Deadlock prevention and prevention of autonomous vehicles

By using sensors and control modules in autonomous vehicle systems to predict trajectory and identify conflicts, generating deadlock prevention notifications and coordination instructions, the problem of ADAS system deadlock at intersections is solved, achieving more efficient vehicle operation and better operation user experience.

CN118918719APending Publication Date: 2024-11-08GM GLOBAL TECHNOLOGY OPERATIONS LLC
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202310508388.5
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2023-05-06
Publication Date
2024-11-08

AI Technical Summary

Technical Problem

As the number of ADAS-equipped vehicles grows, multiple ADAS-equipped vehicles cannot move when intersections may cause deadlocks, resulting in traffic jams, and prior art is difficult to do without adding complexity and no standardization. Prevent deadlocks and improve ADAS functionality, redundancy and speed while maintaining vehicle operator comfort.

Method used

Adopt the autonomous vehicle system, which captures the vehicle status information and environmental information through sensors, uses the control module to execute programming logic, predict trajectories, identify multi-directional conflicts, transmits messages to cloud computing devices, generates deadlock prevention notifications and coordination instructions, and then sends them to the cloud computing device. The actuator implements ADAS function to avoid deadlocks at intersections.

Benefits of technology

Effectively prevent and prevent deadlocks at intersections, improve ADAS functionality, redundancy and speed, reduce pressure on vehicle operators, maintain or improve vehicle operator comfort without increasing complexity and standardization requirements These effects are achieved in the case.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118918719A_ABST
    Figure CN118918719A_ABST
Patent Text Reader

Abstract

A system for deadlock prevention and prevention in an autonomous vehicle (AV) includes an AV having an actuator and a sensor. An actuator changes the dynamic state of the AV, and a sensor captures vehicle state and environmental information. The system receives vehicle status information and predicts a main AV trajectory and a trajectory of other vehicles approaching, entering, or within an intersection. Multi-direction conflict identification (MDCIPD) messages and vehicle status information for each driving direction are transmitted to a cloud computing device. A wait relationship for preventing deadlock is based on the MDCIPD message and the vehicle status information, and a deadlock prevention notification (DPN) and / or a deadlock prevention coordination instruction (DPCI) is generated. The AV implements an advanced driver assistance system (ADAS) function via an actuator in response to the DPN and / or the DPCI and prevents the AV from deadlocking at the intersection.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present disclosure relates to vehicle control systems, and more particularly to an automated driver assistance system (ADAS) that automatically controls vehicle driving. Background Art

[0002] An increasing number of vehicles are being equipped with ADAS systems for controlling various vehicle dynamic functions. Some vehicles may have varying degrees of "hands-off" functionality, where the vehicle can operate without human input for varying periods of time, or until a specific condition or set of conditions occurs. Vehicles equipped with ADAS functionality require input from a variety of different sensors, and many vehicles also communicate with other vehicles through vehicle-to-vehicle (V2V) communication systems, or with infrastructure through vehicle-to-infrastructure (V2I) communication systems.

[0003] ADAS-equipped vehicles can handle a wide variety of driving situations. However, as the number of ADAS-equipped vehicles in operation continues to grow, new situations may arise where multiple ADAS-equipped vehicles are driving into each other in situations that could result in a deadlock, where the ADAS-equipped driving vehicles are effectively unable to move. Therefore, while current systems and methods for controlling vehicle functions via ADAS operate as intended, there is a need in the art for new and improved ADAS systems and methods that operate to allow ADAS-equipped driving vehicles to prevent deadlock and traffic jam situations without increasing complexity and without the need for standardization, while improving ADAS functionality, redundancy, and speed, and maintaining or enhancing vehicle operator comfort. Summary of the Invention

[0004] According to various aspects of the present disclosure, a system for deadlock prevention and control in autonomous vehicles includes one or more autonomous vehicles (AVs). The system also includes one or more actuators and one or more sensors. The one or more actuators selectively change the dynamic state of the one or more AVs. The one or more sensors are disposed on at least one of the one or more AVs and capture vehicle state information and information about the one or more AVs' surroundings. The system also includes one or more control modules. At least one of the one or more AVs having the one or more sensors is also equipped with at least one of the one or more control modules. Each of the one or more control modules includes a memory, a processor, and one or more input / output (I / O) ports. The memory stores programmed control logic. The processor executes the programmed control logic, and the I / O ports communicate with the one or more sensors and the one or more actuators. The programmed control logic includes at least first control logic, second control logic, third control logic, fourth control logic, fifth control logic, and sixth control logic. The first control logic receives vehicle state information. The second control logic predicts the trajectory of a host AV approaching an intersection on a road and predicts the trajectory of one or more other vehicles approaching, at, or within the intersection. The third control logic transmits a Multi-Directional Conflicts Identification Per Driving Direction (MDCIPD) message and vehicle status information for each driving direction to the cloud computing device. The fourth control logic determines a waiting relationship for deadlock prevention at the intersection based on the MDCIPD message and the vehicle status information. The fifth control logic transmits one or more of the following from the cloud computing device and is received by the master AV and one or more AVs: a Deadlock Precaution Notification (DPN) and a Deadlock Prevention Coordination Instruction (DPCI). The sixth control logic implements one or more advanced driver assistance system (ADAS) functions via one or more actuators in response to receiving one or more of the DPN and DPCI, and enables one or more AVs to avoid deadlock at the intersection.

[0005] In another aspect of the present disclosure, the first control logic further includes control logic for receiving a Basic Safety Message (BSM) from one or more other vehicles. The BSM includes vehicle size, vehicle type, vehicle speed, vehicle acceleration, vehicle position, vehicle heading, and vehicle turn signal status.

[0006] In yet another aspect of the present disclosure, the second control logic further includes control logic for obtaining trajectory-related data for each of the host AV and the one or more other vehicles approaching the intersection by combining vehicle state information, information about the host AV's surroundings, and BSMs from the one or more other vehicles. The second control logic further includes control logic for inferring the trajectory of each of the one or more vehicles and the host AV's trajectory from the trajectory-related data.

[0007] In another aspect of the present disclosure, the third control logic also includes control logic for determining MDCIPD messages, including predicting one or more conflicts between the master AV and the other vehicles given the current driving direction and vehicle status information of each of the master AV and the other vehicles.

[0008] In another aspect of the present disclosure, the fourth control logic also includes control logic for determining a waiting relationship for avoiding deadlock based on the driving direction, location, and time of the main AV and each of the other vehicles approaching, entering, or within the intersection, as well as the topology, laws, signs, and rules of the intersection.

[0009] In yet another aspect of the present disclosure, the waiting relationship defines a right-of-way calculation for each of the host AV and the other vehicles. The right-of-way calculation utilizes the driving direction, location, and time of each of the host AV and the other vehicles approaching, entering, or within the intersection, the topology and regulations of the intersection, local laws, and signs present at or near the intersection.

[0010] In another aspect of the present disclosure, the fifth control logic further includes control logic for generating a DPN based on the MDCI PD message. The fifth control logic further includes control logic for generating a DPCI. The DPN includes information directly indicating whether there is a possibility of deadlock. The DPCI includes information directly indicating the control actions required for each of the primary AV and any other AVs approaching, entering, or within the intersection to avoid deadlock at the intersection.

[0011] In yet another aspect of the present disclosure, the control logic for generating DPCI further includes control logic for selectively generating DPCI upon determining that two or more AVs capable of sending MDCI PD messages are at, entering, or within an intersection.

[0012] In yet another aspect of the present disclosure, the control logic for selectively generating the DPCI further includes control logic for causing the master AV to rely solely on the master AV's guidance system and ADAS functions to traverse the intersection when it is determined that there is no possibility of deadlock at the intersection. The control logic for selectively generating the DPCI further includes control logic for causing the master AV to rely solely on the master AV's guidance system and ADAS functions to avoid deadlock at the intersection when it is determined that there is a possibility of deadlock at the intersection based on the presence of at least one other vehicle at the intersection. The control logic for selectively generating the DPCI further includes control logic for causing the master AV and at least one other AV to coordinate their guidance systems and ADAS functions to jointly avoid and prevent deadlock at the intersection when it is determined that there is a possibility of deadlock at the intersection based on the presence of at least one other AV at the intersection.

[0013] In another aspect of the present disclosure, a method for deadlock prevention and control in autonomous vehicles includes utilizing one or more autonomous vehicles (AVs) and selectively changing the dynamic state of the one or more AVs using one or more actuators. The method also includes capturing vehicle state information and information about the surrounding environment of the one or more AVs using one or more sensors, the one or more sensors being disposed on at least one of the one or more AVs. The method also includes executing programmed control logic stored in a memory of one or more control modules, at least one of the one or more AVs having the one or more sensors being further equipped with at least one of the one or more control modules, each of the one or more control modules further having a processor and one or more input / output (I / O) ports, the processor executing the programmed control logic, and the I / O ports communicating with the one or more sensors and the one or more actuators. The programmed control logic receives the vehicle state information from the one or more sensors. The programmed control logic predicts a trajectory of a host AV approaching an intersection on a roadway and predicts trajectories of one or more other vehicles approaching, at, or within the intersection. The programmed control logic transmits a multi-directional conflict identification (MDCIPD) message for each driving direction and the vehicle state information to a cloud computing device. The programmable control logic determines a waiting relationship for preventing deadlock at an intersection based on the MDCIPD message and vehicle status information. The programmable control logic transmits one or more of a deadlock prevention notification (DPN) and a deadlock prevention coordination instruction (DPCI) from a cloud computing device and receives one or more of the deadlock prevention notifications (DPNs) and deadlock prevention coordination instructions (DPCIs) from a master AV among the one or more AVs. In response to receiving one or more of the DPNs and DPCIs, the programmable control logic implements one or more advanced driver assistance system (ADAS) functions via one or more actuators, thereby preventing the one or more AVs from deadlocking at the intersection.

[0014] In another aspect of the present disclosure, the method further includes receiving a basic safety message (BSM) from one or more other vehicles. The BSM includes vehicle size, vehicle type, vehicle speed, vehicle acceleration, vehicle position, vehicle heading, and vehicle turn signal status.

[0015] In yet another aspect of the present disclosure, the method further includes obtaining trajectory-related data for each of the host AV and the one or more other vehicles approaching the intersection by combining vehicle state information, information about the host AV's surroundings, and BSMs from the one or more other vehicles. The method further includes inferring a trajectory of each of the one or more vehicles and the host AV's trajectory from the trajectory-related data.

[0016] In yet another aspect of the present disclosure, the method further includes determining an MDCIPD message including predicting one or more conflicts between the host AV and the other vehicles given a current driving direction and vehicle state information for each of the host AV and the other vehicles.

[0017] In yet another aspect of the present disclosure, the method further includes determining a waiting relationship based on the driving direction, location, and time of the primary AV and each of the other vehicles approaching, entering, or within the intersection, as well as the topology, laws, signs, and regulations of the intersection to avoid deadlock.

[0018] In yet another aspect of the present disclosure, determining the waiting relationship further includes defining a right-of-way calculation for each of the host AV and the other vehicles. The right-of-way calculation utilizes the driving direction, location, and time of each of the host AV and the other vehicles approaching, entering, or within the intersection, as well as the topology and regulations of the intersection, local laws, and signs present at or near the intersection.

[0019] In another aspect of the present disclosure, the method further includes generating a DPN based on the MDCI PD message and generating a DPCI. The DPN includes information directly indicating whether there is a possibility of deadlock. The DPCI includes information directly indicating the control actions required by each of the host AV and any other AVs approaching, entering, or within the intersection to avoid deadlock at the intersection.

[0020] In yet another aspect of the present disclosure, generating the DPCI further includes selectively generating the DPCI upon determining that two or more AVs capable of sending MDCI PD messages are at, entering, or within an intersection.

[0021] In another aspect of the present disclosure, selectively generating the DPCI further includes: upon determining that there is no possibility of deadlock at the intersection, causing the host AV to traverse the intersection solely relying on its guidance system and ADAS functionality. Upon determining that there is a possibility of deadlock at the intersection based on the presence of at least one other vehicle at the intersection, causing the host AV to rely solely on its guidance system and ADAS functionality to avoid the deadlock at the intersection. Upon determining that there is a possibility of deadlock at the intersection based on the presence of at least one other AV at the intersection, causing the host AV and the at least one other AV to coordinate their guidance systems and ADAS functionality to jointly avoid and prevent the deadlock at the intersection.

[0022] In another aspect of the present disclosure, a method for deadlock prevention and control in autonomous vehicles includes utilizing one or more autonomous vehicles (AVs) and selectively changing the dynamic state of the one or more AVs using one or more actuators. The method also includes capturing vehicle state information and information about the environment surrounding the one or more AVs using one or more sensors, the one or more sensors being disposed on at least one of the one or more AVs. The method also includes executing programmed control logic stored in a memory of one or more control modules, wherein at least one of the one or more AVs having the one or more sensors is also equipped with at least one of the one or more control modules, each of the one or more control modules further comprising a processor and one or more input / output (I / O) ports, the processor executing the programmed control logic, and the I / O ports communicating with the one or more sensors and the one or more actuators. The programmed control logic receives vehicle state information from the one or more sensors, including receiving basic safety messages (BSMs) from one or more other vehicles. The BSMs include vehicle size, vehicle type, vehicle speed, vehicle acceleration, vehicle position, vehicle heading, and vehicle turn signal status. The programmed control logic also includes control logic for obtaining trajectory-related data for each of the one or more other vehicles and the host AV approaching the intersection by combining vehicle state information, information about the host AV's surroundings, and BSMs from the one or more other vehicles. The programmed control logic also includes control logic for predicting the trajectory of the host AV approaching the intersection on the road by inferring the trajectory of each of the one or more vehicles and the trajectory of the host AV from the trajectory-related data, and predicting the trajectories of one or more other vehicles approaching, at, or within the intersection. The programmed control logic also includes control logic for determining a multi-directional conflict identification (MDCIPD) message for each driving direction, including predicting one or more conflicts between the host AV and the other vehicles given the current driving direction and vehicle state information of each of the host AV and the other vehicles, and transmitting the MDCIPD message and the vehicle state information to a cloud computing device. The programmed control logic also includes control logic for determining a wait relationship from the MDCIPD messages and vehicle status information to prevent deadlock based on the driving direction, location, and time of each of the master AV and other vehicles approaching, entering, or within the intersection, as well as the topology, laws, signs, and regulations of the intersection. The wait relationship defines a right-of-way calculation for each of the master AV and other vehicles. The right-of-way calculation utilizes the driving direction, location, and time of each of the master AV and other vehicles approaching, entering, or within the intersection, as well as the topology and regulations of the intersection, local laws, and signs present at or near the intersection.The programmed control logic also includes control logic for generating one or more of a deadlock prevention notification (DPN) and a deadlock prevention coordination instruction (DPCI). The DPN is generated based on the MDCI PD message and includes information that directly indicates whether there is a possibility of deadlock. The DPCI includes information that directly indicates the control actions required for each of the master AV and any other AV approaching, entering, or within the intersection to avoid deadlock at the intersection. The programmed control logic also includes control logic for transmitting from a cloud computing device and receiving by the master AV and one or more AVs one or more of the DPN and DPCI. The programmed control logic also includes control logic for implementing one or more advanced driver assistance system (ADAS) functions via one or more actuators in response to receiving one or more of the DPN and DPCI, and causing the one or more AVs to avoid deadlock at the intersection.

[0023] In another aspect of the present disclosure, generating the DPCI further includes selectively generating the DPCI when it is determined that two or more AVs capable of sending MDCI PD messages are at, entering, or within an intersection. When it is determined that there is no possibility of deadlock at the intersection, the method causes the master AV to rely solely on the master AV's guidance system and ADAS functions to traverse the intersection. When it is determined that there is a possibility of deadlock at the intersection based on the presence of at least one other vehicle at the intersection, the method causes the master AV to rely solely on the master AV's guidance system and ADAS functions to avoid the deadlock at the intersection. When it is determined that there is a possibility of deadlock at the intersection based on the presence of at least one other AV at the intersection, the method causes the master AV and at least one other AV to coordinate their guidance systems and ADAS functions to jointly avoid and prevent the deadlock at the intersection.

[0024] Further areas of applicability will become apparent from the description provided herein.It should be understood that the description and specific examples are intended for purposes of illustration only and are not intended to limit the scope of the present disclosure. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] The drawings described herein are for illustration purposes only and are not intended to limit the scope of the present disclosure in any way.

[0026] Figure 1 is an environmental view of a system for deadlock prevention and prevention of an autonomous vehicle according to an aspect of the present disclosure;

[0027] Figure 2 is a diagram illustrating a method for performing a method according to one aspect of the present disclosure. Figure 1 A flowchart of a method for deadlock prevention and prevention of an autonomous vehicle; and

[0028] Figure 3 is a diagram illustrating a method for performing a method according to one aspect of the present disclosure. Figure 1 and Figure 2 A flow chart of an exemplary series of decisions made by an autonomous vehicle that applies the systems and methods for deadlock prevention and control when approaching an intersection. DETAILED DESCRIPTION

[0029] The following description is merely exemplary in nature and is not intended to limit the present disclosure, application, or uses.

[0030] Reference Figure 1 , a system 10 for deadlock prevention and control of a vehicle 12 is shown. More specifically, the illustrated vehicle 12 includes one or more autonomous vehicles (AVs) 12A-12D. However, it should be understood that while at least one of the vehicles 12A-12D is an AV, any one or more of the illustrated vehicles 12 may be an AV without departing from the scope or intent of the present disclosure. The AVs 12A-12D may be any of a variety of vehicles 12 without departing from the scope or intent of the present disclosure. In some examples, the AVs 12A-12D may be cars, trucks, sport utility vehicles (SUVs), buses, semi-tractors, tractors used in agriculture or construction, etc., watercraft, aircraft such as airplanes or helicopters, etc. The AVs 12A-12D are shown traveling at an intersection 16 on a road 14. The road has a plurality of lanes 18 defined by lane markings 20. The AVs 12A-12D may travel along any type of road 14 having any type of road surface, including any type of hard pavement, gravel, dirt, ice, or other types of surfaces.

[0031] The system 10 generally includes one or more sensors 22 in communication with one or more control modules 24. Although only a single sensor 22 is shown on each of the AVs 12A-12D, it should be understood that the illustrated sensor 22 is intended to represent any of a variety of sensors 22 that may be linked to and communicate on a communication network 26, including one or more of a vehicle-to-vehicle (V2V) communication network 28 and a vehicle-to-infrastructure (V2I) communication network 30. For simplicity, the term "infrastructure" is used to collectively refer to network entities that enable point-to-point (ad hoc) wireless communications in direct links or side links, as well as cellular communication networks owned by service providers, so that vehicle-to-network (V2N) communications are also included within the scope of V2I.

[0032] The sensors 22 can capture the AV 12's position information, electromagnetic, and / or optical information at various wavelengths (including wavelengths visible to humans, as well as infrared, ultraviolet, and other portions of the spectrum not visible to humans). Specifically, the sensors 22 can be global positioning system (GPS) sensors, cameras, LiDAR sensors, radar sensors, sonar sensors, ultrasonic sensors, inertial measurement units (IMUs), wheel speed sensors, or any of a variety of other sensors 22 that can be used to determine the dynamic state and position of the AV 12 relative to the environment without departing from the scope or intent of the present disclosure. The sensors 22 can be integrated with the AVs 12A-12D by the AV 12 manufacturer, or they can be installed as part of an after-sales service performed by the AV 12 manufacturer or sensor manufacturer, by a customer, or by another third party without departing from the scope or intent of the present disclosure. In some examples, one or more sensors 22 can be part of a system of standalone sensors 22 without departing from the scope or intent of the present disclosure.

[0033] Control module 24 is a non-general purpose electronic control device having: a pre-programmed digital computer or processor 32; a non-transitory computer-readable medium or memory 34 for storing data such as control logic, software applications, instructions, computer code, data, look-up tables, etc.; and one or more input / output (I / O) ports 36. Computer-readable medium or memory 34 includes any type of media that can be accessed by a computer, such as read-only memory (ROM), random-access memory (RAM), a hard drive, a compact disc (CD), a digital video disc (DVD), solid-state memory, or any other type of memory. "Non-transitory" computer-readable medium or memory 34 does not include wireless, optical, or other communication links that transmit electrical or other signals. Non-transitory computer-readable medium or memory 34 includes media that can permanently store data and media that can store data and later rewrite data, such as any type of program code (including source code, object code, and executable code). Processor 32 is configured to execute the code or instructions. In some examples, control module 24 may be a dedicated wireless or Wi-Fi controller. The I / O port 36 is configured to communicate via wired or wireless means, such as the communication network 26 , including the V2V communication network 28 and / or the V2I communication network 30 using Wi-Fi protocols under IEEE 802.11x, Bluetooth communication protocols, radio frequency (RF) protocols, etc.

[0034] In various aspects, the control module 24 includes one or more applications 38. An application 38 is a software program configured to perform a specific function or group of functions. The application 38 may include one or more computer programs, software components, instruction sets, procedures, functions, objects, classes, instances, related data, or portions thereof adapted to be implemented in suitable computer-readable program code. The application 38 may be stored in the memory 34, or in an additional or separate memory 34. Examples of the application 38 include audio or video streaming services, audio or visual processing services, deadlock prevention and prevention applications (DPA) 40, and the like.

[0035] In various aspects, the DPA 40 selectively utilizes an advanced driver assistance system (ADAS) 42 to alter the dynamics of the AV 12 via actuators of the vehicle dynamics system 44 of the AV 12. In some examples, the actuators of the vehicle dynamics system 44 may include a powertrain 46 that provides motive rotational force to one or more wheels 48 of the AV 12. The powertrain 46 may include an internal combustion engine, a transmission, a drive shaft, and / or one or more electric motors, or other such prime movers without departing from the scope or intent of the present disclosure. Additional actuators of the vehicle dynamics system 44 include one or more brakes 50 that operate to directly reduce, limit, or completely stop the rotation of the wheels 48 and / or the powertrain 46. The actuators of the vehicle dynamics system 44 may also include a steering system 52 that can change the angular orientation of one or more wheels 48 of the AV 12 relative to a longitudinal axis 54 of the AV 12. By changing the angular orientation of one or more wheels 48, the direction of travel of the AV 12 can be changed.

[0036] One or more of the vehicles 12A-12D is a connected autonomous vehicle (CAV) that communicates with at least one of the other vehicles 12A-12D and / or infrastructure such as traffic lights, cloud computing devices 56, etc. via a V2V communication network 28 and / or a V2I communication network 30. Figure 1As the vehicles 12 in the illustrated example approach the intersection 16, they may pass each other as shown. In manually operated vehicles, the vehicle operator determines when and how to proceed through the intersection 16 based on laws, signage, and the location and time of each vehicle 12 approaching, entering, or within the intersection 16. In contrast, in the AVs 12A-12D, the ADAS 42 utilizes data from the sensors 22 and data communicated via the V2V communication network 28 and / or the V2I communication network 30 to determine when and how each vehicle 12 should proceed through the intersection 16. In certain situations, such as when multiple vehicles 12 approach or arrive at the intersection 16 at substantially the same time, the ADAS 42 may have difficulty determining when and how each vehicle 12 should proceed. Thus, the system 10 of the present disclosure utilizes data communicated via the V2V communication network 28 and / or the V2I communication network 30 within the DPA 40 to avoid traffic jams, collisions, or other such traffic flow obstructions that may result from a deadlock between one or more AVs 12A-12D at the intersection 16. The data communicated via the V2V communication network 28 and / or the V2I communication network 30 may include a basic safety message (BSM) 58, a multi-directional conflict identification per driving direction (MDCIPD) message 60, a deadlock prevention notification (DPN) 62, and / or a deadlock prevention coordination instruction (DPCI) 64, each of which will be described in greater detail below.

[0037] Deadlock is a situation where no member of a group can move forward because each member is waiting for another member to take action before taking any action. Deadlock is common in multiprocessing systems, parallel computing systems, and distributed systems where systems such as ADAS use software and / or hardware locks to arbitrate shared resources and synchronize processes.

[0038] Now refer to Figure 2 And continue to refer to Figure 1 , which shows the DPA 40 in more detail in the form of a flow chart. More specifically, Figure 2 A deadlock prevention and control method utilizing the programmed control logic of the DPA 40 is described. In various aspects, the DPA 40 includes a plurality of subroutines or programmed control logic portions that enable the AVs 12A-12D to avoid deadlock situations. Figure 1 , the DPA 40 can be heuristically checked to verify whether certain steps can effectively serve the intended goals using a simple representation of the various communications sent over the communication network 26. As the AVs 12A-12D approach the intersection 16, one or more of the AVs 12A-12D activate the DPA 40 to avoid deadlock.

[0039] In a non-limiting example, the method of the DPA 40 begins at block 100 for a master AV 12A approaching an intersection 16. As the master AV 12A is about to enter the intersection 16, blocks 102 and 104 are executed to determine whether the master AV 12A can anticipate a multi-directional conflict with other vehicles 12 at the intersection 16. More specifically, at block 102, the master AV 12A communicates with other AVs 12B-12D approaching or in the intersection 16 via the V2V communication network 28 and obtains BSM 58 information from each of the other AVs 12B-12D communicating on the V2V communication network 28. The BSM 58 can include a variety of different types of data, but it should be understood that it is a standardized communication packet transmitted on a regular, periodic basis. In various aspects, the BSM 58 can be standardized by the Society of Automotive Engineers International or other technical organizations. In some examples, the regular, periodic basis can be every tenth of a second. The BSM 58 contains data regarding the status of the communicating vehicle 12, such as vehicle size and type, speed, acceleration, position, heading, turn signal status, and the like.

[0040] At block 104 , the host AV 12A obtains sensor data from each of the sensors 22 onboard the host AV 12A. The sensor data may include, but is not limited to, data from a GPS sensor, optical and / or electromagnetic information from cameras, LiDAR, SONAR, ultrasonic sensors, IMUs, wheel speed sensors, etc. Each of the other vehicles 12B-12D approaching the intersection 16 may present a potential collision risk to the host AV 12A.

[0041] Thus, at block 106, the sensor data and the BSM 58 data serve as inputs to a programmed control logic portion that predicts the trajectories of the other vehicles 12B-12D approaching the intersection 16. That is, based on the BSMs 58 from the other vehicles 12B-12D and the data from the onboard sensors 22 of the master AV 12A, the DPA 40 determines the probability that the master AV 12A will approach one or more of the other vehicles 12B-12D at the intersection 16 when the other vehicles 12B-12D are approaching from other directions. It should be understood that the precise distance between the master AV 12A and the other vehicles 12B-12D at the intersection 16 may vary significantly depending on the intersection 16, the type and size of the vehicle 12, and many other factors. However, a predetermined threshold safety distance may be used as a metric on which the trajectory-based proximity prediction is based.

[0042] If the DPA 40 determines that the probability that the master AV 12A and the other vehicles 12B-12D will be approaching is sufficiently high based on the trajectories of each of the master AV 12A and the other vehicles 12B-12D, and that the probability is less than a predetermined threshold, the DPA 40 will continue to adjust the driving function of the master AV 12A, and if one or more of the other vehicles 12B-12D are also equipped with ADAS 42 functionality and the DPA 40, the DPA 40 will adjust the driving function of the one or more other vehicles 12B-12D with the DPA 40.

[0043] In some examples, when the master AV 12A approaches an intersection, if two or more other vehicles 12B-12D are approaching the intersection 16 and their trajectory-related data indicates a likelihood of approaching less than a predetermined threshold, the master AV 12A may be considered to be in a multi-directional conflict. That is, the master AV 12A is expected to conflict with the other vehicles 12B-12D from two or more directions. In some examples, the other vehicles 12B-12D may be classified as potential impact vehicles, indicating that the other vehicles 12B-12D are already or about to enter the intersection 16 from a driving direction other than the direction of travel of the master AV 12A.

[0044] The trajectory-related data for each of the other vehicles 12B-12D may include position, speed, driving intent (i.e., driving straight through the intersection 16 or turning left or right at the intersection 16), path history, and any other such information that can be interpreted as relevant to the trajectory of the other vehicles 12B-12D. This data may come from the BSM 58 and / or the master AV 12A onboard perception functionality based on data from one or more sensors 22. When only one data source (i.e., only the BSM 58 or only the onboard sensors 22) is available, the DPA 40 may assign a lower confidence level to the trajectory-related data for each of the other vehicles 12B-12D. Conversely, when both the BSM 58 and onboard sensor 22 perception data are available to the master AV 12A and the DPA 40, a higher confidence level, greater than the lower confidence level, may be assigned to the trajectory-related data for the other vehicles 12B-12D. The confidence levels described above may be applied to the type of action taken by the master AV 12A in response to the trajectory-related data. When high confidence level data is available, right-of-way calculations for actuators of the vehicle dynamics system 44 and commands for ADAS 42 actuation may be performed at different magnitudes, speeds, or accelerations than when only lower confidence level data is available.

[0045] At block 108, the DPA 40 packages the MDCIPD message 60 including the predicted trajectories of the other vehicles 12B-12D approaching the intersection 16. For the master AV 12A, given the current driving direction and state information of the master AV 12A, the MDCIPD message 60 utilizes the predicted trajectories of the other vehicles 12B-12D approaching the intersection 16 and the trajectory of the master AV 12A to determine whether a conflict may exist between the master AV 12A and the other vehicles 12B-12D at the intersection 16. Additionally, at block 108, the DPA 40 sends the MDCIPD message 60 to the cloud computing device 56 for further processing. The MDCIPD message 60 may indicate whether the master AV 12A anticipates a multi-directional conflict, which may be useful in situations such as Figure 1 The MCDIPD message 60 is sufficient for use in most situations generally described in

[15] . In other examples, the MCDIPD message 60 may include additional information, such as the destination direction of the primary AV 12A, the original direction associated with the identified conflict, etc. The additional information may be used for additional programmed control logic functions of the DPA 40, such as determining the waiting relationship of all vehicles 12A-12D at the intersection 16, right-of-way calculations, deadlock prevention coordination strategies, etc.

[0046] At box 110, the cloud computing device 56 receives the MDCIPD message 60 from one or more vehicles 12 or AVs 12A-12D approaching the relevant intersection 16. At box 112, predetermined driving directions based on the topology and rules of the intersection 16 are retrieved from the sensors 22 and from the memory 34 of the control module 24 of the master AV 12A and the memory 34 of any other vehicles 12B-12D at the intersection 16 that happen to be AVs with DPA 40.

[0047] The MDCIPD message 60 and driving direction information from blocks 110 and 112 are then analyzed at block 114 to determine waiting relationships for the right-of-way and, if possible, to develop a deadlock prevention coordination strategy. That is, at block 114, the programmatic control logic is executed on the received MDCIPD message 60 along with the driving direction, topology, and / or rules of the intersection 16 to deduce whether a potential deadlock may occur at the intersection 16. Upon determining that a potential deadlock exists at the intersection 16, the programmatic control logic of the DPA 40 executed at block 114 derives a deadlock prevention coordination strategy. In the example where the master AV 12A is the only AV approaching the intersection 16 during the predetermined time window, the MDCIPD message 60 is received at block 110 only regarding the driving direction of the master AV 12A. Since the master AV 12A is the only vehicle 12 with a DPA 40 in this example, the DPN 62 to be sent directly indicates that no other vehicle 12 has sent an MDCI PD message 60, and thus no DPCI 64 message should be generated because coordination between the master AV 12A and the other vehicles 12B-12D cannot be achieved.

[0048] However, in the example where one or more of the other vehicles 12B-12D at the intersection 16 are AVs 12 utilizing the DPA 40, an MDCIPD message 60 is received regarding two or more originating directions of two or more AVs 12 at the intersection 16, and the DPN 62 to be sent directly indicates and includes information regarding multi-directional conflicts that can be anticipated based on all available originating directions of the vehicles 12. A deadlock prevention coordination strategy is determined following a potential deadlock based on the prioritization of the involved AVs 12A-12D at the intersection 16 and the wait cycles used for right-of-way calculations and timing.

[0049] At block 116 , the results of the analysis performed at block 114 are packaged into the DPN 62 and DPCI 64 (when two or more AVs with DPAs 40 are at the intersection 16 ) and sent to the relevant AVs at the intersection 16 . The DPA 40 then proceeds to block 118 .

[0050] At block 118, the master AV 12A receives the DPN 62 and DPCI 64 (when available) from the cloud computing device 56. At block 120, the master AV 12A takes action based on the received DPN 62 and DPCI 64 (when available). When the received DPN 62 indicates that no other AVs are present at the intersection 16 and no other AVs have sent MDCI PD messages 60, the master AV 12A relies on itself to avoid deadlock. That is, the master AV 12A will take action without instructions from the cloud computing device 56. In other examples, the master AV 12A may assume there is no potential deadlock and take action, or adjust its own path and / or motion plan based on the received DPCI 64 to prevent deadlock.

[0051] In various aspects, blocks 102-108, 118, and 120 are performed locally on the master AV 12A using the DPA 40, while blocks 110-116 are performed in and on the cloud computing device 56 in communication with the AV using the DPA 40. Furthermore, it should be understood that the DPA 40 can be executed continuously, periodically, or upon the occurrence of specific stimuli. In various aspects, the master AV 12A can execute the DPA 40 upon reaching the vicinity of any and all intersections 16, only certain predetermined intersections 16, and so forth, without departing from the scope or intent of the present disclosure.

[0052] Now refer to Figure 3 And continue to refer to Figure 1 and Figure 2, an example of simplified conditional branching corresponding to a simple MDCIPD message 60 and DPN 62 representation is shown in flowchart form. The example begins at block 200. At block 202, the master AV 12A determines whether one or more multi-directional conflicts are likely to occur. When the master AV 12A anticipates no multi-directional conflicts at the intersection 16 (e.g., when it is the only vehicle 12 present), there is no potential for deadlock, and the example proceeds to block 204, where the master AV 12A proceeds through the intersection 16, assuming there is no potential for deadlock. Conversely, if, at block 202, the master AV 12A determines that there is a potential for multi-directional conflicts with more than one vehicle 12 approaching the intersection 16, then, at block 206, the DPA 40 evaluates, based on the DPN 62, whether any other AVs are present at the intersection 16 and have sent MDCIPD messages 60. When no other AV has sent an MDCI PD message 60, the example proceeds to block 208, and the master AV 12A relies solely on its own guidance system, the DPA 40, and the ADAS 42 to navigate the intersection 16 and avoid deadlock. Conversely, when the received DPN 62 indicates that another AV is present at the intersection 16 and has sent an MDCI PD message 60, the example proceeds to block 210. At block 210, when the DPN 62 received by the master AV 12A indicates that other AVs with DPA 40 are expected to be in a multi-directional conflict, at block 212, the master AV 12A adjusts its path and motion plan based on the received DPCI 64 to avoid and prevent deadlock at the intersection 16. It should be understood that when other AVs are present at the intersection 16 and utilize the DPA 40, each of the AVs at the intersection 16 will adjust its path and motion plan based on the received DPCI 64 to avoid and prevent deadlock. In contrast, when the DPN 62 received by the master AV 12A indicates at block 210 that another AV having a DPA 40 is not expected to have a multi-directional conflict, the master AV 12A deems there is no possibility of deadlock and proceeds according to its own knowledge as previously described at block 204 .

[0053] The disclosed system 10 and method for deadlock prevention and control of autonomous vehicles 12 provide several advantages. These advantages include the ability to operate to allow a master AV 12A to mitigate the risk of a deadlock at an intersection 16 led by an autonomous vehicle, effectively preventing deadlocks in the presence of two or more V2V and V2I-enabled AVs, or at least providing preventative notification of potential deadlocks in the presence of only a single V2V and V2I-enabled AV. By incorporating vehicle-to-everything (V2X) messages (e.g., BSM 58, MDCIPD messages 60, DPN 62, and DPCI 64, which may comply with current standards or require no standardization) and implementing a cloud computing device 56 that implements lightweight computing, the disclosed system 10 and method 100 are highly resource-efficient, thereby improving ADAS functionality, redundancy, and speed without increasing complexity, while maintaining or enhancing vehicle operator comfort.

[0054] The description of the present disclosure is merely exemplary in nature, and variations that do not depart from the gist of the present disclosure are intended to fall within the scope of the present disclosure. Such variations should not be regarded as a departure from the spirit and scope of the present disclosure.

Claims

1. A system for deadlock prevention and prevention in an autonomous vehicle, the system comprising: One or more autonomous vehicles (AVs); one or more actuators that selectively change the dynamic state of the one or more AVs; one or more sensors disposed on at least one of the one or more AVs and capturing vehicle status information and information about the surrounding environment of the one or more AVs; One or more control modules, at least one of the one or more AVs having the one or more sensors is also equipped with at least one of the one or more control modules, each of the one or more control modules has a memory, a processor and one or more input / output I / O ports, the memory stores programmed control logic, the processor executes the programmed control logic, and the I / O port communicates with the one or more sensors and the one or more actuators; wherein the programmed control logic includes: A first control logic for receiving vehicle status information; second control logic for predicting a trajectory of a host AV approaching an intersection on a roadway and predicting trajectories of one or more other vehicles approaching, at, or within the intersection; A third control logic is used to transmit a multi-direction conflict identification MDCIPD message for each driving direction and the vehicle status information to a cloud computing device; a fourth control logic, based on the MDCIPD message and the vehicle status information, determining a waiting relationship at the intersection for preventing deadlock; a fifth control logic, one or more of a deadlock prevention notification DPN and a deadlock prevention coordination instruction DPCI transmitted by the cloud computing device and received by the master AV and the one or more AVs; and and a sixth control logic, in response to receiving one or more of the DPN and the DPCI, implementing one or more advanced driver assistance system (ADAS) functions via the one or more actuators and causing the one or more AVs to avoid deadlock at the intersection.

2. The system according to claim 1, wherein: The first control logic further includes: Control logic for receiving a basic safety message (BSM) from the one or more other vehicles, wherein the BSM comprises: Vehicle size, vehicle type, vehicle speed, vehicle acceleration, vehicle position, vehicle heading, and vehicle turn light status.

3. The system according to claim 2, wherein: The second control logic further includes: control logic for acquiring trajectory-related data for each of the one or more other vehicles and the host AV approaching the intersection by combining the vehicle state information, information about the host AV's surroundings, and the BSMs from the one or more other vehicles; and Control logic for inferring a trajectory of each of the one or more vehicles and the host AV from the trajectory-related data.

4. The system according to claim 1, wherein: The third control logic further includes: Control logic for determining the MDCIPD message includes predicting one or more conflicts between the host AV and the other vehicles given a current driving direction and vehicle state information for each of the host AV and the other vehicles.

5. The system according to claim 1, wherein: The fourth control logic further includes: Control logic for determining a deadlock-avoiding waiting relationship based on the driving direction, location, and time of the host AV and each of the other vehicles approaching, entering, or within the intersection, as well as the topology, laws, signs, and regulations of the intersection.

6. The system according to claim 5, wherein: The waiting relationship defines a right-of-way calculation for the host AV and each of the other vehicles, and wherein the right-of-way calculation utilizes the driving direction, location and time of the host AV and each of the other vehicles approaching, entering or within the intersection, the topology and rules of the intersection, local laws, and signs present at or near the intersection.

7. The system according to claim 1, wherein: The fifth control logic further includes: Control logic for generating the DPN from the MDCIPD message; and Control logic for generating the DPCI; wherein the DPN includes information directly indicating whether there is a possibility of deadlock; and Among other things, the DPCI includes information directly indicating control actions required by each of the host AV and any other AV approaching, entering, or within the intersection to avoid deadlock at the intersection.

8. The system according to claim 7, wherein: The control logic for generating the DPCI further includes: Control logic for selectively generating the DPCI upon determining that two or more AVs capable of sending the MDCI PD message are at, entering, or within the intersection.

9. The system according to claim 8, wherein: The control logic for selectively generating the DPCI further includes: for causing the master AV to rely solely on a guidance system and ADAS functions of the master AV to traverse the intersection when it is determined that there is no possibility of deadlock at the intersection; for causing the host AV to rely solely on the control logic of the guidance system and ADAS functions of the host AV to avoid deadlock at the intersection when it is determined that there is a possibility of deadlock at the intersection based on the presence of at least one other vehicle at the intersection; and Control logic for causing the host AV and at least one other AV to coordinate guidance systems and ADAS functions to jointly avoid and prevent deadlock at the intersection when it is determined that there is a possibility of deadlock at the intersection based on the presence of at least one other AV at the intersection.

10. A method for deadlock prevention and avoidance in an autonomous vehicle, the method comprising: utilizing one or more autonomous vehicles (AVs); selectively changing the dynamic state of the one or more AVs with one or more actuators; capturing vehicle status information and information about the surrounding environment of the one or more AVs using one or more sensors, the one or more sensors being disposed on at least one of the one or more AVs; executing programmed control logic stored in a memory of one or more control modules, wherein at least one of the one or more AVs having the one or more sensors is also equipped with at least one of the one or more control modules, each of the one or more control modules also having a processor and one or more input / output I / O ports, the processor executing the programmed control logic, and the I / O ports communicating with the one or more sensors and the one or more actuators; wherein the programmed control logic comprises: Receiving vehicle status information from the one or more sensors includes: receiving a basic safety message (BSM) from the one or more other vehicles, wherein the BSM comprises: vehicle size, vehicle type, vehicle speed, vehicle acceleration, vehicle position, Vehicle heading and vehicle turn signal status; acquiring trajectory-related data for each of the one or more other vehicles and the host AV approaching an intersection by combining the vehicle state information, information about the host AV's surroundings, and the BSMs from the one or more other vehicles; and predicting a trajectory of the host AV approaching the intersection on a road by inferring the trajectory of each of the one or more vehicles and the trajectory of the host AV from the trajectory-related data, and predicting trajectories of one or more other vehicles approaching, at, or within the intersection; determining a multi-directional conflict identification MDCIPD message for each driving direction, including predicting one or more conflicts between the host AV and each of the other vehicles given a current driving direction and vehicle status information of the host AV and each of the other vehicles, and transmitting the MDCIPD message and the vehicle status information to a cloud computing device; determining a waiting relationship for preventing deadlock based on the MDCIPD message and the vehicle status information based on the driving direction, location, and time of the host AV and each of the other vehicles approaching, entering, or within the intersection, and the topology, laws, signs, and regulations of the intersection, wherein the waiting relationship defines a right-of-way calculation for the host AV and each of the other vehicles, and wherein the right-of-way calculation utilizes the driving direction, location, and time of the host AV and each of the other vehicles approaching, entering, or within the intersection, the topology and regulations of the intersection, local laws, and signs present at or near the intersection; generating one or more of a deadlock prevention notification DPN and a deadlock prevention coordination instruction DPCI; wherein the DPN is generated according to the MDCIPD message and includes information directly indicating whether there is a possibility of deadlock; and wherein the DPCI includes information directly indicating control actions required by each of the master AV and any other AV approaching, entering, or within the intersection to avoid deadlock at the intersection; one or more of the DPN and the DPCI transmitted from the cloud computing device and received by the master AV and the one or more AVs; and One or more advanced driver assistance system (ADAS) functions are implemented via the one or more actuators in response to receiving one or more of the DPN and the DPCI, and the one or more AVs are caused to avoid deadlock at the intersection.