Unmanned speed control method and system based on artificial intelligence

Abstract

The invention discloses an unmanned driving speed control method and system based on artificial intelligence, and relates to the technical field of intelligent driving speed control, and the method comprises the following steps: S1, obtaining the information of surrounding vehicles through a vehicle-mounted camera, judging whether the surrounding vehicles have an overtaking intention or not through the information of the surrounding vehicles, and if there is an overtaking intention vehicle, executing the step S2; s2, if yes, judging whether an overtaking speed-making condition exists or not according to surrounding vehicle information and traffic rules, if yes, executing an overtaking speed-making action, and if not, executing S2; the information of the surrounding vehicles is obtained through the vehicle-mounted camera, whether the surrounding vehicles have the overtaking intention or not is judged according to the states of the surrounding vehicles, and whether the vehicles with the overtaking intention have the conditions for reducing the speed of the unmanned vehicle to facilitate rapid overtaking of the vehicle with the overtaking intention or not is judged, so that convenience is provided for rapid overtaking of the vehicle with the overtaking intention.

Description

Technical Field

[0001] This invention relates to the field of intelligent driving speed control technology, specifically to a speed control method and system for unmanned driving based on artificial intelligence. Background Technology

[0002] Autonomous vehicle systems use multiple sensors, such as lidar, cameras, and millimeter-wave radar, to perceive the environment in real time. They combine high-precision maps and GNSS positioning to achieve precise navigation and rely on algorithms and behavior prediction for decision-making and control. The system needs to handle complex scenarios such as obstacle avoidance, lane changing, and traffic light recognition.

[0003] Currently, the biggest challenge in decision-making for autonomous vehicles is the ethical dilemma. When an autonomous vehicle is waiting at a traffic light or driving normally, and ambulances, police cars, fire trucks, or other vehicles need to pass quickly, human-driven vehicles will make unconventional yielding maneuvers to meet the needs of these emergency vehicles to quickly carry out their missions. However, when faced with overtaking requests from emergency vehicles, autonomous vehicles prioritize making decisions based on traffic rules in their algorithms or their own safety assessments. For example, if an autonomous vehicle could give way by briefly crossing a solid line or slightly below the minimum speed limit, allowing the emergency vehicle to quickly overtake and pass, algorithmic limitations might prevent the autonomous vehicle from making the yielding maneuver. This would prevent the emergency vehicle from quickly overtaking and carrying out its mission. Therefore, when autonomous vehicles face emergencies, the algorithm settings are too rigid, making it difficult to achieve intelligent analysis and decision-making to handle ethical dilemmas. To address this, we propose an AI-based speed control method and system for autonomous driving. Summary of the Invention

[0004] (a) Technical problems to be solved To address the shortcomings of existing technologies, this invention provides a speed control method and system for autonomous driving based on artificial intelligence, in order to solve the aforementioned problems in the prior art.

[0005] (II) Technical Solution To achieve the above objectives, the present invention provides a speed control method for autonomous driving based on artificial intelligence, comprising the following steps: S1: Obtain information about surrounding vehicles through the vehicle-mounted camera, determine whether surrounding vehicles intend to overtake, if there are vehicles intending to overtake, determine whether there are conditions for overtaking and yielding based on the surrounding vehicle information and traffic rules, if there are conditions for overtaking and yielding, execute the overtaking and yielding action, if there are no conditions for overtaking and yielding, execute S2. S2: Locate the unmanned vehicle's position using GPS positioning, obtain the current road restriction rules based on the unmanned vehicle's position information, conduct a risk assessment of overtaking and yielding based on surrounding vehicle information and the current road restriction rules, obtain a yielding risk assessment score, determine whether the conditions for yielding and overtaking are met based on the yielding risk assessment score, if the conditions for overtaking and yielding are met, then execute the overtaking and yielding action; if the conditions for overtaking and yielding are not met, then execute S3. S3: Obtain information on vehicles intending to overtake. Based on this information, determine whether the vehicle intending to overtake is an emergency vehicle. If it is, obtain information on the yielding status of surrounding vehicles through the vehicle-mounted camera. Use this information in conjunction with the current road rules to determine whether there are conditions for yielding speed. If there are conditions, execute the yielding speed action. If there are still no conditions, use the yielding status of surrounding vehicles and the yielding risk assessment score to determine whether to yield. Execute driving actions based on the assessment results.

[0006] Preferably, in S1, information about surrounding vehicles is acquired through an onboard camera, and the surrounding vehicle information is used to determine whether the surrounding vehicles intend to overtake. Specifically: S101: Obtain the turn signal status of the following vehicle through the vehicle camera, determine whether the following vehicle has turned on its turn signal, if the following vehicle has turned on its turn signal, mark that the following vehicle has the intention to overtake, if the following vehicle has not turned on its turn signal, then execute S102. S102: Continuously acquire the distance between the following vehicle and the unmanned vehicle, set a preset distance threshold, and determine whether the distance between the following vehicle and the unmanned vehicle is less than the preset distance threshold. If the distance between the following vehicle and the unmanned vehicle is less than or equal to the preset distance threshold, then acquire the front wheel deflection angle of the following vehicle through the vehicle camera. S103: Set a preset threshold for the front wheel deflection angle of the following vehicle, determine whether the front wheel deflection angle of the following vehicle is greater than the preset threshold. If the front wheel deflection angle of the following vehicle is greater than the preset threshold, mark that the following vehicle has the intention to overtake, and record the overtaking direction according to the direction of the front wheel deflection angle of the following vehicle. If the front wheel deflection angle of the following vehicle is less than or equal to the preset threshold, repeat S101.

[0007] Preferably, in S1, if there is a vehicle intending to overtake, the system determines whether the conditions for yielding to overtake exist based on surrounding vehicle information and traffic rules. Specifically: S104: Obtain the minimum speed limit requirement of the road at the current location of the unmanned vehicle, set a preset threshold for safe speed, and obtain the minimum safe speed by summing the minimum speed limit requirement of the road at the current location of the unmanned vehicle with the preset threshold for safe speed. S105: Obtain the current speed of the unmanned vehicle and determine whether the current speed of the unmanned vehicle is greater than the minimum safe speed. If the current speed of the unmanned vehicle is greater than the minimum safe speed, then execute S106. If the current speed of the unmanned vehicle is equal to the minimum safe speed, then the result is that there is no overtaking speed limit. S106: Obtain the distance between the following vehicle and the unmanned vehicle, set a preset threshold for the safe speed-yielding distance, and determine whether the distance between the following vehicle and the unmanned vehicle is greater than the preset threshold for the safe speed-yielding distance. If the distance between the following vehicle and the unmanned vehicle is greater than the preset threshold for the safe speed-yielding distance, the result is that there is a condition for overtaking and yielding speed. If the distance between the following vehicle and the unmanned vehicle is less than or equal to the preset threshold for the safe speed-yielding distance, the result is that there is no condition for overtaking and yielding speed.

[0008] Preferably, in S2, a risk assessment of overtaking and yielding is performed based on surrounding vehicle information and current road regulations to obtain a yielding risk assessment score, specifically as follows: S201: Obtain the current road restriction rules, determine whether yielding will cross the solid road line, if yielding will cross the solid road line, mark the risk of yielding and crossing the solid line as 1, if yielding will not cross the solid road line, mark the risk of yielding and crossing the solid line as 0. Determine whether yielding will reduce the vehicle speed below the current road minimum speed limit. If yielding will reduce the vehicle speed below the current road minimum speed limit, the risk of speed reduction is marked as 1. If yielding will not reduce the vehicle speed below the current road minimum speed limit, the risk of speed reduction is marked as 0. The risk of slowing down when yielding is summed with the risk of running over a solid line when yielding, resulting in the risk score of the yielding rule. S202: Obtain the speed of the vehicle giving way to the slow lane, obtain the current speed of the unmanned vehicle, obtain the speed difference by subtracting the current speed of the unmanned vehicle from the speed of the vehicle giving way to the slow lane, set a preset threshold for the speed difference, determine whether the speed difference is greater than the preset threshold, if the speed difference is greater than the preset threshold, mark the speed risk as 0, if the speed difference is less than or equal to the preset threshold, mark the speed risk as 1. Obtain the distance between the vehicle giving way to the slow lane and the unmanned vehicle, set a preset threshold for the distance between the vehicles in the slow lane, and determine whether the distance between the vehicle giving way to the slow lane and the unmanned vehicle is greater than the preset threshold. If the distance between the vehicle giving way to the slow lane and the unmanned vehicle is greater than the preset threshold, then mark the distance risk as 0. If the distance between the vehicle giving way to the slow lane and the unmanned vehicle is less than or equal to the preset threshold, then mark the distance risk as 1. The yield safety risk is obtained by summing the risks of vehicle spacing and vehicle speed. S203: Set the yielding rule risk score weight, and obtain the yielding rule risk score by multiplying the yielding rule risk score by the yielding rule risk score weight. Set the yielding safety risk score weight, and obtain the yielding safety risk score by multiplying the yielding safety risk score weight by the yielding safety risk score. Summing the yielding safety risk score and the yielding rule risk score yields the yielding risk assessment score.

[0009] Preferably, in S2, the condition for yielding and overtaking is determined based on the yielding risk assessment score. Specifically, the yielding risk assessment score is obtained, and it is determined whether the yielding risk assessment score is 0. If the yielding risk assessment score is 0, it is marked as having the condition for yielding; if the yielding risk assessment score is not 0, it is marked as not having the condition for yielding.

[0010] Preferably, in S3, information about vehicles intending to overtake is obtained, and based on this information, it is determined whether the vehicle intending to overtake is an emergency or special vehicle. Specifically: S301: Obtain an image of a vehicle intending to overtake using an onboard camera, and identify whether the vehicle intending to overtake is a police car, ambulance, or fire truck based on the image. If the vehicle intending to overtake is not a police car, ambulance, or fire truck, the result is marked as a non-emergency special vehicle. If the vehicle intending to overtake is a police car, ambulance, or fire truck, then execute S302. S302: Acquire the surrounding sounds of the unmanned vehicle through the sound sensor, and use the Mel frequency cepstral coefficient to determine whether there is a siren sound in the surrounding sounds of the unmanned vehicle. If there is a siren sound in the surrounding sounds of the unmanned vehicle, lock the target vehicle through the siren sound, and determine whether the target vehicle is a vehicle intending to overtake. If the target vehicle is a vehicle intending to overtake, execute S303. If the target vehicle is not a vehicle intending to overtake, repeat S301. S303: The vehicle camera determines whether the vehicle preceding the vehicle intending to overtake has yielded to it in the past. If the vehicle preceding the vehicle intending to overtake has yielded to it in the past, the vehicle intending to overtake is defined as an emergency vehicle. If the vehicle preceding the vehicle intending to overtake has not yielded to it in the past, the duration of the siren is obtained, a preset threshold for the duration of the siren is set, and it is determined whether the duration of the siren is greater than the preset threshold. If the duration of the siren is greater than the preset threshold, the vehicle intending to overtake is marked as an emergency vehicle.

[0011] Preferably, in S3, a second determination is made based on the surrounding vehicles' yielding status and the current road traffic rules to determine whether overtaking speed conditions are met. Specifically: S304: Obtain the minimum speed of surrounding vehicles when giving way, obtain the minimum speed limit requirement of the road at the current location of the unmanned vehicle, and determine whether the minimum speed of surrounding vehicles when giving way is lower than the minimum speed limit requirement of the road at the current location of the unmanned vehicle. If the minimum speed of surrounding vehicles when giving way is not lower than the minimum speed limit requirement of the road at the current location of the unmanned vehicle, the result of the second judgment is that there is no overtaking speed limit condition. If the minimum speed of surrounding vehicles when giving way is lower than the minimum speed limit requirement of the road at the current location of the unmanned vehicle, then execute S305. S305: Obtain the distance between the following vehicle and the unmanned vehicle, set a preset threshold for the safe speed-yielding distance, and determine whether the distance between the following vehicle and the unmanned vehicle is greater than the preset threshold for the safe speed-yielding distance. If the distance between the following vehicle and the unmanned vehicle is greater than the preset threshold for the safe speed-yielding distance, the secondary judgment result is that there is a condition for overtaking and yielding speed. If the distance between the following vehicle and the unmanned vehicle is less than or equal to the preset threshold for the safe speed-yielding distance, the secondary judgment result is that there is no condition for overtaking and yielding speed.

[0012] Preferably, in S3, the decision to yield is made based on the yielding status of surrounding vehicles and the yielding risk assessment score, specifically as follows: S306: Obtain images of surrounding vehicles avoiding the vehicle using the vehicle-mounted camera, determine whether the images show vehicles crossing solid lines, and determine whether the minimum speed of surrounding vehicles avoiding the vehicle is lower than the minimum speed limit of the road at the current location of the unmanned vehicle. If the images show vehicles crossing solid lines or the minimum speed of surrounding vehicles avoiding the vehicle is lower than the minimum speed limit of the road at the current location of the unmanned vehicle, then execute S307. If the images show vehicles crossing solid lines and the minimum speed of surrounding vehicles avoiding the vehicle is not lower than the minimum speed limit of the road at the current location of the unmanned vehicle, then the result is that the vehicle cannot give way. S307: Obtain the yielding safety risk score, determine whether the yielding safety risk score is 0. If the yielding safety risk score is 0, the result is that the yielding can be avoided. If the yielding safety risk score is not 0, the result is that the yielding cannot be avoided.

[0013] An artificial intelligence-based speed control system for autonomous driving includes the following modules: The overtaking and yielding speed judgment module obtains surrounding vehicle information through the vehicle-mounted camera and judges whether surrounding vehicles intend to overtake. If there are vehicles intending to overtake, it judges whether the conditions for overtaking and yielding speed are met based on the surrounding vehicle information and traffic rules. If the conditions for overtaking and yielding speed are met, the overtaking and yielding speed action is executed. If the conditions for overtaking and yielding speed are not met, the overtaking and yielding speed judgment module is executed. The overtaking and yielding judgment module uses GPS positioning to lock the location information of the unmanned vehicle, obtains the current road restriction rules from the unmanned vehicle location information, performs a risk assessment of overtaking and yielding based on the surrounding vehicle information and the current road restriction rules, and obtains a yielding risk assessment score. Based on the yielding risk assessment score, it determines whether the conditions for yielding and overtaking are met. If the conditions for overtaking and yielding are met, the overtaking and yielding action is executed. If the conditions for overtaking and yielding are not met, the special vehicle repeated judgment module is executed. The special vehicle repetitive judgment module obtains information on vehicles intending to overtake. Based on this information, it determines whether the vehicle intending to overtake is an emergency special vehicle. If it is, the module uses an onboard camera to obtain information on the yielding status of surrounding vehicles. It then uses this information, along with the current road rules, to make a second judgment on whether there are conditions for yielding when overtaking. If there are conditions for yielding when overtaking, the module executes the overtaking and yielding action. If there are still no conditions for yielding when overtaking, the module uses the yielding status of surrounding vehicles and the yielding risk assessment score to determine whether to yield and execute driving actions based on the judgment result.

[0014] (III) Beneficial Effects This invention provides a speed control method and system for autonomous driving based on artificial intelligence, which has the following beneficial effects: (1) This scheme obtains information about surrounding vehicles through vehicle-mounted cameras, and then judges whether surrounding vehicles intend to overtake based on the status of surrounding vehicles. For vehicles that intend to overtake, it judges whether the unmanned vehicle has the conditions to reduce its speed to facilitate the overtaking vehicle to overtake quickly. This facilitates the overtaking vehicle to overtake quickly. If the conditions for yielding speed are not met, it judges whether the conditions for yielding lanes to overtake are met by analyzing the road conditions and current road traffic rules. This further facilitates the unmanned vehicle to move laterally to the slow lane to give the overtaking vehicle sufficient space to move, avoiding the overtaking vehicle from scraping due to insufficient distance and causing traffic accidents during the overtaking process. This further improves the safety of unmanned vehicle driving. In this scheme, the logic of giving priority to yielding speed and then yielding lanes is used to avoid overtaking vehicles, thus facilitating the overtaking vehicle to overtake quickly while ensuring the driving safety of unmanned vehicle.

[0015] (2) This scheme identifies whether the vehicle intending to overtake is an emergency special vehicle, and re-determines the speed-yielding and lane-yielding conditions based on the identification results. In the secondary determination process, the safety of unmanned vehicles is the main condition. The unmanned vehicle can be judged whether it can briefly run over the solid line or slow down to provide overtaking convenience for the emergency special vehicle by the yielding status of surrounding vehicles. The yielding status of surrounding vehicles can avoid the negative impact caused by the temporary violation of the rules by adopting the herd effect. On the other hand, the yielding status of surrounding vehicles can also be used to re-identify the emergency special vehicle. At the same time, the yielding status of surrounding vehicles can be used to analyze whether the current state of the emergency special vehicle belongs to a special emergency state. Thus, the urgency of the matter can be judged based on the reaction of surrounding vehicles, which facilitates intelligent analysis and decision-making on the ethical dilemma of unmanned vehicles, and facilitates a more humane solution to the emergency situation of unmanned vehicles. Attached Figure Description

[0016] Figure 1 This is a flowchart of an artificial intelligence-based speed control method for autonomous driving according to the present invention. Figure 2 This is a schematic diagram of the module structure of an artificial intelligence-based speed control system for autonomous driving according to the present invention. Detailed Implementation

[0017] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0018] Please see Figures 1-2 This invention provides a speed control method for autonomous driving based on artificial intelligence, comprising the following steps: S1: Obtain information about surrounding vehicles through the vehicle-mounted camera, determine whether surrounding vehicles intend to overtake, if there are vehicles intending to overtake, determine whether there are conditions for overtaking and yielding based on the surrounding vehicle information and traffic rules, if there are conditions for overtaking and yielding, execute the overtaking and yielding action, if there are no conditions for overtaking and yielding, execute S2. S2: Locate the unmanned vehicle's position using GPS positioning, obtain the current road restriction rules based on the unmanned vehicle's position information, conduct a risk assessment of overtaking and yielding based on surrounding vehicle information and the current road restriction rules, obtain a yielding risk assessment score, determine whether the conditions for yielding and overtaking are met based on the yielding risk assessment score, if the conditions for overtaking and yielding are met, then execute the overtaking and yielding action; if the conditions for overtaking and yielding are not met, then execute S3. S3: Obtain information on vehicles intending to overtake. Based on this information, determine whether the vehicle intending to overtake is an emergency vehicle. If it is, obtain information on the yielding status of surrounding vehicles through the vehicle-mounted camera. Use this information in conjunction with the current road rules to determine whether there are conditions for yielding speed. If there are conditions, execute the yielding speed action. If there are still no conditions, use the yielding status of surrounding vehicles and the yielding risk assessment score to determine whether to yield. Execute driving actions based on the assessment results.

[0019] In this embodiment, the solution obtains information about surrounding vehicles through an in-vehicle camera, and then determines whether the surrounding vehicles intend to overtake based on their status. For vehicles that intend to overtake, it determines whether the unmanned vehicle has the conditions to reduce its speed to facilitate the overtaking vehicle's rapid overtaking. This facilitates the overtaking vehicle's rapid overtaking and ensures the driving safety of the unmanned vehicle, preventing the unmanned vehicle from accelerating during the overtaking process and causing overtaking failure and traffic accidents. This solution analyzes road conditions and current traffic rules to determine whether overtaking is possible when there is no opportunity to yield speed. This allows autonomous vehicles to move laterally into the slow lane, providing ample space for overtaking vehicles and preventing collisions caused by insufficient following distance. This further improves the safety of autonomous vehicles. The solution prioritizes yielding speed before yielding lanes to avoid overtaking vehicles, thus ensuring the safety of autonomous vehicles while facilitating rapid overtaking. This solution identifies whether a vehicle intending to overtake is an emergency vehicle and re-determines the yielding and lane-giving conditions based on the identification results. During this secondary determination, ensuring the safety of the unmanned vehicle is the primary consideration. The solution assesses whether the unmanned vehicle can briefly cross a solid line or slow down to facilitate overtaking by observing the yielding behavior of surrounding vehicles. This approach leverages the herd mentality to mitigate the negative impact of temporary rule violations and also serves as a secondary verification of the emergency vehicle's status. Furthermore, it analyzes the yielding behavior of surrounding vehicles to determine if the emergency vehicle's current situation constitutes a special emergency, thus assessing the urgency of the situation based on the reactions of surrounding vehicles. This facilitates intelligent analysis and decision-making regarding ethical dilemmas faced by unmanned vehicles, leading to more humane solutions for unexpected situations involving unmanned vehicles. In this scheme, regardless of whether the vehicle intending to overtake is an emergency vehicle, the risk assessment analysis of the yielding behavior of the unmanned vehicle is first performed based on the current road restriction rules and road environment. Based on the yielding risk assessment score, it is determined whether the vehicle can yield. After the vehicle intending to overtake is identified as an emergency vehicle, a second comprehensive judgment is made based on the yielding status of surrounding vehicles and the previously obtained risk assessment score, thus obtaining a second yielding judgment result. The unmanned vehicle then executes the command based on the second judgment result, which facilitates intelligent analysis and decision-making on the ethical dilemmas of unmanned vehicles, and thus facilitates a more humane solution to unmanned vehicle emergencies. It is worth mentioning that the weight values ​​in this scheme can be obtained through the analytic hierarchy process (AHP), and the preset threshold values ​​can be obtained through the weight analysis method. These will not be elaborated on further here.

[0020] In S1, information about surrounding vehicles is acquired through an onboard camera, and this information is used to determine whether any of the surrounding vehicles intend to overtake. Specifically: S101: Obtain the turn signal status of the following vehicle through the vehicle camera, determine whether the following vehicle has turned on its turn signal, if the following vehicle has turned on its turn signal, mark that the following vehicle has the intention to overtake, if the following vehicle has not turned on its turn signal, then execute S102. S102: Continuously acquire the distance between the following vehicle and the unmanned vehicle, set a preset distance threshold, and determine whether the distance between the following vehicle and the unmanned vehicle is less than the preset distance threshold. If the distance between the following vehicle and the unmanned vehicle is less than or equal to the preset distance threshold, then acquire the front wheel deflection angle of the following vehicle through the vehicle camera. S103: Set a preset threshold for the front wheel deflection angle of the following vehicle, determine whether the front wheel deflection angle of the following vehicle is greater than the preset threshold. If the front wheel deflection angle of the following vehicle is greater than the preset threshold, mark that the following vehicle has the intention to overtake, and record the overtaking direction according to the direction of the front wheel deflection angle of the following vehicle. If the front wheel deflection angle of the following vehicle is less than or equal to the preset threshold, repeat S101.

[0021] In this embodiment, the overtaking intention of the following vehicle is determined by whether the following vehicle has the intention to overtake. However, in real life, there may be situations where drivers forget to turn on their turn signals when overtaking. Therefore, by recognizing changes in the distance between vehicles and by observing changes in the front wheel deflection angle of the following vehicle, a comprehensive judgment can be made as to whether the following vehicle has the intention to overtake. Furthermore, the overtaking direction of the following vehicle can be determined based on the direction of the front wheel deflection angle and the direction of the turn signal, thereby facilitating the driving safety of unmanned vehicles.

[0022] In S1, if there is a vehicle intending to overtake, the system determines whether the conditions for yielding to overtake are met based on surrounding vehicle information and traffic rules. Specifically: S104: Obtain the minimum speed limit requirement of the road at the current location of the unmanned vehicle, set a preset threshold for safe speed, and obtain the minimum safe speed by summing the minimum speed limit requirement of the road at the current location of the unmanned vehicle with the preset threshold for safe speed. S105: Obtain the current speed of the unmanned vehicle and determine whether the current speed of the unmanned vehicle is greater than the minimum safe speed. If the current speed of the unmanned vehicle is greater than the minimum safe speed, then execute S106. If the current speed of the unmanned vehicle is equal to the minimum safe speed, then the result is that there is no overtaking speed limit. S106: Obtain the distance between the following vehicle and the unmanned vehicle, set a preset threshold for the safe speed-yielding distance, and determine whether the distance between the following vehicle and the unmanned vehicle is greater than the preset threshold for the safe speed-yielding distance. If the distance between the following vehicle and the unmanned vehicle is greater than the preset threshold for the safe speed-yielding distance, the result is that there is a condition for overtaking and yielding speed. If the distance between the following vehicle and the unmanned vehicle is less than or equal to the preset threshold for the safe speed-yielding distance, the result is that there is no condition for overtaking and yielding speed.

[0023] In this embodiment, by comparing the current vehicle speed with the minimum speed limit of the road, it is easier to determine whether there are conditions for yielding speed to overtake, so as to prioritize the safety of unmanned vehicles when yielding speed to overtake.

[0024] In S2, a risk assessment of overtaking and yielding is conducted based on surrounding vehicle information and current road regulations to obtain a yielding risk assessment score, specifically: S201: Obtain the current road restriction rules, determine whether yielding will cross the solid road line, if yielding will cross the solid road line, mark the risk of yielding and crossing the solid line as 1, if yielding will not cross the solid road line, mark the risk of yielding and crossing the solid line as 0. Determine whether yielding will reduce the vehicle speed below the current road minimum speed limit. If yielding will reduce the vehicle speed below the current road minimum speed limit, the risk of speed reduction is marked as 1. If yielding will not reduce the vehicle speed below the current road minimum speed limit, the risk of speed reduction is marked as 0. The risk of slowing down when yielding is summed with the risk of running over a solid line when yielding, resulting in the risk score of the yielding rule. S202: Obtain the speed of the vehicle giving way to the slow lane, obtain the current speed of the unmanned vehicle, obtain the speed difference by subtracting the current speed of the unmanned vehicle from the speed of the vehicle giving way to the slow lane, set a preset threshold for the speed difference, determine whether the speed difference is greater than the preset threshold, if the speed difference is greater than the preset threshold, mark the speed risk as 0, if the speed difference is less than or equal to the preset threshold, mark the speed risk as 1. Obtain the distance between the vehicle giving way to the slow lane and the unmanned vehicle, set a preset threshold for the distance between the vehicles in the slow lane, and determine whether the distance between the vehicle giving way to the slow lane and the unmanned vehicle is greater than the preset threshold. If the distance between the vehicle giving way to the slow lane and the unmanned vehicle is greater than the preset threshold, then mark the distance risk as 0. If the distance between the vehicle giving way to the slow lane and the unmanned vehicle is less than or equal to the preset threshold, then mark the distance risk as 1. The yield safety risk is obtained by summing the risks of vehicle spacing and vehicle speed. S203: Set the yielding rule risk score weight, and obtain the yielding rule risk score by multiplying the yielding rule risk score by the yielding rule risk score weight. Set the yielding safety risk score weight, and obtain the yielding safety risk score by multiplying the yielding safety risk score weight by the yielding safety risk score. Summing the yielding safety risk score and the yielding rule risk score yields the yielding risk assessment score.

[0025] In this embodiment, it is easy to determine whether yielding will violate traffic rules by judging whether yielding will cross the solid line or fall below the minimum speed limit. The speed and distance between the unmanned vehicle and the vehicle in the turning lane are used to determine whether yielding is safe. By fully considering the rules and safety, a yielding risk assessment score is obtained, which facilitates the subsequent judgment on whether the conditions for yielding are met. It is worth mentioning that in determining whether a vehicle will run over a solid line and whether it is safe to drive over vehicles in adjacent lanes, prediction is required. Specifically, the prediction results can be obtained through a Hidden Markov Model (HMM). HMM is an existing technology, and the specific operation steps will not be elaborated on here.

[0026] In S2, the condition for yielding and overtaking is determined based on the yielding risk assessment score. Specifically, the yielding risk assessment score is obtained, and it is determined whether the yielding risk assessment score is 0. If the yielding risk assessment score is 0, it is marked as having the condition for yielding; if the yielding risk assessment score is not 0, it is marked as not having the condition for yielding.

[0027] In S3, information about vehicles intending to overtake is obtained, and based on this information, it is determined whether the vehicle intending to overtake is an emergency or special vehicle. Specifically: S301: Obtain an image of a vehicle intending to overtake using an onboard camera, and identify whether the vehicle intending to overtake is a police car, ambulance, or fire truck based on the image. If the vehicle intending to overtake is not a police car, ambulance, or fire truck, the result is marked as a non-emergency special vehicle. If the vehicle intending to overtake is a police car, ambulance, or fire truck, then execute S302. S302: Acquire the surrounding sounds of the unmanned vehicle through the sound sensor, and use the Mel frequency cepstral coefficient to determine whether there is a siren sound in the surrounding sounds of the unmanned vehicle. If there is a siren sound in the surrounding sounds of the unmanned vehicle, lock the target vehicle through the siren sound, and determine whether the target vehicle is a vehicle intending to overtake. If the target vehicle is a vehicle intending to overtake, execute S303. If the target vehicle is not a vehicle intending to overtake, repeat S301. S303: The vehicle camera determines whether the vehicle preceding the vehicle intending to overtake has yielded to it in the past. If the vehicle preceding the vehicle intending to overtake has yielded to it in the past, the vehicle intending to overtake is defined as an emergency vehicle. If the vehicle preceding the vehicle intending to overtake has not yielded to it in the past, the duration of the siren is obtained, a preset threshold for the duration of the siren is set, and it is determined whether the duration of the siren is greater than the preset threshold. If the duration of the siren is greater than the preset threshold, the vehicle intending to overtake is marked as an emergency vehicle.

[0028] In this embodiment, the vehicle is first identified as an emergency special vehicle through visual image recognition, and then identified as an emergency special vehicle performing a mission through audio recognition of the siren sound. This helps to avoid identifying special vehicles without an emergency mission as emergency special vehicles, thus facilitating more intelligent and accurate identification of emergency special vehicles. The avoidance behavior of surrounding vehicles towards the overtaking vehicle can further corroborate whether the overtaking vehicle is an emergency special vehicle, and the urgency of avoidance can be judged by the duration of the siren sound.

[0029] In S3, the system uses a second assessment of surrounding vehicles' yielding behavior and current road regulations to determine whether overtaking speed conditions are met. Specifically: S304: Obtain the minimum speed of surrounding vehicles when giving way, obtain the minimum speed limit requirement of the road at the current location of the unmanned vehicle, and determine whether the minimum speed of surrounding vehicles when giving way is lower than the minimum speed limit requirement of the road at the current location of the unmanned vehicle. If the minimum speed of surrounding vehicles when giving way is not lower than the minimum speed limit requirement of the road at the current location of the unmanned vehicle, the result of the second judgment is that there is no overtaking speed limit condition. If the minimum speed of surrounding vehicles when giving way is lower than the minimum speed limit requirement of the road at the current location of the unmanned vehicle, then execute S305. S305: Obtain the distance between the following vehicle and the unmanned vehicle, set a preset threshold for the safe speed-yielding distance, and determine whether the distance between the following vehicle and the unmanned vehicle is greater than the preset threshold for the safe speed-yielding distance. If the distance between the following vehicle and the unmanned vehicle is greater than the preset threshold for the safe speed-yielding distance, the secondary judgment result is that there is a condition for overtaking and yielding speed. If the distance between the following vehicle and the unmanned vehicle is less than or equal to the preset threshold for the safe speed-yielding distance, the secondary judgment result is that there is no condition for overtaking and yielding speed.

[0030] In this embodiment, ensuring the safety of unmanned vehicles is the primary condition in the secondary judgment process. The decision on whether an unmanned vehicle can briefly cross a solid line or slow down to provide overtaking convenience for an emergency vehicle is determined by observing the yielding behavior of surrounding vehicles. On the one hand, the yielding behavior of surrounding vehicles can mitigate the negative impact of temporary rule violations through the bandwagon effect. On the other hand, the yielding behavior of surrounding vehicles can also be used to re-identify the emergency vehicle. Furthermore, the yielding behavior of surrounding vehicles can be used to analyze whether the current state of the emergency vehicle constitutes a special emergency. By judging the urgency of the situation based on the reactions of surrounding vehicles, it is easier to intelligently analyze and resolve ethical dilemmas related to unmanned vehicles, thereby facilitating a more humane solution to unmanned vehicle emergencies.

[0031] In S3, the decision to yield is determined based on the yielding status of surrounding vehicles and the yielding risk assessment score, specifically as follows: S306: Obtain images of surrounding vehicles avoiding the vehicle using the vehicle-mounted camera, determine whether the images show vehicles crossing solid lines, and determine whether the minimum speed of surrounding vehicles avoiding the vehicle is lower than the minimum speed limit of the road at the current location of the unmanned vehicle. If the images show vehicles crossing solid lines or the minimum speed of surrounding vehicles avoiding the vehicle is lower than the minimum speed limit of the road at the current location of the unmanned vehicle, then execute S307. If the images show vehicles crossing solid lines and the minimum speed of surrounding vehicles avoiding the vehicle is not lower than the minimum speed limit of the road at the current location of the unmanned vehicle, then the result is that the vehicle cannot give way. S307: Obtain the yielding safety risk score, determine whether the yielding safety risk score is 0. If the yielding safety risk score is 0, the result is that the yielding can be avoided. If the yielding safety risk score is not 0, the result is that the yielding cannot be avoided.

[0032] In this embodiment, while ensuring the avoidance of emergency special vehicles, priority is given to ensuring the safety of unmanned vehicles, thereby achieving an intelligent and humane solution to the ethical dilemma between traffic rules and emergency situations while ensuring the safety of traffic vehicles.

[0033] Please see Figures 1-2 This invention provides a speed control system for autonomous driving based on artificial intelligence, comprising the following modules: The overtaking and yielding speed judgment module obtains surrounding vehicle information through the vehicle-mounted camera and judges whether surrounding vehicles intend to overtake. If there are vehicles intending to overtake, it judges whether the conditions for overtaking and yielding speed are met based on the surrounding vehicle information and traffic rules. If the conditions for overtaking and yielding speed are met, the overtaking and yielding speed action is executed. If the conditions for overtaking and yielding speed are not met, the overtaking and yielding speed judgment module is executed. The overtaking and yielding judgment module uses GPS positioning to lock the location information of the unmanned vehicle, obtains the current road restriction rules from the unmanned vehicle location information, performs a risk assessment of overtaking and yielding based on the surrounding vehicle information and the current road restriction rules, and obtains a yielding risk assessment score. Based on the yielding risk assessment score, it determines whether the conditions for yielding and overtaking are met. If the conditions for overtaking and yielding are met, the overtaking and yielding action is executed. If the conditions for overtaking and yielding are not met, the special vehicle repeated judgment module is executed. The special vehicle repetitive judgment module obtains information on vehicles intending to overtake. Based on this information, it determines whether the vehicle intending to overtake is an emergency special vehicle. If it is, the module uses an onboard camera to obtain information on the yielding status of surrounding vehicles. It then uses this information, along with the current road rules, to make a second judgment on whether there are conditions for yielding when overtaking. If there are conditions for yielding when overtaking, the module executes the overtaking and yielding action. If there are still no conditions for yielding when overtaking, the module uses the yielding status of surrounding vehicles and the yielding risk assessment score to determine whether to yield and execute driving actions based on the judgment result.

[0034] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented in software, the above embodiments can be implemented, in whole or in part, as a computer program product. Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution.

[0035] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0036] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application.

Claims

1. A speed control method for autonomous driving based on artificial intelligence, characterized in that, Includes the following steps: S1: Obtain information about surrounding vehicles through the vehicle-mounted camera, determine whether surrounding vehicles intend to overtake, if there are vehicles intending to overtake, determine whether there are conditions for overtaking and yielding based on the surrounding vehicle information and traffic rules, if there are conditions for overtaking and yielding, execute the overtaking and yielding action, if there are no conditions for overtaking and yielding, execute S2. S2: Locate the unmanned vehicle's position using GPS positioning, obtain the current road restriction rules based on the unmanned vehicle's position information, conduct a risk assessment of overtaking and yielding based on surrounding vehicle information and the current road restriction rules, obtain a yielding risk assessment score, determine whether the conditions for yielding and overtaking are met based on the yielding risk assessment score, if the conditions for overtaking and yielding are met, then execute the overtaking and yielding action; if the conditions for overtaking and yielding are not met, then execute S3. S3: Obtain information on vehicles intending to overtake. Based on this information, determine whether the vehicle intending to overtake is an emergency vehicle. If it is, obtain information on the yielding status of surrounding vehicles through the vehicle-mounted camera. Use this information in conjunction with the current road rules to determine whether there are conditions for yielding speed. If there are conditions, execute the yielding speed action. If there are still no conditions, use the yielding status of surrounding vehicles and the yielding risk assessment score to determine whether to yield. Execute driving actions based on the assessment results.

2. The speed control method for autonomous driving based on artificial intelligence according to claim 1, characterized in that: In S1, information about surrounding vehicles is acquired through an onboard camera, and this information is used to determine whether any of the surrounding vehicles intend to overtake. Specifically: S101: Obtain the turn signal status of the following vehicle through the vehicle camera, determine whether the following vehicle has turned on its turn signal, if the following vehicle has turned on its turn signal, mark that the following vehicle has the intention to overtake, if the following vehicle has not turned on its turn signal, then execute S102. S102: Continuously acquire the distance between the following vehicle and the unmanned vehicle, set a preset distance threshold, and determine whether the distance between the following vehicle and the unmanned vehicle is less than the preset distance threshold. If the distance between the following vehicle and the unmanned vehicle is less than or equal to the preset distance threshold, then acquire the front wheel deflection angle of the following vehicle through the vehicle camera. S103: Set a preset threshold for the front wheel deflection angle of the following vehicle, determine whether the front wheel deflection angle of the following vehicle is greater than the preset threshold. If the front wheel deflection angle of the following vehicle is greater than the preset threshold, mark that the following vehicle has the intention to overtake, and record the overtaking direction according to the direction of the front wheel deflection angle of the following vehicle. If the front wheel deflection angle of the following vehicle is less than or equal to the preset threshold, repeat S101.

3. The speed control method for autonomous driving based on artificial intelligence according to claim 1, characterized in that: In S1, if a vehicle intends to overtake, the system determines whether the conditions for yielding to overtake are met based on surrounding vehicle information and traffic rules. Specifically: S104: Obtain the minimum speed limit requirement of the road at the current location of the unmanned vehicle, set a preset threshold for safe speed, and obtain the minimum safe speed by summing the minimum speed limit requirement of the road at the current location of the unmanned vehicle with the preset threshold for safe speed. S105: Obtain the current speed of the unmanned vehicle and determine whether the current speed of the unmanned vehicle is greater than the minimum safe speed. If the current speed of the unmanned vehicle is greater than the minimum safe speed, then execute S106. If the current speed of the unmanned vehicle is equal to the minimum safe speed, then the result is that there is no overtaking speed limit. S106: Obtain the distance between the following vehicle and the unmanned vehicle, set a preset threshold for the safe speed-yielding distance, and determine whether the distance between the following vehicle and the unmanned vehicle is greater than the preset threshold for the safe speed-yielding distance. If the distance between the following vehicle and the unmanned vehicle is greater than the preset threshold for the safe speed-yielding distance, the result is that there is a condition for overtaking and yielding speed. If the distance between the following vehicle and the unmanned vehicle is less than or equal to the preset threshold for the safe speed-yielding distance, the result is that there is no condition for overtaking and yielding speed.

4. The speed control method for autonomous driving based on artificial intelligence according to claim 1, characterized in that: In S2, a risk assessment of overtaking and yielding is conducted based on surrounding vehicle information and current road regulations to obtain a yielding risk assessment score, specifically: S201: Obtain the current road restriction rules, determine whether yielding will cross the solid road line, if yielding will cross the solid road line, mark the risk of yielding and crossing the solid line as 1, if yielding will not cross the solid road line, mark the risk of yielding and crossing the solid line as 0. Determine whether yielding will reduce the vehicle speed below the current road minimum speed limit. If yielding will reduce the vehicle speed below the current road minimum speed limit, the risk of speed reduction is marked as 1. If yielding will not reduce the vehicle speed below the current road minimum speed limit, the risk of speed reduction is marked as 0. The risk of slowing down when yielding is summed with the risk of running over a solid line when yielding, resulting in the risk score of the yielding rule. S202: Obtain the speed of the vehicle giving way to the slow lane, obtain the current speed of the unmanned vehicle, obtain the speed difference by subtracting the current speed of the unmanned vehicle from the speed of the vehicle giving way to the slow lane, set a preset threshold for the speed difference, determine whether the speed difference is greater than the preset threshold, if the speed difference is greater than the preset threshold, mark the speed risk as 0, if the speed difference is less than or equal to the preset threshold, mark the speed risk as 1. Obtain the distance between the vehicle giving way to the slow lane and the unmanned vehicle, set a preset threshold for the distance between the vehicles in the slow lane, and determine whether the distance between the vehicle giving way to the slow lane and the unmanned vehicle is greater than the preset threshold. If the distance between the vehicle giving way to the slow lane and the unmanned vehicle is greater than the preset threshold, then mark the distance risk as 0. If the distance between the vehicle giving way to the slow lane and the unmanned vehicle is less than or equal to the preset threshold, then mark the distance risk as 1. The yield safety risk is obtained by summing the risks of vehicle spacing and vehicle speed. S203: Set the yielding rule risk score weight, and obtain the yielding rule risk score by multiplying the yielding rule risk score by the yielding rule risk score weight. Set the yielding safety risk score weight, and obtain the yielding safety risk score by multiplying the yielding safety risk score weight by the yielding safety risk score. Summing the yielding safety risk score and the yielding rule risk score yields the yielding risk assessment score.

5. The speed control method for autonomous driving based on artificial intelligence according to claim 1, characterized in that: In S2, the condition for yielding and overtaking is determined based on the yielding risk assessment score. Specifically, the yielding risk assessment score is obtained, and it is determined whether the yielding risk assessment score is 0. If the yielding risk assessment score is 0, it is marked as having the condition for yielding; if the yielding risk assessment score is not 0, it is marked as not having the condition for yielding.

6. The speed control method for autonomous driving based on artificial intelligence according to claim 1, characterized in that: In S3, information about vehicles intending to overtake is obtained, and based on this information, it is determined whether the vehicle intending to overtake is an emergency or special vehicle. Specifically: S301: Obtain an image of a vehicle intending to overtake using an onboard camera, and identify whether the vehicle intending to overtake is a police car, ambulance, or fire truck based on the image. If the vehicle intending to overtake is not a police car, ambulance, or fire truck, the result is marked as a non-emergency special vehicle. If the vehicle intending to overtake is a police car, ambulance, or fire truck, then execute S302. S302: Acquire the surrounding sounds of the unmanned vehicle through the sound sensor, and use the Mel frequency cepstral coefficient to determine whether there is a siren sound in the surrounding sounds of the unmanned vehicle. If there is a siren sound in the surrounding sounds of the unmanned vehicle, lock the target vehicle through the siren sound, and determine whether the target vehicle is a vehicle intending to overtake. If the target vehicle is a vehicle intending to overtake, execute S303. If the target vehicle is not a vehicle intending to overtake, repeat S301. S303: The vehicle camera determines whether the vehicle preceding the vehicle intending to overtake has yielded to it in the past. If the vehicle preceding the vehicle intending to overtake has yielded to it in the past, the vehicle intending to overtake is defined as an emergency vehicle. If the vehicle preceding the vehicle intending to overtake has not yielded to it in the past, the duration of the siren is obtained, a preset threshold for the duration of the siren is set, and it is determined whether the duration of the siren is greater than the preset threshold. If the duration of the siren is greater than the preset threshold, the vehicle intending to overtake is marked as an emergency vehicle.

7. The speed control method for autonomous driving based on artificial intelligence according to claim 1, characterized in that: In S3, the system uses a second assessment of surrounding vehicles' yielding behavior and current road regulations to determine whether overtaking speed conditions are met. Specifically: S304: Obtain the minimum speed of surrounding vehicles when giving way, obtain the minimum speed limit requirement of the road at the current location of the unmanned vehicle, and determine whether the minimum speed of surrounding vehicles when giving way is lower than the minimum speed limit requirement of the road at the current location of the unmanned vehicle. If the minimum speed of surrounding vehicles when giving way is not lower than the minimum speed limit requirement of the road at the current location of the unmanned vehicle, the result of the second judgment is that there is no overtaking speed limit condition. If the minimum speed of surrounding vehicles when giving way is lower than the minimum speed limit requirement of the road at the current location of the unmanned vehicle, then execute S305. S305: Obtain the distance between the following vehicle and the unmanned vehicle, set a preset threshold for the safe speed-yielding distance, and determine whether the distance between the following vehicle and the unmanned vehicle is greater than the preset threshold for the safe speed-yielding distance. If the distance between the following vehicle and the unmanned vehicle is greater than the preset threshold for the safe speed-yielding distance, the secondary judgment result is that there is a condition for overtaking and yielding speed. If the distance between the following vehicle and the unmanned vehicle is less than or equal to the preset threshold for the safe speed-yielding distance, the secondary judgment result is that there is no condition for overtaking and yielding speed.

8. The speed control method for autonomous driving based on artificial intelligence according to claim 4, characterized in that: In S3, the decision to yield is determined based on the yielding status of surrounding vehicles and the yielding risk assessment score, specifically as follows: S306: Obtain images of surrounding vehicles avoiding the vehicle using the vehicle-mounted camera, determine whether the images show vehicles crossing solid lines, and determine whether the minimum speed of surrounding vehicles avoiding the vehicle is lower than the minimum speed limit of the road at the current location of the unmanned vehicle. If the images show vehicles crossing solid lines or the minimum speed of surrounding vehicles avoiding the vehicle is lower than the minimum speed limit of the road at the current location of the unmanned vehicle, then execute S307. If the images show vehicles crossing solid lines and the minimum speed of surrounding vehicles avoiding the vehicle is not lower than the minimum speed limit of the road at the current location of the unmanned vehicle, then the result is that the vehicle cannot give way. S307: Obtain the yielding safety risk score, determine whether the yielding safety risk score is 0. If the yielding safety risk score is 0, the result is that the yielding can be avoided. If the yielding safety risk score is not 0, the result is that the yielding cannot be avoided.

9. A speed control system for autonomous driving based on artificial intelligence, applied to the speed control method for autonomous driving based on artificial intelligence as described in any one of claims 1-8, characterized in that, Includes the following modules: The overtaking and yielding speed judgment module obtains surrounding vehicle information through the vehicle-mounted camera and judges whether surrounding vehicles intend to overtake. If there are vehicles intending to overtake, it judges whether the conditions for overtaking and yielding speed are met based on the surrounding vehicle information and traffic rules. If the conditions for overtaking and yielding speed are met, the overtaking and yielding speed action is executed. If the conditions for overtaking and yielding speed are not met, the overtaking and yielding speed judgment module is executed. The overtaking and yielding judgment module uses GPS positioning to lock the location information of the unmanned vehicle, obtains the current road restriction rules from the unmanned vehicle location information, performs a risk assessment of overtaking and yielding based on the surrounding vehicle information and the current road restriction rules, and obtains a yielding risk assessment score. Based on the yielding risk assessment score, it determines whether the conditions for yielding and overtaking are met. If the conditions for overtaking and yielding are met, the overtaking and yielding action is executed. If the conditions for overtaking and yielding are not met, the special vehicle repeated judgment module is executed. The special vehicle repetitive judgment module obtains information on vehicles intending to overtake. Based on this information, it determines whether the vehicle intending to overtake is an emergency special vehicle. If it is, the module uses an onboard camera to obtain information on the yielding status of surrounding vehicles. It then uses this information, along with the current road rules, to make a second judgment on whether there are conditions for yielding when overtaking. If there are conditions for yielding when overtaking, the module executes the overtaking and yielding action. If there are still no conditions for yielding when overtaking, the module uses the yielding status of surrounding vehicles and the yielding risk assessment score to determine whether to yield and execute driving actions based on the judgment result.

Images