, and specifically configured to: if the target group in the motion state is a pedestrian or a non-motor vehicle, analyze the real-time motion behavior data of the pedestrian or the non-motor vehicle to obtain a current behavior intention of the pedestrian or the non-motor vehicle; search for a historical behavior intention of the pedestrian or the non-motor vehicle on the current road segment; and couple the current behavior intention with the historical behavior intention to obtain an expected behavior and an execution probability of the pedestrian or the non-motor vehicle; and if the target group in the motion state is a car around the ego vehicle, extract car navigation key data from the real-time motion behavior data of the car; wherein the car navigation key data at least includes a navigation destination, a travel purpose, a lane where the car is located, and a driver gender; predict several car driving behavior intentions according to the lane where the car is located; and refer to the navigation destination, the travel purpose, and the driver gender to predict execution probabilities of the several car driving behavior intentions. a second prediction module configured to predict an expected driving behavior of the ego vehicle according to expected behaviors of the target group in motion, the traffic indication data and the real-time driving data, specifically configured to predict the expected driving behavior of the ego vehicle according to expected behaviors of the pedestrians or non-motor vehicles and probabilities of the expected behaviors, expected driving behavior intentions of the plurality of vehicles and probabilities of the expected driving behavior intentions, the traffic indication data and the real-time driving data; a coupling module configured to couple the expected driving behavior of the ego vehicle with the basic driving decision to obtain a final driving decision for controlling driving, and to automatically drive based on the final driving decision.

6. An intelligent vehicle, comprising a memory, an intelligent domain controller, and a computer program stored on the memory and executable on the intelligent domain controller, characterized in that, The intelligent driving domain controller implements the steps of the method of any one of claims 1-4 when executing the program.

7. A driving control system characterized by comprising: The system comprises a cloud server, and the intelligent driving vehicle of claim 6.

Images