Automatic driving decision method and system for high interaction scene of imitating previous vehicle to play game

Abstract

The application provides a kind of high interaction scene automatic driving decision method and system that imitates preceding vehicle to carry out game, comprising: step S1: ego vehicle obtains map information and position information, and utilizes sensor to collect environmental data in real time;Step S2: ego vehicle obtains road traffic information according to environmental data and identifies traffic participants;Step S3: ego vehicle analyzes surrounding other vehicles, judges whether there is suitable other vehicle to follow, if there is suitable other vehicle to follow, ego vehicle selects it as preceding vehicle to follow, if there is no suitable other vehicle to follow, ego vehicle drives autonomously;Step S4: if ego vehicle follows preceding vehicle to drive, it imitates preceding vehicle to carry out game in high interaction scene to assist self decision;If ego vehicle drives autonomously, it adopts conservative strategy in high interaction scene.The application improves the decision processing capability of automatic driving vehicle in strong interaction scene by making it imitate preceding vehicle to carry out game, and guarantees the effectiveness of this imitation behavior through a set of methods.

Description

Technical Field

[0001] This invention relates to the field of autonomous driving technology, and more specifically, to an autonomous driving decision-making method and system for highly interactive scenarios that mimic the game played by the vehicle in front. Background Technology

[0002] While autonomous driving technology has advanced to the point where it can recognize and handle common scenarios, in highly interactive situations, the interaction and competition between the autonomous vehicle and surrounding vehicles makes it difficult to predict the behavior of surrounding vehicles, thus hindering the autonomous vehicle's ability to make effective decisions. For example, in an unprotected left turn, the behavior of oncoming vehicles may be influenced by the autonomous vehicle, and their final decisions are related to unobservable factors (such as the driver's personality), making it difficult to predict their trajectories and determine how the autonomous vehicle should respond. Human drivers, on the other hand, typically possess higher intelligence and richer driving experience, enabling them to adaptively make appropriate decisions in such highly interactive scenarios.

[0003] Patent document CN106030609B (application number: 201480074153.9) discloses a driver-assisted navigation system for a primary vehicle. The system may include: at least one image capture device configured to acquire multiple images of a region near the primary vehicle; a data interface; and at least one processing device. The at least one processing device may be configured to: locate a preceding vehicle in the multiple images; determine at least one action taken by the preceding vehicle based on the multiple images; and cause the primary vehicle to mimic the at least one action of the preceding vehicle. While this patent provides a system and method for mimicking a preceding vehicle, it only considers the interaction between the preceding vehicle and the primary vehicle, without considering the influence of other traffic participants on the behavior of both. Therefore, its solution is only suitable for simple, low-interaction driving scenarios where there is little interaction between the primary vehicle and its environment.

[0004] Patent document CN103158705B (application number: 201210521885.0) discloses a method and system for monitoring the behavior of surrounding vehicles, aiming to predict and react to impending hazards on the road even when hazards cannot be directly detected. In one embodiment, the method monitors the area around the vehicle and searches for the presence of one or more target vehicles. If a target vehicle is detected, the method evaluates its behavior, categorizes it into one of several classes, and if its behavior suggests a certain type of impending hazard, then an appropriate preemptive response is formulated to control the vehicle. However, this approach does not consider the influence of other road users on the behavior of the vehicle in front and the vehicle itself, making it unsuitable for highly interactive scenarios.

[0005] This invention mimics the behavior of the vehicle in front while taking into account the influence of other traffic participants. It can effectively utilize the decision-making ability of human drivers to assist in handling highly interactive scenarios with greater decision-making difficulty, thereby enhancing the adaptability of autonomous vehicles. Summary of the Invention

[0006] To address the shortcomings of existing technologies, the purpose of this invention is to provide an autonomous driving decision-making method and system for highly interactive scenarios that mimic the game played by the vehicle in front.

[0007] An autonomous driving decision-making method for highly interactive scenarios, which mimics a game between a vehicle ahead and another vehicle, according to the present invention, includes:

[0008] Step S1: The vehicle acquires map and location information and uses sensors to collect environmental data in real time;

[0009] Step S2: The vehicle acquires road traffic information based on environmental data and identifies traffic participants;

[0010] Step S3: The vehicle analyzes the surrounding vehicles to determine if there is a suitable vehicle to follow. If there is a suitable vehicle, the vehicle selects it as the vehicle in front to follow. If there is no suitable vehicle, the vehicle drives autonomously.

[0011] Step S4: If the vehicle follows the vehicle in front, it will imitate the vehicle in front to engage in a game in a highly interactive scenario to assist its own decision-making; if the vehicle drives autonomously, it will adopt a conservative strategy in a highly interactive scenario.

[0012] Preferably, the vehicle is an autonomous vehicle; the map information includes a high-precision map; the location information includes the vehicle's position on the map; the sensors include cameras and lidar; and the environmental data consists of images and point clouds of the vehicle's surrounding environment.

[0013] The road traffic information includes roads, curbs, lane lines, traffic lights, signs, traffic facilities, and obstacles; the traffic participants include motor vehicles, non-motor vehicles, and pedestrians; the vehicle uses target detection algorithms to extract key targets from the surrounding environment images and point clouds, thereby obtaining road traffic information and identifying all traffic participants, and determining the type, size, position, direction, speed, and acceleration of each traffic participant.

[0014] Preferably, the high-interaction scenario refers to a scenario where there are many other traffic participants besides the vehicle itself, and the other traffic participants have route conflicts with the vehicle itself or the vehicle in front; the route conflict includes the fact that the predicted travel routes of other traffic participants and the planned travel routes of the vehicle itself or the predicted travel routes of the vehicle in front will intersect at a certain moment within a preset time period.

[0015] The vehicle obtains its predicted route based on the position, direction, speed, acceleration, and steering information of other traffic participants or vehicles in front; the vehicle performs path planning and speed planning based on the driving target and environmental data to obtain the planned route.

[0016] Preferably, the method by which the vehicle determines whether there are other vehicles suitable to follow includes:

[0017] The system uses sensors to obtain real-time information on the position, speed, acceleration, and turn signals of other vehicles in the vicinity.

[0018] By analyzing the above information and combining it with a high-precision map, we can obtain the lane information, direction information, and turning information of other vehicles in the vicinity, and then determine their driving target.

[0019] Choose a car from the surrounding vehicles that has the same destination as your own car and is traveling at a safe speed, as the car you should follow.

[0020] The same driving objective includes the same direction and the same steering choice.

[0021] Preferably, the method for a vehicle to select a suitable other vehicle to follow as the lead vehicle includes:

[0022] Scenario 1: If a suitable vehicle to follow is located in front of the vehicle, in the same lane, and within a preset distance, the vehicle will choose it as the vehicle in front and follow it closely while maintaining a safe distance and speed.

[0023] Scenario 2: If a suitable vehicle to follow is located in front of the vehicle, in the same lane, and at a distance greater than or less than a preset distance range, the vehicle will select the vehicle in front as the vehicle in front, accelerate or decelerate for a certain distance to make the distance between the vehicle in front and the vehicle in front moderate, and then follow the vehicle in front closely while ensuring a safe following distance and a safe speed.

[0024] Scenario 3: If a suitable car to follow is located in front of your car and in a different lane, and the lane of the other car is the same type as your car and you can change lanes, then your car chooses the other car as the car in front, changes lanes to the lane of the other car, then judges the distance between your car and the car in front. If it is too close, slow down; if it is too far, speed up; if it is moderate, maintain the distance, so that the distance between your car and the car in front is moderate. Then, while ensuring a safe following distance and a safe speed, follow the car in front closely.

[0025] Scenario 4: If a suitable car to follow is in front of your car and in a different lane, and the car in the same lane as your car but you cannot change lanes, or if the car in the same lane as your car is in a different lane, then abandon the selection of that car as the car in front and continue to look for other cars to follow from the surrounding vehicles.

[0026] Scenario 5: If a suitable vehicle to follow is behind the vehicle and in the same lane, the vehicle slows down or gives other overtaking signals so that the other vehicle can overtake. If the other vehicle overtakes the vehicle and moves into the same lane within a preset time threshold, the process is handled according to Scenario 1 or Scenario 2. If the other vehicle overtakes the vehicle and moves into a different lane within a preset time threshold, the process is handled according to Scenario 3 or Scenario 4. If the other vehicle does not overtake within the preset time threshold, the vehicle is abandoned as the lead vehicle, and the vehicle continues to search for suitable vehicles to follow from other vehicles in the surrounding area.

[0027] Scenario 6: If the other vehicle that can be followed is behind or parallel to your vehicle and is in a different lane, and the lane that the other vehicle is in is the same type as your lane and you can change lanes, then your vehicle should slow down until it is behind the other vehicle, and then proceed as in Scenario 3.

[0028] Scenario 7: If the other vehicle that can be followed is behind or parallel to your vehicle and is in a different lane, and the lane that the other vehicle is in is the same type as your lane but you cannot change lanes, or the lane that the other vehicle is in is different from your lane, then abandon selecting that vehicle as the lead vehicle and continue to look for other vehicles that can be followed from the surrounding vehicles.

[0029] Preferably, the method of the self-driving vehicle following the vehicle in front and imitating the vehicle in front in a highly interactive scenario includes:

[0030] First, a distance judgment method is used to identify traffic participants whose distance from the vehicle in front or their own vehicle is less than a preset distance threshold as key targets. Then, the following strategies are adopted:

[0031] Strategy 1: If the predicted route of the key target does not intersect with the predicted route of the vehicle in front and the planned route of the vehicle itself, then regardless of the type of traffic participant the key target belongs to, the vehicle will continue to follow the vehicle in front and imitate its driving behavior.

[0032] Strategy 2: If the key target is a motor vehicle and its predicted route intersects with the predicted route of the vehicle in front, then the key target's opponent is the vehicle in front. Regardless of whether the vehicle in front succeeds or fails in the game, the vehicle continues to follow the vehicle in front and imitate its driving behavior.

[0033] Strategy 3: If the key target is a motor vehicle and its predicted route intersects with the planned route of the vehicle, then the key target's opponent is the vehicle itself. If the vehicle wins the game, it continues to follow the vehicle in front. If the vehicle loses the game, it adopts a conservative driving strategy, actively abandons the game, stops following the vehicle in front, and then selects a new vehicle to follow after the game ends.

[0034] Strategy 4: If the key target is a non-motorized vehicle or pedestrian and its predicted route intersects with the predicted route of the vehicle in front, then the key target's opponent is the vehicle in front. Regardless of whether the vehicle in front succeeds or fails in the game, the vehicle continues to follow the vehicle in front and imitate its driving behavior.

[0035] Strategy 5: If the key target is a non-motorized vehicle or pedestrian and their predicted route intersects with the vehicle's planned route, then the key target's opponent is the vehicle itself. The vehicle adopts a conservative driving strategy, only continuing to follow the vehicle in front when it confirms that the other vehicle has stopped or actively given way. Otherwise, the vehicle voluntarily abandons the game and stops following the vehicle in front. After the game ends, the vehicle will choose a new vehicle to follow.

[0036] Preferably, the self-vehicle imitating the driving behavior of the preceding vehicle refers to the self-vehicle imitating the longitudinal and lateral control behaviors of the preceding vehicle, including the self-vehicle imitating the speed, acceleration, and steering of the preceding vehicle;

[0037] A successful negotiation between the vehicle in front and the key target means that the vehicle in front gains the right-of-way from the key target without having to change its original driving state; a failed negotiation between the vehicle in front and the key target means that the vehicle in front fails to gain the right-of-way and needs to stop to let the key target move forward or change direction to avoid the key target.

[0038] When a vehicle engages in a game with a key target that is a motor vehicle, it first maintains its original driving state for a preset time and observes the other party's reaction. If the key target's intention to engage in the game is weak and it relinquishes its right of way, then the vehicle wins the game. If the key target's intention to engage in the game is strong and it strives for the right of way, then the vehicle loses the game.

[0039] Preferably, if, while following another vehicle, the vehicle determines that the vehicle's driving target has changed and is no longer the same as its own, then the vehicle abandons following that vehicle and selects another vehicle from the surrounding vehicles that has the same driving target and is at a safe speed as a suitable vehicle to follow.

[0040] When a vehicle closely follows another vehicle, it maintains the minimum permissible safe distance from the vehicle in front, thereby ensuring that imitating the driving behavior of the vehicle in front can effectively solve decision-making problems in highly interactive scenarios.

[0041] The response time of an autonomous vehicle's emergency braking is shorter than that of a human driver in the same scenario, thus ensuring that the vehicle's braking distance does not exceed a safe following distance, avoiding collisions with the vehicle in front in emergency situations, and enhancing the safety of following the vehicle in front.

[0042] Preferably, the method of autonomous driving and adopting a conservative strategy in highly interactive scenarios includes: when the planned route of the autonomous vehicle intersects with the predicted route of other traffic participants, the autonomous vehicle will only continue to follow the vehicle in front if it confirms that the other party stops first or actively avoids it. In other cases, the autonomous vehicle will actively give up the game and abandon following the vehicle in front, and then select a new vehicle to follow after the game ends.

[0043] An autonomous driving decision-making system for highly interactive scenarios, which mimics a game between a vehicle ahead and another vehicle, according to the present invention, includes:

[0044] Data acquisition module: Acquires map and location information, and uses sensors to collect environmental data in real time;

[0045] Information extraction module: Obtains road traffic information based on environmental data and identifies traffic participants;

[0046] Target recognition module: Analyzes surrounding vehicles to determine if there are any suitable vehicles to follow;

[0047] Lead vehicle following module: If there is a suitable other vehicle to follow, the vehicle will select it as the lead vehicle to follow; if there is no suitable other vehicle to follow, the vehicle will drive autonomously.

[0048] Game Theory Decision Module: If the vehicle is following the vehicle in front, it will imitate the vehicle in front in a high-interaction scenario to assist its own decision-making; if the vehicle is driving autonomously, it will adopt a conservative strategy in a high-interaction scenario.

[0049] Preferably, the system is installed on the vehicle, which is an autonomous vehicle; the map information includes a high-precision map; the location information includes the vehicle's position on the map; the sensors include a camera and a lidar; and the environmental data consists of images and point clouds of the vehicle's surrounding environment.

[0050] The road traffic information includes roads, curbs, lane lines, traffic lights, signs, traffic facilities, and obstacles. The traffic participants include motor vehicles, non-motor vehicles, and pedestrians. The vehicle uses a target detection algorithm to extract key targets from the surrounding environment based on images and point clouds, thereby obtaining road traffic information and identifying all traffic participants, and determining the type, size, position, direction, speed, and acceleration of each traffic participant.

[0051] The high-interaction scenario refers to a scenario where there are many other traffic participants besides the vehicle itself, and the other traffic participants have route conflicts with the vehicle itself or the vehicle in front of it; the route conflict includes the fact that the predicted travel routes of other traffic participants and the planned travel routes of the vehicle itself or the predicted travel routes of the vehicle in front of it will intersect at a certain moment within a preset time period.

[0052] Based on the position, direction, speed, acceleration, and steering information of other traffic participants or vehicles ahead, the predicted route is obtained; based on the vehicle's own driving goal and environmental data, path planning and speed planning are performed to obtain the planned route.

[0053] Preferably, the method by which the target recognition module determines whether there is a suitable other vehicle to follow includes:

[0054] The system uses sensors to obtain real-time information on the position, speed, acceleration, and turn signals of other vehicles in the vicinity.

[0055] By analyzing the above information and combining it with a high-precision map, we can obtain the lane information, direction information, and turning information of other vehicles in the vicinity, and then determine their driving target.

[0056] Choose a car from the surrounding vehicles that has the same destination as your own car and is traveling at a safe speed, as the car you should follow.

[0057] The same driving objective includes the same direction and the same steering choice.

[0058] Preferably, the method by which the leading vehicle following module causes the vehicle to select a suitable other vehicle to follow as the leading vehicle includes:

[0059] Scenario 1: If a suitable vehicle to follow is located in front of the vehicle, in the same lane, and within a preset distance, the vehicle will select that vehicle as the lead vehicle and follow it closely while maintaining a safe distance and speed.

[0060] Scenario 2: If a suitable vehicle to follow is located in front of the vehicle, in the same lane, and at a distance greater than or less than a preset distance range, the vehicle will select the vehicle in front as the vehicle in front, accelerate or decelerate for a certain distance to make the distance between the vehicle in front and the vehicle in front moderate, and then follow the vehicle in front closely while ensuring a safe following distance and a safe speed.

[0061] Scenario 3: If a suitable car to follow is located in front of your car and in a different lane, and the lane of the other car is the same type as your car and you can change lanes, then your car should choose the other car as the car in front, change lanes to the lane of the other car, and then judge the distance between your car and the car in front. If it is too close, slow down; if it is too far, speed up; if it is moderate, maintain the distance, so that the distance between your car and the car in front is moderate. Then, while ensuring a safe following distance and a safe speed, follow the car in front closely.

[0062] Scenario 4: If a suitable car to follow is in front of your car and is in a different lane, if the car is in the same type of lane as your car but you cannot change lanes, or if the car is in a different type of lane than your car, then your car will abandon selecting that car as the car in front and continue to look for a suitable car to follow from other vehicles around.

[0063] Scenario 5: If a suitable vehicle to follow is behind the vehicle and in the same lane, the vehicle should slow down or give other overtaking signals to allow the other vehicle to overtake. If the other vehicle overtakes the vehicle within a preset time threshold and is in the same lane, the process should be handled according to Scenario 1 or Scenario 2. If the other vehicle overtakes the vehicle within a preset time threshold and is in a different lane, the process should be handled according to Scenario 3 or Scenario 4. If the other vehicle does not overtake within the preset time threshold, the vehicle should abandon selecting that vehicle as the lead vehicle and continue searching for suitable vehicles to follow from other vehicles in the surrounding area.

[0064] Scenario 6: If the other vehicle that can be followed is behind or parallel to your vehicle and is in a different lane, and the lane that the other vehicle is in is the same type as your lane and you can change lanes, then slow down your vehicle until you are behind the other vehicle, and then proceed as in Scenario 3.

[0065] Scenario 7: If the other vehicle that can be followed is behind or parallel to the vehicle and is in a different lane, and the lane of the other vehicle is the same type as the vehicle's lane but the vehicle cannot change lanes, or if the lane of the other vehicle is different from the vehicle's lane, then the vehicle should abandon selecting that vehicle as the vehicle in front and continue to look for other vehicles that can be followed from the surrounding vehicles.

[0066] Preferably, the method by which the game decision-making module causes the vehicle to follow the vehicle in front and imitate the vehicle in front in a highly interactive scenario includes:

[0067] First, a distance judgment method is used to identify traffic participants whose distance from the vehicle in front or their own vehicle is less than a preset distance threshold as key targets. Then, the following strategies are adopted:

[0068] Strategy 1: If the predicted route of the key target does not intersect with the predicted route of the vehicle in front and the planned route of the vehicle itself, then regardless of the type of traffic participant the key target belongs to, the vehicle should continue to follow the vehicle in front and imitate its driving behavior.

[0069] Strategy 2: If the key target is a motor vehicle and its predicted route intersects with the predicted route of the vehicle in front, then the key target's opponent is the vehicle in front. Regardless of whether the vehicle in front succeeds or fails in the game, the vehicle continues to follow the vehicle in front and imitates its driving behavior.

[0070] Strategy 3: If the key target is a motor vehicle and its predicted route intersects with the planned route of the vehicle in front, then the key target's opponent is the vehicle in front. If the vehicle in front wins the game, it continues to follow the vehicle in front. If the vehicle in front loses the game, it adopts a conservative driving strategy, actively abandons the game, and stops following the vehicle in front. Then, after the game ends, it selects a new vehicle to follow.

[0071] Strategy 4: If the key target is a non-motorized vehicle or pedestrian and its predicted route intersects with the predicted route of the vehicle in front, then the key target's opponent is the vehicle in front. Regardless of whether the vehicle in front succeeds or fails in the game, the vehicle continues to follow the vehicle in front and imitates its driving behavior.

[0072] Strategy 5: If the key target is a non-motorized vehicle or pedestrian and their predicted route intersects with the planned route of the vehicle, then the key target is the vehicle itself. The vehicle adopts a conservative driving strategy, and only continues to follow the vehicle in front when it is confirmed that the other party stops first or actively avoids it. In other cases, the vehicle actively gives up the game and stops following the vehicle in front. After the game ends, the vehicle will choose a new vehicle to follow.

[0073] The self-vehicle imitating the driving behavior of the vehicle in front refers to the self-vehicle imitating the longitudinal and lateral control behaviors of the vehicle in front, including imitating the speed, acceleration, and steering of the vehicle in front; the self-vehicle successfully engaging in a game with the key target means that the self-vehicle wins the right-of-way from the key target without changing its original driving state; the self-vehicle fails to win the right-of-way means that the self-vehicle fails to win the right-of-way and needs to stop to let the key target move forward or change direction to avoid the key target; when the self-vehicle engages in a game with a key target that is a motor vehicle, it first maintains its original driving state for a preset time and observes the other party's reaction. If the key target's intention to engage in the game is weak and it relinquishes the right-of-way, then the self-vehicle wins the game; if the key target's intention to engage in the game is strong and it strives for the right-of-way, then the self-vehicle fails the game.

[0074] The game decision-making module enables the vehicle to drive autonomously and adopt a conservative strategy in highly interactive scenarios. This includes the following: when the planned route of the vehicle intersects with the predicted route of other traffic participants, the vehicle will only continue to follow the vehicle in front if it confirms that the other party has stopped or actively given way. In other cases, the vehicle will actively give up the game and abandon following the vehicle in front. Then, after the game ends, it will select a new vehicle to follow.

[0075] Preferably, if, during the process of following the vehicle in front, the system determines that the driving target of the vehicle in front has changed and is no longer the same as the driving target of the vehicle in front, then the vehicle in front is instructed to give up following the vehicle in front, and then another vehicle in the surrounding area that has the same driving target as the vehicle in front and is at a safe speed is selected as a suitable vehicle to follow.

[0076] When the vehicle is following closely behind the vehicle in front, the distance between the vehicle and the vehicle in front should be kept at the minimum allowable safe distance, so as to ensure that imitating the driving behavior of the vehicle in front can effectively solve the decision-making problem in highly interactive scenarios.

[0077] This ensures that the vehicle's emergency braking response time is shorter than that of a human driver in the same scenario, thereby guaranteeing that the vehicle's braking distance does not exceed the safe following distance, avoiding collisions with the vehicle in front in emergency situations, and enhancing the safety of following the vehicle in front.

[0078] Compared with the prior art, the present invention has the following beneficial effects: the present invention improves the decision-making and processing capabilities of autonomous vehicles in highly interactive scenarios by having them imitate the vehicles in front of them in a game, and ensures the effectiveness of this imitation behavior through a set of methods. Attached Figure Description

[0079] Other features, objects, and advantages of the present invention will become more apparent from the following detailed description of non-limiting embodiments with reference to the accompanying drawings:

[0080] Figure 1 This is a flowchart of the autonomous driving decision-making method for high-interaction scenarios provided in the embodiments of this application;

[0081] Figure 2 This is a schematic diagram of scenario one in the embodiments of this application;

[0082] Figure 3 This is a schematic diagram of scenario two in the embodiments of this application;

[0083] Figure 4 This is a schematic diagram of scenario three in the embodiments of this application;

[0084] Figure 5 This is a schematic diagram of scenario four in the embodiments of this application;

[0085] Figure 6 This is a schematic diagram of scenario five in the embodiments of this application;

[0086] Figure 7 This is a schematic diagram of scenario six in the embodiments of this application;

[0087] Figure 8 This is a schematic diagram of scenario seven in the embodiments of this application;

[0088] Figure 9 This is a schematic diagram of Strategy 1 in the embodiments of this application;

[0089] Figure 10 This is a schematic diagram of Strategy 2 in the embodiments of this application;

[0090] Figure 11 This is a schematic diagram of strategy three in the embodiments of this application;

[0091] Figure 12 This is a schematic diagram of strategy four in the embodiments of this application;

[0092] Figure 13 This is a schematic diagram of strategy five in the embodiments of this application.

[0093] Among them, 1-self-vehicle, 2-other vehicles suitable for following, 3-vehicle in front, 4-key target of motor vehicle, 5-key target of non-motor vehicle or pedestrian. Detailed Implementation

[0094] The present invention will now be described in detail with reference to specific embodiments. These embodiments will help those skilled in the art to further understand the present invention, but do not limit the invention in any way. It should be noted that those skilled in the art can make several changes and improvements without departing from the concept of the present invention. These all fall within the protection scope of the present invention.

[0095] One embodiment of the present invention provides an autonomous driving decision-making method for highly interactive scenarios that mimic the game-like interaction of the vehicle in front, such as... Figure 1 As shown, it includes:

[0096] Step S1: The vehicle acquires map and location information and uses sensors to collect environmental data in real time;

[0097] Step S2: The vehicle acquires road traffic information based on environmental data and identifies traffic participants;

[0098] Step S3: The vehicle analyzes the surrounding vehicles to determine if there is a suitable vehicle to follow. If there is a suitable vehicle, the vehicle selects it as the vehicle in front to follow. If there is no suitable vehicle, the vehicle drives autonomously.

[0099] Step S4: If the vehicle follows the vehicle in front, it will imitate the vehicle in front to engage in a game in a highly interactive scenario to assist its own decision-making; if the vehicle drives autonomously, it will adopt a conservative strategy in a highly interactive scenario.

[0100] In one embodiment, the vehicle is an autonomous vehicle; the map information includes a high-precision map; the location information includes the vehicle's position on the map; the sensors include cameras and lidar; and the environmental data consists of images and point clouds of the vehicle's surrounding environment.

[0101] The road traffic information includes roads, curbs, lane lines, traffic lights, signs, traffic facilities, and obstacles; the traffic participants include motor vehicles, non-motor vehicles, and pedestrians; the vehicle uses target detection algorithms to extract key targets from the surrounding environment images and point clouds, thereby obtaining road traffic information and identifying all traffic participants, and determining the type, size, position, direction, speed, and acceleration of each traffic participant.

[0102] In one embodiment, the high-interaction scenario refers to a scenario where there are numerous other traffic participants besides the vehicle itself, and these other traffic participants have route conflicts with the vehicle itself or the vehicle in front of it; the route conflict includes the fact that the predicted travel routes of other traffic participants and the planned travel routes of the vehicle itself or the predicted travel routes of the vehicle in front of it will intersect at a certain moment within a preset time period.

[0103] The vehicle obtains its predicted route based on the position, direction, speed, acceleration, and steering information of other traffic participants or vehicles in front; the vehicle performs path planning and speed planning based on the driving target and environmental data to obtain the planned route.

[0104] In one embodiment, the method by which the vehicle determines whether there are other vehicles suitable to follow includes:

[0105] The system uses sensors to obtain real-time information on the position, speed, acceleration, and turn signals of other vehicles in the vicinity.

[0106] By analyzing the above information and combining it with a high-precision map, we can obtain the lane information, direction information, and turning information of other vehicles in the vicinity, and then determine their driving target.

[0107] Choose a car from the surrounding vehicles that has the same destination as your own car and is traveling at a safe speed, as the car you should follow.

[0108] The same driving objective includes the same direction and the same steering choice.

[0109] In one embodiment, the method for a vehicle to select a suitable other vehicle to follow as the lead vehicle includes:

[0110] Scenario 1: For example Figure 2 As shown, if a suitable vehicle to follow is located in front of the vehicle, in the same lane as the vehicle, and within a preset distance range, the vehicle will select the vehicle in front as the vehicle in front and follow it closely while ensuring a safe distance and a safe speed.

[0111] The preset distance range is determined based on traffic regulations and the vehicle's speed, ensuring a suitable distance that allows for safe braking while maintaining a good following distance. In one embodiment, a 3-second rule is used; if the vehicle speed is v (in km / h), the preset distance range is [v / 1.2, v / 1.2+5] (in meters). In another embodiment, a 2-second rule is used; if the vehicle speed is v (in km / h), the preset distance range is [v / 1.8, v / 1.8+5] (in meters).

[0112] Scenario 2: For example Figure 3As shown, if a suitable vehicle to follow is located in front of the vehicle, in the same lane, and at a distance greater than or less than a preset distance range, the vehicle selects it as the vehicle in front, accelerates or decelerates for a certain distance to ensure the distance between the vehicle and the vehicle in front meets the preset conditions, and then follows closely behind the vehicle in front while maintaining a safe following distance and a safe speed; Scenario 3: Figure 4 As shown, if a suitable vehicle to follow is located in front of the vehicle and in a different lane, and the lane of the vehicle is the same type as the vehicle's lane and lane changing is allowed, the vehicle selects the vehicle as the vehicle in front, changes lanes to the lane of the vehicle in front, then judges the distance between the vehicle and the vehicle in front. If it is close, the vehicle slows down; if it is far, the vehicle speeds up. If the preset conditions are met, the vehicle maintains the distance, thus ensuring that the distance between the vehicle and the vehicle in front meets the preset requirements. Then, the vehicle closely follows the vehicle in front while ensuring a safe following distance and a safe speed.

[0113] The lane types include one or two of the five permitted modes of traffic: straight, oncoming, left turn, right turn, and no-entry. When a vehicle you are following is in a lane that includes two permitted modes of traffic, if one of these modes is the same as your vehicle's travel destination, then the two vehicles are considered to be in the same lane type.

[0114] Scenario 4: For example Figure 5 As shown, if a suitable vehicle to follow is located in front of the vehicle and in a different lane, and the vehicle is in the same type of lane as the vehicle but cannot change lanes, or if the vehicle is in a different type of lane, then the vehicle should not be selected as the vehicle in front, and the vehicle should continue to search for suitable vehicles to follow from other vehicles around.

[0115] Scenario 5: For example Figure 6 As shown, if a suitable vehicle to follow is located behind the vehicle and in the same lane, the vehicle slows down or gives other overtaking signals so that the other vehicle can overtake. If the other vehicle overtakes the vehicle within a preset time threshold and is in the same lane, the process is handled according to scenario one or scenario two. If the other vehicle overtakes the vehicle within a preset time threshold and is in a different lane, the process is handled according to scenario three or scenario four. If the other vehicle does not overtake within the preset time threshold, the vehicle is abandoned as the lead vehicle, and the vehicle continues to search for suitable vehicles to follow from other vehicles in the surrounding area.

[0116] In one embodiment, the preset time threshold is 3 seconds.

[0117] Scenario Six: For example Figure 7 As shown, if the other vehicle that can be followed is behind or parallel to the vehicle and is in a different lane, and the lane of the other vehicle is the same type as the lane of the vehicle and lane changing is allowed, the vehicle slows down until it is behind the other vehicle, and then the process is handled according to the method in scenario three.

[0118] Scenario 7: For example Figure 8 As shown, if a vehicle that can be followed is located behind or parallel to the vehicle and is in a different lane, and if the lane of that vehicle is the same type as the vehicle's lane but lane change is not allowed, or if the lane of that vehicle is different from the vehicle's lane, then the vehicle that can be followed is abandoned, and the vehicle continues to search for other vehicles that can be followed from the surrounding vehicles.

[0119] In one embodiment, a method for a self-driving vehicle to follow a vehicle in front and mimic the game-playing behavior of the vehicle in a highly interactive scenario includes:

[0120] First, a distance-based judgment method is used to identify key targets from other traffic participants whose distance from the vehicle in front or their own vehicle is less than a preset distance threshold. In one embodiment, the preset distance threshold is 10 meters. In another embodiment, the preset distance threshold is 20 meters for motor vehicles and 10 meters for non-motorized vehicles or pedestrians.

[0121] Then, the following strategy will be adopted:

[0122] Strategy 1: such as Figure 9 As shown, if the predicted route of the key target does not intersect with the predicted route of the vehicle in front or the planned route of the vehicle itself, then regardless of which type of traffic participant the key target belongs to, the vehicle will continue to follow the vehicle in front and imitate its driving behavior.

[0123] Strategy Two: Figure 10 As shown, if the key target is a motor vehicle and its predicted route intersects with the predicted route of the vehicle in front, then the key target's game partner is the vehicle in front. Regardless of whether the vehicle in front succeeds or fails in the game, the vehicle continues to follow the vehicle in front and imitate its driving behavior.

[0124] Strategy 3: such as Figure 11 As shown, if the key objective is a motor vehicle and its predicted route intersects with the planned route of the vehicle, then the key objective's game partner is the vehicle itself. If the vehicle wins the game, it continues to follow the vehicle in front. If the vehicle loses the game, it adopts a conservative driving strategy, actively abandons the game, gives up following the vehicle in front, and then selects a new vehicle to follow after the game ends.

[0125] Strategy 4: such as Figure 12 As shown, if the key target is a non-motorized vehicle or pedestrian and its predicted route intersects with the predicted route of the vehicle in front, then the key target's game partner is the vehicle in front. Regardless of whether the vehicle in front succeeds or fails in the game, the vehicle continues to follow the vehicle in front and imitate its driving behavior.

[0126] Strategy 5: such as Figure 13As shown, if the key target is a non-motorized vehicle or pedestrian and its predicted route intersects with the planned route of the vehicle, then the game partner of the key target is the vehicle itself. The vehicle adopts a conservative driving strategy and will only continue to follow the vehicle in front if it confirms that the other party stops first or actively avoids it. In other cases, the vehicle will actively give up the game and give up following the vehicle in front. Then, after the game ends, the vehicle will select a new vehicle to follow.

[0127] In one embodiment, the self-vehicle imitating the driving behavior of the preceding vehicle refers to the self-vehicle imitating the longitudinal and lateral control behaviors of the preceding vehicle, including the self-vehicle imitating the speed, acceleration, and steering of the preceding vehicle.

[0128] A successful negotiation between the vehicle in front and the key target means that the vehicle in front gains the right-of-way from the key target without having to change its original driving state; a failed negotiation between the vehicle in front and the key target means that the vehicle in front fails to gain the right-of-way and needs to stop to let the key target move forward or change direction to avoid the key target.

[0129] When a vehicle engages in a game with a key target that is a motor vehicle, it first maintains its original driving state for a preset time and observes the other party's reaction. If the key target's intention to engage in the game is weak and it relinquishes its right of way, then the vehicle wins the game. If the key target's intention to engage in the game is strong and it strives for the right of way, then the vehicle loses the game.

[0130] In one embodiment, if the vehicle determines that the driving target of the vehicle in front has changed and is no longer the same as the driving target of the vehicle itself while following the vehicle in front, the vehicle will give up following the vehicle and then select another vehicle from the surrounding vehicles that has the same driving target as the vehicle and is at a safe speed as a suitable vehicle to follow.

[0131] When a vehicle closely follows another vehicle, it maintains the minimum permissible safe distance from the vehicle in front, thereby ensuring that imitating the driving behavior of the vehicle in front can effectively solve decision-making problems in highly interactive scenarios.

[0132] In one embodiment, the 3-second rule is adopted. If the vehicle speed is v in km / h, the minimum allowable safe following distance is v / 1.2 in meters. In another embodiment, the 2-second rule is adopted. If the vehicle speed is v in km / h, the minimum allowable safe following distance is v / 1.8 in meters.

[0133] The response time of an autonomous vehicle's emergency braking is shorter than that of a human driver in the same scenario, thus ensuring that the vehicle's braking distance does not exceed a safe following distance, avoiding collisions with the vehicle in front in emergency situations, and enhancing the safety of following the vehicle in front.

[0134] In one embodiment, the human driver's response time is 3 seconds. In another embodiment, the human driver's response time is 2 seconds.

[0135] In one embodiment, the method of autonomous driving and adopting a conservative strategy in highly interactive scenarios includes: when the planned route of the autonomous vehicle intersects with the predicted route of other traffic participants, the autonomous vehicle will only continue to follow the vehicle in front if it confirms that the other party stops first or actively avoids it. In other cases, the autonomous vehicle will actively give up the game and abandon following the vehicle in front, and then select a new vehicle to follow after the game ends.

[0136] One embodiment of the present invention provides an autonomous driving decision-making system for highly interactive scenarios that mimic the game played by the vehicle in front, comprising:

[0137] Data acquisition module: Acquires map and location information, and uses sensors to collect environmental data in real time;

[0138] Information extraction module: Obtains road traffic information based on environmental data and identifies traffic participants;

[0139] Target recognition module: Analyzes surrounding vehicles to determine if there are any suitable vehicles to follow;

[0140] Lead vehicle following module: If there is a suitable other vehicle to follow, the vehicle will select it as the lead vehicle to follow; if there is no suitable other vehicle to follow, the vehicle will drive autonomously.

[0141] Game Theory Decision Module: If the vehicle is following the vehicle in front, it will imitate the vehicle in front in a high-interaction scenario to assist its own decision-making; if the vehicle is driving autonomously, it will adopt a conservative strategy in a high-interaction scenario.

[0142] In one embodiment, the system is installed on a vehicle, which is an autonomous vehicle; the map information includes a high-precision map; the location information includes the vehicle's position on the map; the sensors include cameras and lidar; and the environmental data consists of images and point clouds of the vehicle's surrounding environment.

[0143] The road traffic information includes roads, curbs, lane lines, traffic lights, signs, traffic facilities, and obstacles. The traffic participants include motor vehicles, non-motor vehicles, and pedestrians. The vehicle uses a target detection algorithm to extract key targets from the surrounding environment based on images and point clouds, thereby obtaining road traffic information and identifying all traffic participants, and determining the type, size, position, direction, speed, and acceleration of each traffic participant.

[0144] The high-interaction scenario refers to a scenario where there are many other traffic participants besides the vehicle itself, and the other traffic participants have route conflicts with the vehicle itself or the vehicle in front of it; the route conflict includes the fact that the predicted travel routes of other traffic participants and the planned travel routes of the vehicle itself or the predicted travel routes of the vehicle in front of it will intersect at a certain moment within a preset time period.

[0145] Based on the position, direction, speed, acceleration, and steering information of other traffic participants or vehicles ahead, the predicted route is obtained; based on the vehicle's own driving goal and environmental data, path planning and speed planning are performed to obtain the planned route.

[0146] In one embodiment, the method by which the target recognition module determines whether there is another vehicle suitable for following includes:

[0147] The system uses sensors to obtain real-time information on the position, speed, acceleration, and turn signals of other vehicles in the vicinity.

[0148] By analyzing the above information and combining it with a high-precision map, we can obtain the lane information, direction information, and turning information of other vehicles in the vicinity, and then determine their driving target.

[0149] Choose a car from the surrounding vehicles that has the same destination as your own car and is traveling at a safe speed, as the car you should follow.

[0150] The same driving objective includes the same direction and the same steering choice.

[0151] In one embodiment, the method by which the leading vehicle following module causes the vehicle to select a suitable other vehicle to follow as the leading vehicle includes:

[0152] Scenario 1: For example Figure 2 As shown, if a suitable vehicle to follow is located in front of the vehicle, in the same lane, and within a preset distance, the vehicle will select that vehicle as the vehicle in front and follow it closely while ensuring a safe following distance and a safe speed.

[0153] The preset distance range is determined based on traffic regulations and the vehicle's speed, ensuring a suitable distance that allows for safe braking while maintaining a good following distance. In one embodiment, a 3-second rule is used; if the vehicle speed is v (in km / h), the preset distance range is [v / 1.2, v / 1.2+5] (in meters). In another embodiment, a 2-second rule is used; if the vehicle speed is v (in km / h), the preset distance range is [v / 1.8, v / 1.8+5] (in meters).

[0154] Scenario 2: For example Figure 3 As shown, if a suitable vehicle to follow is located in front of the vehicle, in the same lane as the vehicle, and at a distance greater than or less than a preset distance range, the vehicle will select the vehicle in front as the vehicle in front, accelerate or decelerate for a certain distance to make the distance between the vehicle in front and the vehicle in front moderate, and then follow the vehicle in front closely while ensuring a safe following distance and a safe speed.

[0155] Scenario 3: For example Figure 4As shown, if a suitable vehicle to follow is located in front of the vehicle and in a different lane, and the lane of that vehicle is the same type as the vehicle's lane and lane changing is allowed, then the vehicle should choose that vehicle as its leader, change lanes to the lane of the vehicle in front, and then judge the distance between the vehicle and the vehicle in front. If it is too close, slow down; if it is too far, speed up; if it is moderate, maintain the distance, so that the distance between the vehicle and the vehicle in front is moderate. Then, while ensuring a safe following distance and a safe speed, follow the vehicle in front closely.

[0156] The lane types include one or two of the five permitted modes of traffic: straight, oncoming, left turn, right turn, and no-entry. When a vehicle you are following is in a lane that includes two permitted modes of traffic, if one of these modes is the same as your vehicle's travel destination, then the two vehicles are considered to be in the same lane type.

[0157] Scenario 4: For example Figure 5 As shown, if a suitable vehicle to follow is located in front of the vehicle and in a different lane, if the vehicle's lane is the same type as the vehicle's lane but the vehicle cannot change lanes, or if the vehicle's lane is different from the vehicle's lane, the vehicle will abandon selecting that vehicle as the vehicle in front and continue to search for a suitable vehicle to follow from other vehicles around it.

[0158] Scenario 5: For example Figure 6 As shown, if a suitable vehicle to follow is located behind the vehicle and in the same lane, the vehicle will slow down or give other overtaking signals so that the vehicle can overtake. If the vehicle overtakes the vehicle within a preset time threshold and is in the same lane, the process will be handled according to scenario one or scenario two. If the vehicle overtakes the vehicle within a preset time threshold and is in a different lane, the process will be handled according to scenario three or scenario four. If the vehicle does not overtake within the preset time threshold, the vehicle will abandon selecting the vehicle as the lead vehicle and continue to search for suitable vehicles to follow from other vehicles in the surrounding area.

[0159] In one embodiment, the preset time threshold is 3 seconds.

[0160] Scenario Six: For example Figure 7 As shown, if the other vehicle that can be followed is behind or parallel to the vehicle and is in a different lane, and the lane of the other vehicle is the same type as the lane of the vehicle and lane changing is allowed, then the vehicle should slow down until it is behind the other vehicle, and then the process should be handled according to the method in scenario three.

[0161] Scenario 7: For example Figure 8As shown, if a vehicle that can be followed is located behind or parallel to the vehicle and is in a different lane, and the lane of the other vehicle is the same type as the vehicle's lane but the vehicle cannot change lanes, or if the lane of the other vehicle is different from the vehicle's lane, the vehicle will abandon selecting that vehicle as the lead vehicle and continue to search for other vehicles that can be followed from the surrounding vehicles.

[0162] In one embodiment, the game decision-making module enables the vehicle to follow the vehicle in front and imitate the vehicle in front in a highly interactive scenario, including the following method:

[0163] First, a distance-based judgment method is used to identify key targets from other traffic participants whose distance from the vehicle in front or their own vehicle is less than a preset distance threshold. In one embodiment, the preset distance threshold is 10 meters. In another embodiment, the preset distance threshold is 20 meters for motor vehicles and 10 meters for non-motorized vehicles or pedestrians.

[0164] Then, the following strategy will be adopted:

[0165] Strategy 1: such as Figure 9 As shown, if the predicted route of the key target does not intersect with the predicted route of the vehicle in front or the planned route of the vehicle itself, then regardless of which type of traffic participant the key target belongs to, the vehicle will continue to follow the vehicle in front and imitate its driving behavior.

[0166] Strategy Two: Figure 10 As shown, if the key target is a motor vehicle and its predicted route intersects with the predicted route of the vehicle in front, then the key target's game partner is the vehicle in front. Regardless of whether the vehicle in front succeeds or fails in the game, the vehicle continues to follow the vehicle in front and imitates its driving behavior.

[0167] Strategy 3: such as Figure 11 As shown, if the key objective is a motor vehicle and its predicted route intersects with the planned route of the vehicle, then the key objective's game partner is the vehicle itself. If the vehicle wins the game, it continues to follow the vehicle in front. If the vehicle loses the game, it adopts a conservative driving strategy, actively abandons the game, and stops following the vehicle in front. Then, after the game ends, it selects a new vehicle to follow.

[0168] Strategy 4: such as Figure 12 As shown, if the key target is a non-motorized vehicle or pedestrian and its predicted route intersects with the predicted route of the vehicle in front, then the key target's game partner is the vehicle in front. Regardless of whether the vehicle in front succeeds or fails in the game, the vehicle continues to follow the vehicle in front and imitates its driving behavior.

[0169] Strategy 5: such as Figure 13As shown, if the key target is a non-motorized vehicle or pedestrian and its predicted route intersects with the planned route of the vehicle, then the game partner of the key target is the vehicle itself. The vehicle adopts a conservative driving strategy and only continues to follow the vehicle in front when it confirms that the other party stops first or actively avoids it. In other cases, the vehicle actively gives up the game and stops following the vehicle in front. Then, after the game ends, the vehicle selects a new vehicle to follow.

[0170] The self-vehicle imitating the driving behavior of the vehicle in front refers to the self-vehicle imitating the longitudinal and lateral control behaviors of the vehicle in front, including imitating the speed, acceleration, and steering of the vehicle in front; the self-vehicle successfully engaging in a game with the key target means that the self-vehicle wins the right-of-way from the key target without changing its original driving state; the self-vehicle fails to win the right-of-way means that the self-vehicle fails to win the right-of-way and needs to stop to let the key target move forward or change direction to avoid the key target; when the self-vehicle engages in a game with a key target that is a motor vehicle, it first maintains its original driving state for a preset time and observes the other party's reaction. If the key target's intention to engage in the game is weak and it relinquishes the right-of-way, then the self-vehicle wins the game; if the key target's intention to engage in the game is strong and it strives for the right-of-way, then the self-vehicle fails the game.

[0171] The game decision-making module enables the vehicle to drive autonomously and adopt a conservative strategy in highly interactive scenarios. This includes the following: when the planned route of the vehicle intersects with the predicted route of other traffic participants, the vehicle will only continue to follow the vehicle in front if it confirms that the other party has stopped or actively given way. In other cases, the vehicle will actively give up the game and abandon following the vehicle in front. Then, after the game ends, it will select a new vehicle to follow.

[0172] In one embodiment, if, while the vehicle is following the vehicle in front, the system determines that the vehicle's driving target has changed and is no longer the same as the vehicle's driving target, then the vehicle is instructed to abandon following the vehicle, and another vehicle with the same driving target and a safe speed is selected from the surrounding vehicles as a suitable vehicle to follow.

[0173] When the vehicle is following closely behind the vehicle in front, the distance between the vehicle and the vehicle in front should be kept at the minimum allowable safe distance, so as to ensure that imitating the driving behavior of the vehicle in front can effectively solve the decision-making problem in highly interactive scenarios.

[0174] In one embodiment, using the 3-second rule, if the vehicle speed is v in km / h, then the minimum permissible safe following distance is v / 1.2 in meters. In another embodiment, using the 2-second rule, if the vehicle speed is v in km / h, then the minimum permissible safe following distance is v / 1.8 in meters.

[0175] This ensures that the vehicle's emergency braking response time is shorter than that of a human driver in the same scenario, thereby guaranteeing that the vehicle's braking distance does not exceed the safe following distance, avoiding collisions with the vehicle in front in emergency situations, and enhancing the safety of following the vehicle in front.

[0176] In one embodiment, the human driver's response time is 3 seconds. In another embodiment, the human driver's response time is 2 seconds.

[0177] Those skilled in the art will understand that, in addition to implementing the system, apparatus, and their modules provided by this invention in purely computer-readable program code, the same program can be implemented in the form of logic gates, switches, application-specific integrated circuits, programmable logic controllers, and embedded microcontrollers by logically programming the method steps. Therefore, the system, apparatus, and their modules provided by this invention can be considered a hardware component, and the modules included therein for implementing various programs can also be considered structures within the hardware component; alternatively, modules for implementing various functions can be considered both software programs implementing the method and structures within the hardware component.

[0178] Specific embodiments of the present invention have been described above. It should be understood that the present invention is not limited to the specific embodiments described above, and those skilled in the art can make various changes or modifications within the scope of the claims, which do not affect the essence of the present invention. Unless otherwise specified, the embodiments and features described in this application can be arbitrarily combined with each other.

Claims

1. A decision-making method for autonomous driving in highly interactive scenarios that mimics the game-like interaction of the vehicle ahead, characterized in that, include: Step S1: The vehicle acquires map and location information and uses sensors to collect environmental data in real time; Step S2: The vehicle acquires road traffic information based on environmental data and identifies traffic participants; Step S3: The vehicle analyzes the surrounding vehicles to determine if there is a suitable vehicle to follow. If there is a suitable vehicle, the vehicle selects it as the vehicle in front to follow. If there is no suitable vehicle, the vehicle drives autonomously. Step S4: If the vehicle follows the vehicle in front, it will imitate the vehicle in front to engage in a game in a highly interactive scenario to assist its own decision-making; if the vehicle drives autonomously, it will adopt a conservative strategy in a highly interactive scenario. Methods for a self-driving vehicle to follow the vehicle in front and mimic the game-playing behavior of the vehicle in highly interactive scenarios include: First, a distance judgment method is used to identify traffic participants whose distance from the vehicle in front or their own vehicle is less than a preset distance threshold as key targets. Then, the following strategies are adopted: Strategy 1: If the predicted route of the key target does not intersect with the predicted route of the vehicle in front and the planned route of the vehicle itself, then regardless of the type of traffic participant the key target belongs to, the vehicle will continue to follow the vehicle in front and imitate its driving behavior. Strategy 2: If the key target is a motor vehicle and its predicted route intersects with the predicted route of the vehicle in front, then the key target's opponent is the vehicle in front. Regardless of whether the vehicle in front succeeds or fails in the game, the vehicle continues to follow the vehicle in front and imitate its driving behavior. Strategy 3: If the key target is a motor vehicle and its predicted route intersects with the planned route of the vehicle, then the key target's opponent is the vehicle itself. If the vehicle wins the game, it continues to follow the vehicle in front. If the vehicle loses the game, it adopts a conservative driving strategy, actively abandons the game, stops following the vehicle in front, and then selects a new vehicle to follow after the game ends. Strategy 4: If the key target is a non-motorized vehicle or pedestrian and its predicted route intersects with the predicted route of the vehicle in front, then the key target's opponent is the vehicle in front. Regardless of whether the vehicle in front succeeds or fails in the game, the vehicle continues to follow the vehicle in front and imitate its driving behavior. Strategy 5: If the key target is a non-motorized vehicle or pedestrian and their predicted route intersects with the vehicle's planned route, then the key target's opponent is the vehicle itself. The vehicle adopts a conservative driving strategy, only continuing to follow the vehicle in front when it confirms that the other vehicle has stopped or actively given way. Otherwise, the vehicle voluntarily abandons the game and stops following the vehicle in front. After the game ends, the vehicle will choose a new vehicle to follow.

2. The autonomous driving decision-making method for highly interactive scenarios that mimics the game between vehicles ahead, as described in claim 1, is characterized in that... The vehicle is an autonomous vehicle; the map information includes a high-precision map; the location information includes the vehicle's position on the map; the sensors include cameras and lidar; the environmental data consists of images and point clouds of the vehicle's surrounding environment. The road traffic information includes roads, curbs, lane lines, traffic lights, signs, traffic facilities, and obstacles; the traffic participants include motor vehicles, non-motor vehicles, and pedestrians; the vehicle uses target detection algorithms to extract key targets from the surrounding environment images and point clouds, thereby obtaining road traffic information and identifying all traffic participants, and determining the type, size, position, direction, speed, and acceleration of each traffic participant.

3. The autonomous driving decision-making method for highly interactive scenarios that mimics the game between vehicles ahead, as described in claim 1, is characterized in that... The high-interaction scenario refers to a scenario where there are many other traffic participants besides the vehicle itself, and the other traffic participants have route conflicts with the vehicle itself or the vehicle in front of it; the route conflict includes the fact that the predicted travel routes of other traffic participants and the planned travel routes of the vehicle itself or the predicted travel routes of the vehicle in front of it will intersect at a certain moment within a preset time period. The vehicle obtains its predicted route based on the position, direction, speed, acceleration, and steering information of other traffic participants or vehicles in front; The vehicle performs path and speed planning based on the driving target and environmental data to obtain the planned travel route.

4. The autonomous driving decision-making method for highly interactive scenarios that mimics the game between the vehicle ahead and the driver, as described in claim 1, is characterized in that... Methods for a driver to determine if there are suitable other vehicles to follow include: The system uses sensors to obtain real-time information on the position, speed, acceleration, and turn signals of other vehicles in the vicinity. By analyzing the above information and combining it with a high-precision map, we can obtain the lane information, direction information, and turning information of other vehicles in the vicinity, and then determine their driving target. Choose a car from the surrounding vehicles that has the same destination as your own car and is traveling at a safe speed, as the car you should follow. The same driving objective includes the same direction and the same steering choice.

5. The autonomous driving decision-making method for highly interactive scenarios that mimics the game played by the vehicle in front, as described in claim 1, is characterized in that... Methods for a self-driving vehicle to choose a suitable other vehicle as the lead vehicle for following include: Scenario 1: If a suitable vehicle to follow is located in front of the vehicle, in the same lane, and within a preset distance, the vehicle will choose it as the vehicle in front and follow it closely while maintaining a safe distance and speed. Scenario 2: If a suitable vehicle to follow is located in front of the vehicle, in the same lane, and at a distance greater than or less than a preset distance range, the vehicle will select the vehicle in front as the vehicle in front, accelerate or decelerate for a certain distance to make the distance between the vehicle in front and the vehicle in front meet the preset conditions, and then follow the vehicle in front closely while ensuring a safe following distance and a safe speed. Scenario 3: If a suitable car to follow is located in front of the vehicle and in a different lane, and the lane of the other car is the same type as the vehicle's lane and lane changing is allowed, then the vehicle selects the other car as the lead car, changes lanes to the lane of the other car, and then judges the distance between the vehicle and the other car. If it is close, the vehicle slows down; if it is far, the vehicle speeds up. If the preset conditions are met, the vehicle maintains the distance, thus ensuring that the distance between the vehicle and the other car meets the preset requirements. Then, the vehicle closely follows the other car while ensuring a safe following distance and a safe speed. Scenario 4: If a suitable car to follow is in front of your car and in a different lane, and the car in the same lane as your car but you cannot change lanes, or if the car in the same lane as your car is in a different lane, then abandon the selection of that car as the car in front and continue to look for other cars to follow from the surrounding vehicles. Scenario 5: If a suitable vehicle to follow is behind the vehicle and in the same lane, the vehicle slows down or gives other overtaking signals so that the other vehicle can overtake. If the other vehicle overtakes the vehicle and moves into the same lane within a preset time threshold, the process is handled according to Scenario 1 or Scenario 2. If the other vehicle overtakes the vehicle and moves into a different lane within a preset time threshold, the process is handled according to Scenario 3 or Scenario 4. If the other vehicle does not overtake within the preset time threshold, the vehicle is abandoned as the lead vehicle, and the vehicle continues to search for suitable vehicles to follow from other vehicles in the surrounding area. Scenario 6: If the other vehicle that can be followed is behind or parallel to your vehicle and is in a different lane, and the lane that the other vehicle is in is the same type as your lane and you can change lanes, then your vehicle should slow down until it is behind the other vehicle, and then proceed as in Scenario 3. Scenario 7: If the other vehicle that can be followed is behind or parallel to your vehicle and is in a different lane, and the lane that the other vehicle is in is the same type as your lane but you cannot change lanes, or the lane that the other vehicle is in is different from your lane, then abandon selecting that vehicle as the lead vehicle and continue to look for other vehicles that can be followed from the surrounding vehicles.

6. The autonomous driving decision-making method for highly interactive scenarios that mimics the game between the vehicle ahead and the driver, as described in claim 5, is characterized in that... The self-vehicle imitating the driving behavior of the vehicle in front refers to the self-vehicle imitating the longitudinal and lateral control behaviors of the vehicle in front, including the self-vehicle imitating the speed, acceleration and steering of the vehicle in front; A successful negotiation between the vehicle in front and the key target means that the vehicle in front gains the right-of-way from the key target without having to change its original driving state; a failed negotiation between the vehicle in front and the key target means that the vehicle in front fails to gain the right-of-way and needs to stop to let the key target move forward or change direction to avoid the key target. When a vehicle engages in a game with a key target that is a motor vehicle, it first maintains its original driving state for a preset time and observes the other party's reaction. If the key target's intention to engage in the game is weak and it relinquishes its right of way, then the vehicle wins the game. If the key target's intention to engage in the game is strong and it strives for the right of way, then the vehicle loses the game.

7. The autonomous driving decision-making method for highly interactive scenarios that mimics the game between the vehicle ahead and the driver, as described in any one of claims 4-6, is characterized in that... If, while following another vehicle, the vehicle determines that the vehicle's driving goal has changed and is no longer the same as its own, then the vehicle will abandon following that vehicle and select another vehicle from the surrounding area that has the same driving goal and is traveling at a safe speed as a suitable vehicle to follow. When a vehicle closely follows another vehicle, it maintains the minimum permissible safe distance from the vehicle in front, thereby ensuring that imitating the driving behavior of the vehicle in front can effectively solve decision-making problems in highly interactive scenarios. The response time of an autonomous vehicle's emergency braking is shorter than that of a human driver in the same scenario, thus ensuring that the vehicle's braking distance does not exceed a safe following distance, avoiding collisions with the vehicle in front in emergency situations, and enhancing the safety of following the vehicle in front.

8. The autonomous driving decision-making method for high-interaction scenarios that mimics the game between the vehicle ahead and the driver, as described in claim 1, is characterized in that... Methods for autonomous driving and adopting a conservative strategy in highly interactive scenarios include: when the planned route of the autonomous vehicle intersects with the predicted route of other traffic participants, the autonomous vehicle will only continue to follow the vehicle in front if it confirms that the other party has stopped or actively given way. In other cases, the autonomous vehicle will actively give up the game and abandon following the vehicle in front, and then select a new vehicle to follow after the game ends.

9. An autonomous driving decision-making system for highly interactive scenarios that mimics the game-like interaction of the vehicle ahead, characterized in that, include: Data acquisition module: Acquires map and location information, and uses sensors to collect environmental data in real time; Information extraction module: Obtains road traffic information based on environmental data and identifies traffic participants; Target recognition module: Analyzes surrounding vehicles to determine if there are any suitable vehicles to follow; Lead vehicle following module: If there is a suitable other vehicle to follow, the vehicle will select it as the lead vehicle to follow; if there is no suitable other vehicle to follow, the vehicle will drive autonomously. Game Theory Decision Module: If the vehicle is following the vehicle in front, it will imitate the vehicle in front in a high-interaction scenario to assist its own decision-making; if the vehicle is driving autonomously, it will adopt a conservative strategy in a high-interaction scenario. The game-playing decision-making module enables the vehicle to follow the vehicle in front and mimic the vehicle in front to engage in game-playing scenarios, including the following methods: First, a distance judgment method is used to identify traffic participants whose distance from the vehicle in front or their own vehicle is less than a preset distance threshold as key targets. Then, the following strategies are adopted: Strategy 1: If the predicted route of the key target does not intersect with the predicted route of the vehicle in front and the planned route of the vehicle itself, then regardless of the type of traffic participant the key target belongs to, the vehicle should continue to follow the vehicle in front and imitate its driving behavior. Strategy 2: If the key target is a motor vehicle and its predicted route intersects with the predicted route of the vehicle in front, then the key target's opponent is the vehicle in front. Regardless of whether the vehicle in front succeeds or fails in the game, the vehicle continues to follow the vehicle in front and imitates its driving behavior. Strategy 3: If the key target is a motor vehicle and its predicted route intersects with the planned route of the vehicle in front, then the key target's opponent is the vehicle in front. If the vehicle in front wins the game, it continues to follow the vehicle in front. If the vehicle in front loses the game, it adopts a conservative driving strategy, actively abandons the game, and stops following the vehicle in front. Then, after the game ends, it selects a new vehicle to follow. Strategy 4: If the key target is a non-motorized vehicle or pedestrian and its predicted route intersects with the predicted route of the vehicle in front, then the key target's opponent is the vehicle in front. Regardless of whether the vehicle in front succeeds or fails in the game, the vehicle continues to follow the vehicle in front and imitates its driving behavior. Strategy 5: If the key target is a non-motorized vehicle or pedestrian and their predicted route intersects with the vehicle's planned route, then the key target's opponent is the vehicle itself. The vehicle adopts a conservative driving strategy, only continuing to follow the vehicle in front when it confirms that the other vehicle has stopped or actively given way. Otherwise, the vehicle voluntarily abandons the game and stops following the vehicle in front. After the game ends, the vehicle will choose a new vehicle to follow.

10. The highly interactive autonomous driving decision-making system for imitating the preceding vehicle in a game, as described in claim 9, is characterized in that... The system is installed on the vehicle, which is an autonomous vehicle. The map information includes a high-precision map, the location information includes the vehicle's position on the map, the sensors include cameras and lidar, and the environmental data consists of images and point clouds of the vehicle's surrounding environment. The road traffic information includes roads, curbs, lane lines, traffic lights, signs, traffic facilities, and obstacles. The traffic participants include motor vehicles, non-motor vehicles, and pedestrians. The vehicle uses a target detection algorithm to extract key targets from the surrounding environment based on images and point clouds, thereby obtaining road traffic information and identifying all traffic participants, and determining the type, size, position, direction, speed, and acceleration of each traffic participant. The high-interaction scenario refers to a scenario where there are many other traffic participants besides the vehicle itself, and the other traffic participants have route conflicts with the vehicle itself or the vehicle in front of it; the route conflict includes the fact that the predicted travel routes of other traffic participants and the planned travel routes of the vehicle itself or the predicted travel routes of the vehicle in front of it will intersect at a certain moment within a preset time period. Based on the position, direction, speed, acceleration, and steering information of other traffic participants or vehicles ahead, the predicted route is obtained; based on the vehicle's own driving goal and environmental data, path planning and speed planning are performed to obtain the planned route.

11. The highly interactive autonomous driving decision-making system for imitating the preceding vehicle in a game, as described in claim 9, is characterized in that... The target recognition module determines whether there is a suitable other vehicle to follow using the following methods: The system uses sensors to obtain real-time information on the position, speed, acceleration, and turn signals of other vehicles in the vicinity. By analyzing the above information and combining it with a high-precision map, we can obtain the lane information, direction information, and turning information of other vehicles in the vicinity, and then determine their driving target. Choose a car from the surrounding vehicles that has the same destination as your own car and is traveling at a safe speed, as the car you should follow. The same driving objective includes the same direction and the same steering choice.

12. The highly interactive autonomous driving decision-making system for imitating the preceding vehicle in a game, as described in claim 9, is characterized in that... The method by which the lead vehicle following module enables the vehicle to select a suitable other vehicle to follow as the lead vehicle includes: Scenario 1: If a suitable vehicle to follow is located in front of the vehicle, in the same lane, and within a preset distance, the vehicle will select that vehicle as the lead vehicle and follow it closely while maintaining a safe distance and speed. Scenario 2: If a suitable vehicle to follow is located in front of the vehicle, in the same lane, and at a distance greater than or less than a preset distance range, the vehicle will select the vehicle in front as the vehicle in front, accelerate or decelerate for a certain distance to make the distance between the vehicle in front and the vehicle in front meet the preset requirements, and then follow the vehicle in front closely while ensuring a safe following distance and a safe speed. Scenario 3: If a suitable car to follow is located in front of the vehicle and in a different lane, and the lane of the other car is the same type as the vehicle's lane and lane changing is allowed, then the vehicle selects the other car as the lead car, changes lanes to the lane of the other car, and then judges the distance between the vehicle and the other car. If it is close, the vehicle slows down; if it is far, the vehicle speeds up. If the preset requirements are met, the vehicle maintains the distance between the vehicle and the other car, so that the distance between the vehicle and the other car meets the preset requirements. Then, the vehicle closely follows the other car while ensuring a safe following distance and a safe speed. Scenario 4: If a suitable car to follow is in front of your car and is in a different lane, if the car is in the same type of lane as your car but you cannot change lanes, or if the car is in a different type of lane than your car, then your car will abandon selecting that car as the car in front and continue to look for a suitable car to follow from other vehicles around. Scenario 5: If a suitable vehicle to follow is behind the vehicle and in the same lane, the vehicle should slow down or give other overtaking signals to allow the other vehicle to overtake. If the other vehicle overtakes the vehicle within a preset time threshold and is in the same lane, the process should be handled according to Scenario 1 or Scenario 2. If the other vehicle overtakes the vehicle within a preset time threshold and is in a different lane, the process should be handled according to Scenario 3 or Scenario 4. If the other vehicle does not overtake within the preset time threshold, the vehicle should abandon selecting that vehicle as the lead vehicle and continue searching for suitable vehicles to follow from other vehicles in the surrounding area. Scenario 6: If the other vehicle that can be followed is behind or parallel to your vehicle and is in a different lane, and the lane that the other vehicle is in is the same type as your lane and you can change lanes, then slow down your vehicle until you are behind the other vehicle, and then proceed as in Scenario 3. Scenario 7: If the other vehicle that can be followed is behind or parallel to the vehicle and is in a different lane, and the lane of the other vehicle is the same type as the vehicle's lane but the vehicle cannot change lanes, or if the lane of the other vehicle is different from the vehicle's lane, then the vehicle should abandon selecting that vehicle as the vehicle in front and continue to look for other vehicles that can be followed from the surrounding vehicles.

13. The highly interactive autonomous driving decision-making system for imitating the preceding vehicle in a game, as described in claim 9, is characterized in that... The self-vehicle imitating the driving behavior of the vehicle in front refers to the self-vehicle imitating the longitudinal and lateral control behaviors of the vehicle in front, including imitating the speed, acceleration, and steering of the vehicle in front; the self-vehicle successfully engaging in a game with the key target means that the self-vehicle wins the right-of-way from the key target without changing its original driving state; the self-vehicle fails to win the right-of-way means that the self-vehicle fails to win the right-of-way and needs to stop to let the key target move forward or change direction to avoid the key target; when the self-vehicle engages in a game with a key target that is a motor vehicle, it first maintains its original driving state for a preset time and observes the other party's reaction. If the key target's intention to engage in the game is weak and it gives up the right-of-way, then the self-vehicle wins the game; if the key target's intention to engage in the game is strong and it strives for the right-of-way, then the self-vehicle loses the game. The game decision-making module enables the vehicle to drive autonomously and adopt a conservative strategy in highly interactive scenarios. This includes the following: when the planned route of the vehicle intersects with the predicted route of other traffic participants, the vehicle will only continue to follow the vehicle in front if it confirms that the other party has stopped or actively given way. In other cases, the vehicle will actively give up the game and abandon following the vehicle in front. Then, after the game ends, it will select a new vehicle to follow.

14. The highly interactive autonomous driving decision-making system for imitating the preceding vehicle in a game, as described in any one of claims 9-13, is characterized in that... If, while the vehicle is following the vehicle in front, the system determines that the vehicle's driving target has changed and is no longer the same as the vehicle's driving target, then the vehicle will stop following the vehicle and select another vehicle from the surrounding vehicles that has the same driving target and is at a safe speed as a suitable vehicle to follow. When the vehicle is following closely behind the vehicle in front, the distance between the vehicle and the vehicle in front should be kept at the minimum allowable safe distance, so as to ensure that imitating the driving behavior of the vehicle in front can effectively solve the decision-making problem in highly interactive scenarios. This ensures that the vehicle's emergency braking response time is shorter than that of a human driver in the same scenario, thereby guaranteeing that the vehicle's braking distance does not exceed the safe following distance, avoiding collisions with the vehicle in front in emergency situations, and enhancing the safety of following the vehicle in front.

Images