CN113032144B - Unmanned vehicle group forming method based on edge calculation under expressway scene and self-help model construction method - Google Patents
Unmanned vehicle group forming method based on edge calculation under expressway scene and self-help model construction method Download PDFInfo
- Publication number
- CN113032144B CN113032144B CN202110274590.7A CN202110274590A CN113032144B CN 113032144 B CN113032144 B CN 113032144B CN 202110274590 A CN202110274590 A CN 202110274590A CN 113032144 B CN113032144 B CN 113032144B
- Authority
- CN
- China
- Prior art keywords
- vehicle
- node
- unmanned vehicle
- nodes
- unmanned
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F9/00—Arrangements for program control, e.g. control units
- G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
- G06F9/46—Multiprogramming arrangements
- G06F9/50—Allocation of resources, e.g. of the central processing unit [CPU]
- G06F9/5061—Partitioning or combining of resources
- G06F9/5072—Grid computing
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/004—Artificial life, i.e. computing arrangements simulating life
- G06N3/006—Artificial life, i.e. computing arrangements simulating life based on simulated virtual individual or collective life forms, e.g. social simulations or particle swarm optimisation [PSO]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2111/00—Details relating to CAD techniques
- G06F2111/06—Multi-objective optimisation, e.g. Pareto optimisation using simulated annealing [SA], ant colony algorithms or genetic algorithms [GA]
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Software Systems (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Evolutionary Computation (AREA)
- Mathematical Physics (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- Health & Medical Sciences (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Geometry (AREA)
- Computer Hardware Design (AREA)
- Control Of Position, Course, Altitude, Or Attitude Of Moving Bodies (AREA)
Abstract
Disclosed are an unmanned vehicle group forming algorithm based on edge calculation and a self-coordination model construction method under a highway scene. Step 1, defining, step 2, forming an algorithm of the unmanned vehicle group, and step 3, constructing a self-help model of the unmanned vehicle group. The invention considers the problem that the open type unmanned vehicle group can not effectively keep stable and ordered under the interference of the unmanned vehicle nodes and roadside marks under the expressway background, and provides the unmanned vehicle group forming method of the layered multi-role nodes based on the edge calculation idea, so that tasks are distributed to different role nodes, the repeated calculation of the nodes is effectively avoided, and the cooperation between the corresponding nodes is kept; giving a self-help model of the unmanned vehicle group by considering the connectivity and stability of the unmanned vehicle group network; the unmanned vehicle group self-help model is optimized and solved by using a multi-objective optimization method, so that the optimal self-help model is obtained, and the reasonability and the usability of the unmanned vehicle group self-help model are effectively guaranteed.
Description
Technical Field
The invention relates to the field of unmanned driving, in particular to an edge calculation-based unmanned vehicle cluster self-help model construction method in a highway scene.
Background
The final aim of the automatic driving technology is to replace the work of a driver under any traffic condition, and the vehicle senses the change of the surrounding environment by using the sensing equipment carried by the vehicle and makes corresponding judgment on the change, so that the motion safety of the vehicle is ensured. The researchers have made relevant studies on the vehicle motion behavior at present as follows.
The automatic driving vehicle usually utilizes the sensing equipment carried by the automatic driving vehicle to acquire the surrounding road environment information, and on the basis, various motion planning algorithms are used for controlling the vehicle to run in the actual road environment. The main work of motion planning is a method for controlling a vehicle to avoid obstacles to travel all the way to a set destination on the basis of perceiving and understanding the current road environment. Among them, kiss D et al propose a geometric path planning method aimed at generating a high-quality path navigation with continuous curvature in a cluttered environment (containing narrow regions), the final generated path being similar to the path planning made by human drivers facing similar scenes, so as to make it better applicable to real vehicles. Guidolini R et al propose a Model Predictive Control method (n-mpc) Based on a Neural network to solve the delay problem of the steering device of the unmanned vehicle, and experimental results show that the n-mpc can greatly reduce the influence from the delay of the IARA steering device. Hiraishi H et al performed some experiments on the relationship between total efficiency (total efficiency) and diffusivity (diffusion rate) of an unmanned vehicle, and proposed a method based on passenger conditions to determine whether to generate an avoidance route according to the conditions of different passengers. Chen Z et al use a Convolutional Neural Network (CNN) model with raw image frames as inputs and output accordingly the optimal steering angle of the vehicle when facing a turn, which ensures that the vehicle is always driving on one lane without unnecessary lane changes. Bohren J et al propose a design based on a land based unmanned vehicle sensing, planning, navigation and drive system to enable vehicles to travel in a complex urban environment full of uncertainty. Chen Y L et al describes the overall system architecture of a racing vehicle (Terramax), the processing of sensors and sensor information, the mission planning system, and the autonomous behavior control of the vehicle. Patz B J et al discusses processing of sensor data (including data returned by radar), steering control of vehicles, and presents a practical method for efficient navigation of unmanned vehicles in urban environments. Anderson S J et al, on the basis of design and execution selection constraints, propose a new method applied to semi-autonomous unmanned vehicle hazard avoidance and stability control, and utilize a threat-based feedback mechanism to achieve an input limit that meets vehicle dynamics constraints to ensure that the vehicle stably avoids collisions. Bacha a et al employ a hybrid negotiation/reaction architecture to analyze the current situation and select appropriate behavior, enabling the vehicle to plan a safe path when confronted with intersections, road merging, automatic parking, and obstacle avoidance. Kala R et al propose a 4-layer planning algorithm, including road selection, path distribution and trajectory generation, which is helpful to realize the final road planning of unmanned vehicles in the case of complex obstacle networks. Urmson C et al combine neural network computational tasks, vehicle intelligent motion behaviors and trajectory motion planning to provide a three-layer planning system, use vehicle-mounted sensors to track other vehicles and detect static obstacles, and localize relative to a road model. Wherein the mission planning layer considers which candidate path to take to achieve the mission objective. The behavior layer decides when to change the lane and priority of the intersection and performs an error recovery operation. The motion planning layer generates action behaviors to avoid obstacles while stepping toward local targets. Ferguson D et al propose a motion planning framework for navigation of unmanned vehicles in urban environments, attempting to solve the problems of vehicle over-reliability, high-speed operation, complex workshop interaction, parking in large unstructured locations, and limited mobility. Likhachev Maxi and Ferguson Dave propose an algorithm based on arbitrary incremental search for multi-resolution, dynamically feasible grid states between unmanned vehicles traveling at high speeds over long distances. Montemerlo M et al propose an architecture model for an unmanned vehicle that can sense and interact with other vehicles to perform driving operations in various urban environments, including lane changing, turning around, parking, and road merging. Dolgov D et al use a modified a-algorithm to obtain feasible motion trajectories and improve the quality of the trajectories by numerical nonlinear optimization. Ziegler J et al use a notification search algorithm in combination with kinematic constraints of the vehicle and free-space topology of the vehicle to solve problems of accurate vehicle parking operations, narrow-lane turns, long-distance navigation, etc. Pivtoraiko M et al propose an algorithm for constraining path planning based on a special search space, which can generate an efficient path planner at a selected resolution. T M Howard et al propose an efficient state space sampling algorithm that enables high speed navigation in highly constrained and partially known environments. McNaughton M et al propose a motion planner for autonomous highway driving in a public road structured environment, transformed from a state grid framework for planetary rover navigation, allowing the search algorithm to be performed in real time, systematically and efficiently in the spatial and temporal dimensions. Gu T et al propose a novel double-deck movement planning system, solve the unmanned problem under city and highway these two kinds of circumstances under single frame, greatly optimized the calculated amount of system. Werling M et al propose a semi-reactive planning strategy that effectively accomplishes long-term maneuvering tasks for vehicles, such as lane changing, merging, distance keeping, speed keeping, precision stopping, etc., and can avoid collisions for a long period of time.
From the current state of research on the motion behavior of unmanned vehicle nodes at home and abroad, a method for planning the track and the path of the vehicle node is the direction in which the existing researchers put most attention. Although the above methods can solve the problems of the motion behaviors of the vehicle nodes to a certain extent, the methods rely on the vehicle nodes of the single intelligent agents to obtain and identify the surrounding environment information by the sensing equipment carried by the vehicle nodes, and then plan the motion behaviors, and no effective information circulation channel exists between the single intelligent agents. The method specifically comprises the following steps: unnecessary repeated calculation exists for the incidence relation of the surrounding vehicle nodes and the perception of the surrounding environment, the decision of a single intelligent agent is only made for the obstacles in the range of the sight distance of the single vehicle, the situation in the blind area of the sight field cannot be effectively responded, and the intelligent cooperation cannot be achieved, so that the wide application of the automatic driving technology is fundamentally restricted.
Disclosure of Invention
The invention aims to:
the highway road environment is single, and the external interference factor is small, so that a safe, stable and efficient application environment can be provided for unmanned driving. At present, research on unmanned motion behaviors in expressway scenes mainly focuses on single-agent motion behavior decision. There are the following problems: firstly, each single agent node needs to calculate the data sensed by the surrounding environment, and the calculation results cannot be shared mutually; secondly, the unmanned single intelligent agent relies on self independent perception and recognition, information interaction among vehicle nodes is lacked, and the intelligent decision of the motion behaviors has defects. The invention considers the problem that the open type unmanned vehicle group can not effectively keep stable and ordered under the interference of the manned vehicle nodes and roadside identification under the background of the highway, and provides the unmanned vehicle group forming method of the layered multi-role nodes (leading nodes, secondary leading nodes and common nodes) based on the edge computing thought, so that tasks are distributed to different role nodes, the repeated computing of the nodes is effectively avoided, and the cooperation between the corresponding nodes is kept; giving a self-help model of the unmanned vehicle group by considering the connectivity and stability of the unmanned vehicle group network; the unmanned vehicle group self-help model is optimized and solved by using a multi-objective optimization method, so that the optimal self-help model is obtained, and the reasonability and the usability of the unmanned vehicle group self-help model are effectively guaranteed. The problem is solved, and the self-help technology of the unmanned vehicle group can be practically applied in the expressway scene.
Therefore, the invention specifically provides the following technical scheme:
an edge calculation-based unmanned vehicle cluster forming algorithm in a highway scene is characterized by comprising the following steps:
step 1. Related definition
(1) Highway interference identification definition
Definition 1 highway Manned Vehicle Identification (MVI): the method comprises the following steps that a manned vehicle and an unmanned vehicle coexist on a highway, the manned vehicle is an obstacle moving at a high speed relative to the unmanned vehicle, in order to distinguish the manned vehicle and the unmanned vehicle, each vehicle node marks the identity of the vehicle node according to the state of the vehicle node by setting a highway manned vehicle mark, and the mathematical expression of the vehicle node is (1):
wherein, MVI represents the manned vehicle identification of vehicle v in the highway, when MVI v If =1, it means that the vehicle v is a manned vehicle and is used as an obstacle moving at a high speed in the subsequent algorithm; on the contrary, when MVI v =0, it indicates that the vehicle v is an unmanned vehicle;
definition 2 Highway Road sign (HRS, highway Road Signs): a plurality of road information signboard examples exist on the highway, and the road signboard can guide the movement of the unmanned vehicle;
(2) Edge computing based vehicle role and associated definitions
Definition 3 driverless vehicle Cluster lead node CH (Cluster Head): one unmanned vehicle group can select a unique unmanned vehicle node as a leading node of the whole vehicle group, and the leading node is responsible for communicating with leading nodes in other vehicle groups and managing various information of the vehicle group;
definition 4 driverless vehicle Cluster secondary guidance node CR (Cluster Relay): the secondary leading node CR determines whether the common node of the unmanned vehicle which is not the vehicle group can join the vehicle group or not;
defining 5 common nodes CM of the unmanned vehicle group: the unmanned vehicle cluster except the leading node CH and the secondary leading node CR belongs to common nodes, and all the unmanned vehicle nodes are common nodes before the vehicle cluster is formed;
definition of 6ch v : the pilotless vehicle cluster leading node CH corresponding to the vehicle node v is represented;
definition 7cr v : representing a secondary leading node CR of the unmanned vehicle group corresponding to the vehicle node v;
definition 8d max : indicates that in an unmanned vehicle group, the leading node CH of the unmanned vehicle group is the maximum distance from all vehicle nodes in the unmanned vehicle groupThe number of hops;
(3) Vehicle group attribute definition
Define 9 Driverless Vehicle connection Stability Index CSI (Driverless Vehicle Connectivity Stability Index): representing the stability of a direct connection between two unmanned vehicle nodes, the mathematical expression for which is (3):
wherein, MDCR (Max drive Communication Range) represents the maximum Communication Range of the Communication detection device carried by the unmanned vehicle, D t (v i ,v j ) Showing two unmanned vehicles v at time t i And v j The physical distance between; vehicle node v e { v | MVI v =0}; if the distance between the two unmanned vehicle nodes is smaller than or equal to the MDCR, namely the two unmanned vehicle nodes are located in the maximum communication range, the closer the vehicle nodes are, the larger the connection stability index CSI is, the higher the stability degree of the direct connection between the two vehicle nodes is, and the closer the connection between the two vehicle nodes is;
define 10 Neighbor Vehicle nodes NVN (Neighbor Vehicle Node): if a certain unmanned vehicle node v on the highway i With another vehicle node v j Between CSI satisfies CSI>Delta (delta is a set threshold value which ensures the reliability of the connection of the vehicle nodes), the connection between the two vehicles is stable, and the vehicle node v is called at the moment i And vehicle node v j The mutual neighbor node relationship is expressed as a node v in a network topological graph i And node v j There is an edge therebetween, and the mathematical expression thereof is (4):
wherein CSI represents an unmanned vehicle connection stability index, v i Indicating unmanned vehicles on the highway, v j Means that the vehicle on the highway is an unmanned vehicle or a manned vehicle;
definition 11 unmanned vehicle node v on a highway i Neighbor vehicle node set NVNS i，t : indicating a node v with a certain unmanned vehicle on a highway at time t i The vehicle set is composed of unmanned vehicle nodes in a neighbor node relation, and the mathematical expression of the vehicle set is (5):
NVNS i，t ＝{v j |NVN(v i ，v j )＝1and MVI v ＝0} (5)
wherein, NVN (v) i ，v j ) Representing unmanned vehicle node v i And v j Are a neighbor node relationship between v i Indicating unmanned vehicles on the highway, v j Means that the vehicle on the highway is an unmanned vehicle or a manned vehicle;
defining 12 a network topology DVN (V, E) composed of all unmanned vehicle nodes on the highway: DVN (DriverlessVehicle Network) is an undirected graph; where V represents the set of all unmanned vehicle nodes V = { V | MVI v =0, E denotes the set of all edges in the topological graph; v. of i Representing a certain unmanned vehicle node in the topological graph node set V;
defining 13 a connected component CCG (V, E) consisting of all unmanned vehicle nodes in the highway: the CCG (V, E) (Connected Component Graph) represents a greatly Connected subgraph in the network topology DVN, all vehicle nodes in the CCG (V, E) can be directly Connected with each other, and the mathematical expression is (6):
wherein the Arrive function is a discriminant function, where Arrive (u, v) = True indicates that there is a path that can connect the unmanned vehicle node u to the node v;
define 14 hyper-Connected Component HCCG (V, E) (High Connected Component Graph): obtained by adding edges to the connected component CCG (V, E), the mathematical expression is as follows (7):
HCCG(V，E)＝Addedge(u，v)toCCG(V，E)
wherein CCG (V, E) represents a connected component formed by all unmanned vehicle nodes on the expressway, a gap function is a discriminant function, and gap (u, V) = True represents that the distance between the node u and the node V is 2 hops;
define 15 dominating sets DS (DominatingSet): set V of partial points being hyperconnected components HCCG (V, E) * (ii) a The nodes in HCCG (V, E) are a set of vertices V * Or with a set of vertices V * One node in the set is adjacent, and the mathematical expression of the dominating set DS is (8):
define 16 a through path: defined as a path w = v composed of connected vehicle nodes in a highway network topology i e 0 v i+1 e 1 …v i+k e k v i+k+1 (ii) a The length of the path is k;
step 2. Unmanned vehicle group forming algorithm
Further comprises a step 3. The pilotless vehicle group self-help model construction method
Step 3.1 driverless vehicle group network connectivity
Step 3.2 stability of driverless vehicle fleet network
And 3.3, the unmanned vehicle group self-help model.
And 4, performing multi-objective optimization on the autonomous vehicle group self-help model.
Advantageous effects
The invention aims to disclose a method for guaranteeing the motion behavior of an unmanned vehicle group to be intelligent and to enable unmanned vehicles to be applied to an expressway scene, aiming at an open unmanned vehicle group driving environment in the expressway scene.
Description of the attached tables
TABLE 1 description of symbols in the present invention
TABLE 2 simulation experiment parameters
Drawings
FIG. 1 freeway scene schematic
FIG. 2 is a schematic view of a driverless vehicle cluster structure
FIG. 3 schematic representation of HCCG (V, E)
FIG. 4 unmanned vehicle cluster formation algorithm
FIG. 5 illustrates drone vehicle cluster connectivity
FIG. 6 is a schematic view of the stability of the unmanned vehicle cluster
FIG. 7 restriction of vehicle behavior by road conditions and traffic regulations
FIG. 8 comparison of average vehicle cluster switching times before and after auto-negotiation model optimization
FIG. 9 comparison of average residence time before and after optimization of the autocontrol model
FIG. 10 is a flow chart of the method of the present invention
Detailed Description
The specific implementation process of the invention is shown in fig. 10, and comprises the following 5 aspects:
(1) correlation definition
(2) Unmanned vehicle group forming algorithm
(3) Method for constructing self-help model of unmanned vehicle group
(4) Multi-objective optimization of self-help model of unmanned vehicle group
(5) Verification of simulation experiment
①
Correlation definition
The invention is based on main symbols required in the construction of the open unmanned vehicle group self-help model under the expressway scene, and the specific meaning description is shown in table 1.
In order to research the construction method of the open unmanned vehicle group self-help model based on the expressway scene, the relevant definitions are as follows:
(1) Highway interference identification definition
Definition 1 highway Manned Vehicle Identification (MVI): on a highway (the scene diagram is shown in fig. 1), a manned vehicle and an unmanned vehicle coexist, and the manned vehicle is an obstacle moving at a high speed relative to the unmanned vehicle. In order to distinguish the two, each vehicle node can mark the identity of the vehicle node by setting MVI according to the state of the vehicle node, and the mathematical expression is (1):
wherein, MVI represents the manned vehicle identification of vehicle v in the highway, when MVI v If =1, the vehicle v is a manned vehicle and serves as an obstacle moving at a high speed in the subsequent algorithm; on the contrary when MVI v If =0, it indicates that the vehicle v is an unmanned vehicle.
Definition 2 Highway Road sign (HRS, highway Road Signs): on an expressway (a scene diagram is shown in fig. 1), a plurality of road information signboards exist, such as expressway entrance and exit indication information, speed limit information and the like, and the road signboards can effectively guide the movement of the unmanned vehicle.
(2) Edge computing based vehicle roles and associated definitions
The unmanned vehicle cluster multi-role node based on edge calculation is shown in fig. 2.
Definition 3 driverless vehicle Cluster lead node CH (Cluster Head): one unmanned vehicle group can select a unique unmanned vehicle node as a leading node of the whole vehicle group, and the leading node is responsible for communicating with leading nodes in other vehicle groups and managing various information of the vehicle group, such as a vehicle group node set, positions and speeds of all nodes in the vehicle group and the like. To simplify the structure of the vehicle group, the maximum number of hops from the CH to any other node in the vehicle group will not exceed 2 hops.
Definition 4 driverless vehicle Cluster secondary guidance node CR (Cluster Relay): and the secondary leading node CR can determine whether the common node of the unmanned vehicle which is not the vehicle group can join the vehicle group or not.
Defining 5 common nodes CM of the unmanned vehicle group: the driverless vehicle cluster belongs to common nodes except for the leading node CH and the secondary leading node CR, and all driverless vehicle nodes are common nodes before the vehicle cluster is formed.
Definition of 6ch v : and represents the leading node CH of the driverless vehicle group corresponding to the vehicle node v.
Definition 7cr v : and represents the driverless vehicle cluster secondary lead node CR corresponding to the vehicle node v.
Definition 8d max : and the maximum hop count of the leading node CH of the unmanned vehicle group from all vehicle nodes in the unmanned vehicle group is shown.
(3) Vehicle group attribute definition
Define 9 Driverless Vehicle connection Stability Index CSI (drive Vehicle Connectivity Stability Index): representing the degree of stability of a direct connection between two unmanned vehicle nodes, the mathematical expression for which is (3):
wherein MDCR (Max driver Communication Range) represents an unmanned vehicleMaximum communication range of the carried communication probe device, D t (v i ,v j ) Representing two unmanned vehicles v at time t i And v j The physical distance between them. Vehicle node v e { v | MVI v =0}. If the distance between the two unmanned vehicle nodes is smaller than or equal to MDCR, namely the two unmanned vehicle nodes are located in the maximum communication range, the closer the vehicle nodes are, the larger the connection stability index CSI is, which indicates that the higher the stability degree of the direct connection between the two vehicle nodes is, and the closer the connection between the two vehicle nodes is.
Define 10 Neighbor Vehicle nodes NVN (Neighbor Vehicle Node): if a certain unmanned vehicle node v on the highway i With another vehicle node v j Between CSI satisfies CSI>Delta (delta is a set threshold value which ensures the reliability of the connection of the vehicle nodes), the connection between the two vehicles is stable, and the vehicle node v is called at the moment i And vehicle node v j The mutual neighbor node relationship is expressed as a node v in a network topological graph i And node v j There is an edge therebetween, and the mathematical expression thereof is (4):
wherein CSI represents an unmanned vehicle connection stability index, v i Indicating unmanned vehicles on the highway, v j Means that the vehicles on the highway, either unmanned or manned, are possible.
Definition 11 unmanned vehicle node v on a highway i Of neighboring vehicle node set NVNS i，t : indicating a node v with a certain unmanned vehicle on a highway at time t i The vehicle set is composed of unmanned vehicle nodes in a neighbor node relation, and the mathematical expression of the vehicle set is (5):
NVNS i，t ＝{v j |NVN(v i ，v j )＝1 and MVI v ＝0} (5)
wherein, NVN (v) i ，v j ) Representing unmanned vehicle node v i And v j Are a neighbor node relationship between v i Indicating unmanned vehicles on motorways, v j Means that the vehicles on the highway, may be unmanned vehicles, and may also be manned vehicles.
Defining 12 a network topology DVN (V, E) composed of all unmanned vehicle nodes on the highway: DVN (drive vertical Network) is an undirected graph; where V represents the set of all unmanned vehicle nodes V = { V | MVI v =0}, E denotes the set of all edges in the topology; v. of i And (4) representing a certain unmanned vehicle node in the topological graph node set V.
Defining 13 a connected component CCG (V, E) consisting of all unmanned vehicle nodes in the highway: the CCG (V, E) (Connected Component Graph) represents a greatly Connected subgraph in the network topology DVN, all vehicle nodes in the CCG (V, E) can be directly Connected with each other, and the mathematical expression is (6):
where the Arrive function is a discriminant function, where Arrive (u, v) = True indicates that there is a path that can connect the driverless vehicle node u to the node v.
Define 14 hyper-Connected Component HCCG (V, E) (High Connected Component Graph): obtained by adding edges to the connected component CCG (V, E) (the solid line of the double-headed arrow in fig. 3 is an edge newly added to the connected component CCG (V, E)), the mathematical expression is as follows (7):
HCCG(V，E)＝Add edge(u，v)to CCG(V，E)
wherein CCG (V, E) represents a connected component composed of all the nodes of the unmanned vehicle on the highway, the gap function is a discriminant function, and gap (u, V) = True represents that the distance between the node u and the node V is 2 hops.
Define 15 dominating sets DS (DominatingSet): set V of partial points being hyperconnected components HCCG (V, E) * . Nodes in HCCG (V, E) are either a set of vertices V * With one node in, or with a set of vertices V * One node in the set is adjacent, and the mathematical expression of the dominating set DS is (8):
define 16 through paths: is defined as a path w = v formed by connecting vehicle nodes in a highway network topology i e 0 v i+1 e 1 …v i+k e k v i+k+1 . The length of this path is then k.
②
The unmanned vehicle group forming algorithm is adopted,
specifically, as shown in algorithm 1, the flow chart is shown in fig. 4.
Algorithm 1 remarks description:
③
method for constructing self-help model of unmanned vehicle group
In order to keep the completeness of the unmanned vehicle group self-help system which can still keep the intelligent characteristics similar to the group in the external environment interference, the self-help criterion is followed, so that the unmanned vehicle group always keeps the intelligent behavior characteristics, the connectivity and the stability of the network of the unmanned vehicle group are considered, and the unmanned vehicle group self-help model is provided.
(1) Driverless vehicle group network connectivity
The network connectivity of the unmanned vehicle cluster is a scale for measuring the interconnection and intercommunication capacity of nodes in the vehicle cluster and is an important part of the unmanned self-coordination model. In the pilotless vehicle group, frequent information interaction is required between the common nodes of the pilotless vehicle group and the pilotless secondary leading nodes and between the pilotless secondary leading nodes and the pilotless vehicle group leading nodes, so that the communication capacity between the vehicle nodes in the pilotless vehicle group needs to be measured.
Fig. 5 is a schematic diagram of connectivity of an unmanned vehicle group on a highway, and it can be seen from the diagram that the unmanned vehicle group in the left diagram has more through paths, and even if some links in a network are disconnected due to an emergency, vehicle nodes in the unmanned vehicle group can still maintain better connectivity. The density and redundancy of the network reflect the fault tolerance of the network after the links between the facing nodes are disconnected, namely, the network has stronger connectivity. It can be concluded therefrom that the network connectivity of the unmanned vehicle cluster depends on the number of through-paths in the vehicle cluster. But it is difficult to calculate the number of all the through paths in the large-scale network, so the invention uses the ratio of the number of edges in the vehicle group network to the number of vehicle nodes to measure the consistence of the vehicle group network. The unmanned vehicle group network connectivity formula is (16):
wherein the content of the first and second substances,
showing connectivity of the vehicle group network at time t, and V showing unmanned vehicle group at time t
A middle vehicle node set, E represents an unmanned vehicle group at the time t
A set of edges in a network.
The larger the value is, the higher the network connectivity of the vehicle group is.
(2) Driverless vehicle cluster network stability
In the process of moving the unmanned vehicle group on the highway, due to uncertain interference of external environment, the connection of nodes in the unmanned vehicle group is easy to change, which may cause obstacles to information interaction and cause the unmanned vehicle group to be in a chaotic and disordered state. It is therefore necessary to measure the stability of the network connection within the unmanned vehicle fleet.
Fig. 6 is a schematic diagram of stability of an unmanned vehicle cluster on a highway. The numbers in the figure represent the driverless vehicle connection stability index CSI, the greater the connection stability index CSI, the more stable the direct connection between two driverless vehicle nodes is, the closer the connection between the two is. As can be seen from the figure, the connectivity of the unmanned vehicle group in the left figure is stronger than that of the vehicle group in the right figure according to the definition of the network connectivity of the unmanned vehicle group, but since the CSI value of the edges between the vehicle nodes in the vehicle group in the right figure is 1.56 and is larger than the CSI value of the edges between the vehicle nodes in the vehicle group in the left figure by 1.38, the closer the distance between the vehicle nodes of the unmanned vehicle group in the right figure is shown, the higher the stability degree of the direct connection between the vehicle nodes is, and therefore, the higher the stability of the unmanned vehicle group in the right figure can be inferred. The network stability is measured by adopting an average vehicle connection stability index of the unmanned vehicle group network, and the mathematical expression is (17):
wherein the content of the first and second substances,
indicating the vehicle group network at time tStability of the following. E denotes the unmanned vehicle group at time t
A set of edges in a network.
The larger the value is, the higher the stability of the unmanned vehicle group network is.
(3) Self-help model of unmanned vehicle group
In order to keep the completeness of the unmanned vehicle group self-help system which can still keep the intelligent characteristics similar to the group in the interference of the unmanned vehicle group in the external environment, the self-help criterion is followed, so that the unmanned vehicle group always keeps the intelligent behavior characteristics, a self-help model of the unmanned vehicle group is provided, the network interconnection and intercommunication are realized, and the stable state is kept, wherein the mathematical expression of the self-help model is (18):
wherein, the first and the second end of the pipe are connected with each other,
indicating unmanned vehicle group at time t
The larger the value of the network connectivity of the unmanned vehicle group indicates that the network connectivity of the current unmanned vehicle group is stronger,
indicating unmanned vehicle group at time t
The larger the value of the network stability of (2) indicates that the network stability of the current unmanned vehicle group is stronger. Co th And (4) representing a network connectivity threshold value, and setting according to the specific environment requirement of the highway network. The network connectivity is the bar that the pilotless vehicle crowd self-help model needs to satisfy firstWhen the network connectivity value of the vehicle group is larger than the threshold value, the self-help model of the unmanned vehicle group can operate.
④
Multi-objective optimization of self-help model of unmanned vehicle group
The driving behavior of the unmanned vehicle node on the expressway is a process constrained by external environmental conditions such as road conditions and traffic regulations, as shown in fig. 7. The figure shows that the minimum vehicle safety distance between the unmanned vehicles A and B must be kept to avoid rear-end collisions; when the unmanned vehicle C finds that the vehicle node in front of the unmanned vehicle C has a fault, the motion behavior of the unmanned vehicle C needs to be changed into deceleration, stop or detour; the driverless vehicle D finds that the vehicle is about to reach the destination ahead, and the vehicle must be decelerated, and the like, the road conditions and the traffic regulations restrict the movement of the vehicle. The general motion behavior of the unmanned vehicle nodes in the expressway scene is considered, the unmanned self-help model is bound to keep the interior of the vehicle group stable and orderly under the condition that the constraint conditions are met, the network connectivity and the network stability of the unmanned vehicle group are considered, the unmanned vehicle group self-help model is optimized and solved by using a multi-objective optimization method, and the optimal self-help model is obtained, so that the reasonability and the usability of the unmanned vehicle group self-help model are ensured.
(1) Constraint conditions
1) Vehicle dynamics constraints
An unmanned vehicle moving on a highway usually only considers one direction, and a common driving method of the vehicle shows that the speed and the acceleration of the vehicle can only continuously change along with time within a certain limited range, and the mathematical expression of the node acceleration of the vehicle is (19):
wherein C = [ -a, a ] represents a variation interval of the vehicle acceleration, which interval depends on the vehicle's own motion performance. If the position, the self speed and the acceleration of a certain vehicle node i on the expressway are known at the time t, the position, the self speed and the acceleration of the vehicle node i on the expressway can be calculated at the next time as follows:
wherein, the first and the second end of the pipe are connected with each other,
the above equation shows the dynamic behavior of the vehicle node. Based on the unidirectional direction of the expressway, once the relevant information of the vehicle node is known, the motion situation of the vehicle node at the next moment can be calculated. And the mobility of each unmanned vehicle node is limited by other vehicles, particularly its own front and rear vehicles. Assuming that vehicle node j is before vehicle node i and vehicle node k is after vehicle node i on the highway, then there are the following speed constraints for vehicle node i:
2) Traffic regulation constraints
In an expressway scenario, road restriction rules are usually required, such as specifying maximum vehicle speeds, minimum vehicle speeds, and safe vehicle distances between vehicles for different lanes, and the like, and the mathematical expression thereof is:
wherein omega t The driving range of the unmanned vehicle is represented, and c represents a vehicle speed interval allowed by a vehicle node, namely, the vehicle node is allowed to drive at the speed in the interval in the area, wherein the influence of a traffic signal lamp, a road signboard such as an access road junction and the like is considered. dis (x) i (t),x j (t)) represents the distance between two vehicle nodes, D ij Represents the minimum safe distance between any two vehicle nodes, i.e., represents a constraint on vehicle behavior.
(2) Model optimization
For the optimization problem of the autonomous model of the unmanned vehicle group in the expressway scene, a plurality of objective functions in the autonomous model can be decomposed into a plurality of independent sub-problems to be solved, the optimization solution of each sub-problem uses the information of the adjacent sub-problem, and therefore the optimization solution of the model can be completed after each sub-problem is independently optimized and solved. In order to meet constraint conditions in the running process of the unmanned vehicle group self-help model and ensure network connectivity and stability of the unmanned vehicle group self-help model, the unmanned vehicle group self-help model is rapidly optimized in a highway scene by using a multi-objective optimization method, so that the network connectivity and stability objective function of the vehicle group tends to be optimal, and the communication among vehicle nodes is kept stable. At the moment, the pilotless vehicle group self-help model can be converted into:
wherein the content of the first and second substances,
indicates the unmanned vehicle group at the time t
The larger the value of the network connectivity of the unmanned vehicle group, the stronger the network connectivity of the current unmanned vehicle group,
indicates the unmanned vehicle group at the time t
The larger the value of the network stability of (2) indicates that the network stability of the current unmanned vehicle group is stronger. When the temperature is higher than the set temperature
And with
When the self-coordinated model is respectively larger than the corresponding threshold values, the self-coordinated model can reach an optimal state under the condition of meeting the following constraint conditions.
⑤
Simulation experiment verification
The invention realizes the self-help model of the unmanned vehicle group in the expressway scene by adopting a simulation experiment mode so as to further verify the rationality and the effectiveness of the self-help model of the unmanned vehicle group.
(1) Experimental data
In order to verify the self-help model of the unmanned vehicle group, based on the highway scene, SUMO simulation software is adopted to simulate a one-way two-lane highway scene with the length of ten kilometers, an entrance is arranged at each kilometer in the road, 100 unmanned vehicles are randomly arranged, the vehicle speed obeys uniform distribution in the arrangement interval, and an exit in the highway is randomly selected to leave the road. The specific experimental simulation parameters are shown in table 2, and the simulation time is set to 300 seconds. The range in which the unmanned vehicle can guarantee stable connection in network simulation is set to be 200 meters.
(2) Simulation and test result analysis
In order to verify the reasonability and the availability of the model optimization method, the method analyzes and compares various indexes of an unmanned vehicle group on the expressway before and after model optimization in different vehicle speed intervals, and specifically comprises the following steps: average train group switching times and average stay time.
The average population switch times represents the average number of times the unmanned vehicle node switches the population during its lifecycle. Due to the fact that the variability of highway road environments comprises interference of human driving and the like, the unmanned vehicles can generate dynamic evolution events, and then the unmanned vehicles are switched from one unmanned vehicle group i to another unmanned vehicle group j. If the average switching times of the nodes of the unmanned vehicle are large, the self-help model is not stable. The simulation experiment researches the influence of the vehicle speed on the average vehicle group switching times of the unmanned vehicle, different speed intervals are set, the average vehicle group switching times of the self-help model before and after comparison and optimization are calculated under different vehicle speeds, and the experiment result is shown in fig. 8.
Fig. 8 shows a comparison of average vehicle group switching times before and after optimization of the unmanned autonomous model, wherein the average vehicle group switching times become larger as the vehicle speed increases, which is caused by the fact that the vehicle speed difference between unmanned vehicles increases and the node role state changes more frequently due to the increase of the vehicle speed. The average train switching times of the optimized unmanned self-help model is lower than the value before optimization, which proves that the optimized unmanned self-help model is more stable.
The average stay time represents the time for which the unmanned vehicle stays in a certain unmanned vehicle group i in the moving process of the expressway. The unmanned vehicle nodes are bound to be switched among vehicle groups due to various dynamic evolution events in the running process of the unmanned vehicle nodes, and the short average vehicle stay time means that the unmanned vehicles are switched among the vehicle groups more frequently, which indicates that the unmanned self-help model is not stable. The simulation experiment researches the influence of the vehicle speed on the average stay time of the unmanned vehicle node, different speed intervals are set, the average stay time of the self-help model vehicle node before and after comparison and optimization is calculated under different vehicle speeds, and the experiment result is shown in fig. 9.
It can be seen that the average stay time value is continuously reduced with the increase of the vehicle speed, which is caused by the increase of the vehicle speed, the increase of the vehicle speed difference between the unmanned vehicles and the more frequent switching of the vehicle node and the vehicle group. The average stay time of the optimized unmanned self-help model is longer than the value before optimization, which proves that the optimized unmanned self-help model is more stable.
Innovation point
The technical scheme provided by the application of the invention is as follows: based on the edge calculation idea, relevant definitions of unmanned layered multi-role nodes are given, an unmanned vehicle group forming method is given, connectivity and stability of an unmanned vehicle group network are considered, an unmanned vehicle group self-cooperation model of the unmanned vehicle group in an external interference factor environment is given, and the constructed unmanned vehicle group self-cooperation model is optimally solved by utilizing multi-objective optimization. The method aims to construct a self-coordination model that the motion behaviors of the unmanned vehicle group are always kept intelligent under the condition of interference of external factors in the highway scene, so that the unmanned vehicle can be applied to the highway scene.
Attached table of the specification
TABLE 1
TABLE 2
Claims (3)
1. A method for forming an unmanned vehicle group in an expressway scene, wherein an unmanned single agent moves in the expressway scene, is characterized in that: the method specifically comprises the following steps:
step 1. Related definition
(1) Highway interference signature definition
Definition 1 highway Manned Vehicle Identification (MVI): the method comprises the following steps that a manned vehicle and an unmanned vehicle coexist on a highway, the manned vehicle is an obstacle moving at a high speed relative to the unmanned vehicle, in order to distinguish the manned vehicle and the unmanned vehicle, each vehicle node marks the identity of the vehicle node according to the state of the vehicle node by setting a highway manned vehicle mark, and the mathematical expression of the vehicle node is (1):
wherein, MVI represents the manned vehicle identification of vehicle v in the highway, when MVI v If =1, it means that the vehicle v is a manned vehicle and is used as an obstacle moving at a high speed in the subsequent algorithm; on the contrary, when MVI v If =0, it indicates that the vehicle v is an unmanned vehicle;
definition 2 Highway Road sign (HRS, highway Road Signs): a plurality of road information signboard examples exist on the expressway, and the road signboard can guide the movement of the unmanned vehicle;
(2) Edge computing based vehicle roles and associated definitions
Definition 3 driverless vehicle Cluster leading node CH (Cluster Head): one unmanned vehicle group can select a unique unmanned vehicle node as a leading node of the whole vehicle group, and the leading node is responsible for communicating with leading nodes in other vehicle groups and managing various information of the vehicle group;
definition 4 driverless vehicle Cluster secondary guidance node CR (Cluster Relay): the secondary leading node CR is responsible for mutually transmitting messages between the common nodes of the vehicle cluster and the leading nodes of the vehicle cluster and managing the common nodes of the vehicle cluster, namely determining whether the common nodes of the unmanned vehicles which are not the vehicle cluster can be added into the vehicle cluster or not;
defining 5 common nodes CM of the unmanned vehicle group: the unmanned vehicle cluster except the leading node CH and the secondary leading node CR belongs to common nodes, and all the unmanned vehicle nodes are common nodes before the vehicle cluster is formed;
definition of 6ch v : the pilotless vehicle cluster leading node CH corresponding to the vehicle node v is represented;
definition 7cr v : the pilot node CR represents a secondary pilot node CR of the unmanned vehicle group corresponding to the vehicle node v;
definition 8d max : the maximum hop count of the leading node CH of the unmanned vehicle cluster from all vehicle nodes in the unmanned vehicle cluster is shown in the unmanned vehicle cluster;
(3) Vehicle group attribute definition
Define 9 Driverless Vehicle connection Stability Index CSI (Driverless Vehicle Connectivity Stability Index): representing the stability of a direct connection between two unmanned vehicle nodes, the mathematical expression for which is (3):
wherein, MDCR (Max driver Communication Range) represents the maximum Communication Range of the Communication detection device carried by the unmanned vehicle, D t (v i ，v j ) Representing two unmanned vehicles v at time t i And v j The physical distance therebetween; vehicle node v ∈{v|MVI v =0}; if the distance between the two unmanned vehicle nodes is smaller than or equal to the MDCR, namely the two unmanned vehicle nodes are located in the maximum communication range, the closer the vehicle nodes are, the larger the connection stability index CSI is, the higher the stability degree of the direct connection between the two vehicle nodes is, and the closer the connection between the two vehicle nodes is;
define 10 Neighbor Vehicle nodes NVN (Neighbor Vehicle Node): if a certain unmanned vehicle node v on the highway i With another vehicle node v j The CSI between the two vehicles meets the condition that the CSI is larger than delta (delta is a set threshold value to ensure the reliability of the connection of the vehicle nodes), the connection between the two vehicles is stable, and the vehicle node v is called at the moment i And vehicle node v j The mutual neighbor node relationship is expressed as a node v in a network topological graph i And node v j There is an edge therebetween, and the mathematical expression thereof is (4):
wherein CSI represents an unmanned vehicle connection stability index, v i Indicating unmanned vehicles on the highway, v j Means that the vehicle on the highway is an unmanned vehicle or a manned vehicle;
definition 11 driverless vehicle node v on highway i Neighbor vehicle node set NVNS i，t : showing a node v with a certain unmanned vehicle on the highway at time t i The vehicle set composed of unmanned vehicle nodes in a neighbor node relationship comprises the following mathematical expression (5):
NVNS i，t ＝{v j |NVN(v i ，v j )＝1 and MVI v ＝0} (5)
wherein, NVN (v) i ，v j ) Representing unmanned vehicle node v i And v j Is a neighbor node relationship between v i Indicating unmanned vehicles on the highway, v j Indicating vehicles on freewaysA vehicle, which is an unmanned vehicle or a manned vehicle;
defining 12 a network topology DVN (V, E) composed of all unmanned vehicle nodes on the highway: DVN (drive vertical Network) is an undirected graph; where V represents the set of all unmanned vehicle nodes V = { V | MVI v =0, E denotes the set of all edges in the topological graph; v. of i Representing a certain unmanned vehicle node in the topological graph node set V;
defining 13 a connected component CCG (V, E) consisting of all unmanned vehicle nodes in the highway: the CCG (V, E) (Connected Component Graph) represents a greatly Connected subgraph in the network topology DVN, all vehicle nodes in the CCG (V, E) can be directly Connected with each other, and the mathematical expression is (6):
wherein the Arrive function is a discriminant function, where Arrive (u, v) = True indicates that there is a path that can connect the unmanned vehicle node u to the node v;
define 14 hyper-Connected Component HCCG (V, E) (High Connected Component Graph): obtained by adding edges to the connected component CCG (V, E), the mathematical expression is as follows (7):
HCCG(V，E)＝Add edge(u，v)to CCG(V，E)
wherein CCG (V, E) represents a connected component formed by all unmanned vehicle nodes on the highway, the gap function is a discriminant function, and gap (u, V) = True represents that the distance between the node u and the node V is 2 hops;
define 15 Dominating sets DS (domining Set): set V of partial points being hyperconnected components HCCG (V, E) * (ii) a The nodes in HCCG (V, E) are a set of vertices V * Or with a set of vertices V * One node in (1) is adjacent to, and supportsThe match set DS is expressed mathematically as (8):
define 16 a through path: is defined as a path w = v formed by connecting vehicle nodes in a highway network topology i e 0 v i+ 1 e 1 …v i+k e k v i+k+1 (ii) a The length of the path is k;
step 2. Formation method of unmanned vehicle group
2. The method of claim 1, further comprising: also comprises a step 3. The method for constructing the self-help model of the unmanned vehicle group
Step 3.1 driverless vehicle group network connectivity
Step 3.2 stability of driverless vehicle fleet network
And 3.3, the pilotless vehicle group self-help model.
3. The method of claim 1, further comprising: and 4, performing multi-objective optimization on the autonomous driving vehicle group model.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110274590.7A CN113032144B (en) | 2021-03-15 | 2021-03-15 | Unmanned vehicle group forming method based on edge calculation under expressway scene and self-help model construction method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110274590.7A CN113032144B (en) | 2021-03-15 | 2021-03-15 | Unmanned vehicle group forming method based on edge calculation under expressway scene and self-help model construction method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113032144A CN113032144A (en) | 2021-06-25 |
| CN113032144B true CN113032144B (en) | 2022-11-11 |
Family
ID=76468837
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110274590.7A Active CN113032144B (en) | 2021-03-15 | 2021-03-15 | Unmanned vehicle group forming method based on edge calculation under expressway scene and self-help model construction method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113032144B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113742909B (en) * | 2021-08-24 | 2023-11-03 | 同济大学 | Method for constructing unmanned vehicle group self-cooperative model in closed scene |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN204425397U (en) * | 2014-11-27 | 2015-06-24 | 华东师范大学 | A kind of urban service network of pilotless automobile |
| CN108682145A (en) * | 2018-05-31 | 2018-10-19 | 大连理工大学 | The grouping method of unmanned bus |
| CN109353409A (en) * | 2018-12-06 | 2019-02-19 | 吉林大学 | A kind of steering system based on commercial vehicle formation traveling |
| CN109447547A (en) * | 2018-10-12 | 2019-03-08 | 吉林大学 | A kind of freight transport system and its shipping method based on unmanned son, female vehicle |
| CN109979188A (en) * | 2019-04-04 | 2019-07-05 | 吉林大学 | A kind of unmanned sightseeing fleet of intelligence |
| CN112163720A (en) * | 2020-10-22 | 2021-01-01 | 哈尔滨工程大学 | Multi-agent unmanned electric vehicle battery replacement scheduling method based on Internet of vehicles |
-
2021
- 2021-03-15 CN CN202110274590.7A patent/CN113032144B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN204425397U (en) * | 2014-11-27 | 2015-06-24 | 华东师范大学 | A kind of urban service network of pilotless automobile |
| CN108682145A (en) * | 2018-05-31 | 2018-10-19 | 大连理工大学 | The grouping method of unmanned bus |
| CN109447547A (en) * | 2018-10-12 | 2019-03-08 | 吉林大学 | A kind of freight transport system and its shipping method based on unmanned son, female vehicle |
| CN109353409A (en) * | 2018-12-06 | 2019-02-19 | 吉林大学 | A kind of steering system based on commercial vehicle formation traveling |
| CN109979188A (en) * | 2019-04-04 | 2019-07-05 | 吉林大学 | A kind of unmanned sightseeing fleet of intelligence |
| CN112163720A (en) * | 2020-10-22 | 2021-01-01 | 哈尔滨工程大学 | Multi-agent unmanned electric vehicle battery replacement scheduling method based on Internet of vehicles |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113032144A (en) | 2021-06-25 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| JP7214017B2 (en) | Joint control of vehicles traveling on different crossroads | |
| Ajanovic et al. | Search-based optimal motion planning for automated driving | |
| Sharma et al. | Recent advances in motion and behavior planning techniques for software architecture of autonomous vehicles: A state-of-the-art survey | |
| Bai et al. | Hybrid reinforcement learning-based eco-driving strategy for connected and automated vehicles at signalized intersections | |
| Zhan et al. | Spatially-partitioned environmental representation and planning architecture for on-road autonomous driving | |
| Baskar et al. | Hierarchical traffic control and management with intelligent vehicles | |
| US20220340177A1 (en) | Systems and methods for cooperative driving of connected autonomous vehicles in smart cities using responsibility-sensitive safety rules | |
| Pan et al. | Research on the behavior decision of connected and autonomous vehicle at the unsignalized intersection | |
| CN111899509B (en) | Intelligent networking automobile state vector calculation method based on vehicle-road information coupling | |
| CN113032144B (en) | Unmanned vehicle group forming method based on edge calculation under expressway scene and self-help model construction method | |
| Gu et al. | Safe-state enhancement method for autonomous driving via direct hierarchical reinforcement learning | |
| Baby et al. | A suggestion-based fuel efficient control framework for connected and automated vehicles in heterogeneous urban traffic | |
| Wang et al. | Motion planning in complex urban environments: An industrial application on autonomous last‐mile delivery vehicles | |
| Guo et al. | Self-defensive coordinated maneuvering of an intelligent vehicle platoon in mixed traffic | |
| CN114093166A (en) | Centralized hybrid vehicle coordination optimization method for signal-lamp-free fork | |
| Ramaswamy et al. | Lane assignment on automated highway systems | |
| CN115220461B (en) | Robot single system and multi-robot interaction cooperation method in indoor complex environment | |
| Kala et al. | Multi-vehicle planning using RRT-connect | |
| Shiomi et al. | A lane-change maneuver of automated vehicles for improving traffic flow on highways with multiple lanes | |
| Pauca et al. | Trajectory planning and tracking for cooperative automated vehicles in a platoon | |
| Naumann | Probabilistic motion planning for automated vehicles | |
| Zhang et al. | A Hierarchical Multi-Vehicle Coordinated Motion Planning Method based on Interactive Spatio-Temporal Corridors | |
| Zhang et al. | Decision-making and planning framework with prediction-guided strategy tree search algorithm for uncontrolled intersections | |
| Duhautbout et al. | Generic trajectory planning algorithm for urban autonomous driving | |
| Kala et al. | Congestion avoidance in city traffic |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |