Digital twin enabled autonomous vehicle post-incident repair method

Abstract

The application discloses a digital twin-enabled unmanned vehicle post-accident maintenance method, and mainly solves the low vehicle accident processing rate of the prior art.The implementation scheme is as follows: a task model and a network model under a digital twin network are constructed; maintenance information of each maintenance provider is published; a maintenance response strategy of the unmanned vehicle after an accident at time t is established; a maintenance strategy of each unmanned vehicle at each network edge device is determined; a bid of each network edge device for the unmanned vehicle is determined; matching between the network edge device and the unmanned vehicle is completed; a maintenance price to be paid by the unmanned vehicle is determined according to a VCG pricing mechanism; and the price is paid after the maintenance is completed.The application effectively solves the two-way selection problem between the accident vehicle and the maintenance provider, improves the vehicle accident processing rate, the average income of the maintenance provider and the average social welfare, and can be used for maintenance processing of the unmanned vehicle after an accident in a city traffic system.

Description

Technical Field

[0001] This invention belongs to the field of vehicle networking technology, and specifically relates to a method for post-accident repair of unmanned vehicles, which can be used in intelligent transportation systems. Background Technology

[0002] As an important component of intelligent transportation systems, autonomous vehicles typically integrate environmental perception units, computing units, and motion control units. They can not only make safe and accurate driving decisions autonomously, but also sell computing resources to other autonomous vehicles with limited resources, thereby improving traffic safety and efficiency. However, autonomous vehicles controlled by intelligent algorithms may cause accidents in complex and ever-changing traffic conditions due to problems with hardware or recognition algorithms.

[0003] Traditional accident handling mechanisms utilize advanced sensor networks and image analysis models for accident prevention and prediction. By sensing road scenes in real time and identifying potential hazards, vehicles can make effective response decisions. They also leverage advanced communication technologies for post-accident processing, rapidly identifying the location and scale of accidents and responding promptly to reduce the environmental impact of autonomous vehicle accidents. However, traditional mechanisms lack consideration for the impact on autonomous vehicles during an accident and their post-accident benefits within the network, hindering efficient maintenance services. Furthermore, digital twin technology can effectively reduce the waste of communication resources caused by frequent information exchanges between autonomous vehicles and maintenance providers. Therefore, in digital twin networks, a key challenge lies in how autonomous vehicles comprehensively consider the impact and benefits of accidents to select the optimal accident response plan, and how maintenance providers choose a set of autonomous vehicles for maintenance services.

[0004] Patent document CN110047170A discloses a "Method for Emergency Management and Roadside Assistance on Dedicated Lanes for Autonomous Driving," which provides functions such as routing assistance, intelligent judgment of emergency severity, emergency rescue process design, and control of unmanned vehicles and accident emergency vehicles. Unmanned vehicle accidents include vehicle malfunctions and traffic accidents. Emergency vehicles include disabled trucks and police cars. Police cars are dispatched for major accidents, personnel assistance, and system-determined emergencies. When any accident occurs, the disabled vehicle will be dispatched by the connected autonomous vehicle highway CAVH system. The system can provide different management and rescue methods for different road and traffic conditions, implementing one or more of the following functional categories: perception, traffic behavior prediction and management, planning and decision-making, and vehicle control. The management system is supported by road infrastructure, real-time wired and / or wireless communication, power supply networks, and network security and security services. While this method can effectively provide reliable assistance to accident vehicles, it still has the following shortcomings:

[0005] First, this method addresses the issue that large-scale repairs of accident vehicles can lead to significant service delays, making it difficult for service providers to efficiently serve accident vehicles.

[0006] Secondly, this method only considers the subsequent rescue and handling of the accident vehicle, without taking into account the impact of the accident on the performance of the autonomous vehicle or the interaction and selection issues between the autonomous vehicle and the service provider. As a result, it leads to a waste of the autonomous vehicle's computing resources and a low vehicle accident handling rate. Summary of the Invention

[0007] The purpose of this invention is to address the shortcomings of the prior art by proposing a digital twin-enabled post-accident repair method for autonomous vehicles. This method aims to reduce service delays for vehicles involved in large-scale accidents, fully utilize the available computing resources of autonomous vehicles after an accident, determine the optimal accident response strategy for autonomous vehicles, enable two-way selection decisions between autonomous vehicles and service providers, and improve the handling rate of accident vehicles and social welfare.

[0008] To achieve the above objectives, the technical solution adopted by the present invention includes the following steps:

[0009] (1) Constructing the task model and network model under the digital twin network:

[0010] (1a) Construct a task set that includes N computational tasks;

[0011] (1b) Construct a digital twin network system comprising I autonomous vehicles i, J cellular base stations j, I autonomous vehicle digital twin i' and J cellular base station digital twin j';

[0012] (2) Each cellular base station's digital twin j' publishes the service providers it represents in the digital twin network, excluding the unit time maintenance cost P. j External repair information;

[0013] (3) Establish a maintenance response strategy for unmanned vehicles after an accident at time t:

[0014] (3a) Calculate the speed s of the unmanned vehicle i after the accident based on the degree of impact of the accident at time t. i Available computing resources a i Repair time and the maintenance budget P for autonomous vehicle i i r ;

[0015] (3b) Calculate the deadline t for autonomous vehicle i to enter the maintenance market. i,max ;

[0016] (3c) According to a i The value of i is calculated for autonomous vehicles. The cost of time loss

[0017] (3d) Based on loss cost Calculate the autonomous vehicle i in Repair quotes at all times

[0018] (3e) Calculate the autonomous vehicle i in [t,t i,max The best time to enter the accident handling market within the specified time period. and repair quotes

[0019] (4) Determine the maintenance strategy for each autonomous vehicle at each cellular base station and the pricing for each autonomous vehicle from each cellular base station:

[0020] (4a) Calculate the start time for maintenance of the autonomous vehicle i by selecting the service provider of the cellular base station j. and end time

[0021] (4b) Calculate the maintenance cost P of autonomous vehicle i to cellular base station j. i,j :

[0022] (4c) Each cellular base station calculates the maintenance quote P for cellular base station j for autonomous vehicle i based on the maintenance strategy determined by the autonomous vehicle. j,i ;

[0023] (5) Complete the matching between cellular base stations and autonomous vehicles:

[0024] (5a) Each autonomous vehicle digital twin i' selects an acceptable set of cellular base stations:

[0025] J i ={j:j∈J,P i,j -P j,i ≥0}, based on P i,j -P j,i The descending order of the preference list L(J) for generating cellular base stations i );

[0026] (5b) Each cellular base station digital twin j' selects an acceptable set of vehicles:

[0027] V i ={i:i∈I,P i,j -P j,i ≥0}, based on P i,j -P j,i In descending order, generate a vehicle preference list L(V) i );

[0028] (5c) Each autonomous vehicle digital twin i' has a preference list L(J) i The first cellular base station sends the maintenance request;

[0029] (5d) The digital twin of the cellular base station j' comprehensively considers the set of vehicles that send maintenance requests. and the set of vehicles it matches Regenerate a new set of matching repair vehicles And vehicles that fail to find a match should have their repair requests resubmitted;

[0030] (5e) Repeat (5c)-(5d) until the preference list of unmatched autonomous vehicles is exhausted, then end the matching process and obtain the set of matching pairs between autonomous vehicles and cellular base stations.

[0031] (6) Determine the maintenance costs payable for autonomous vehicles using the Vickrey-Clark-Groves VCG pricing mechanism:

[0032] (6a) Based on the set of matching pairs Calculate the total social welfare U resulting from autonomous vehicle i selecting cellular base station j, and the total social welfare U resulting from matching vehicles with cellular base stations in the network when autonomous vehicle i is not present. -i :

[0033]

[0034]

[0035] in, express The set excluding the matching pair of autonomous vehicle i is included; (i^,j^) represents A set of matching pairs between autonomous vehicles and cellular base stations; This represents the repair price offered by the autonomous vehicle i^ for the cellular base station j^. This indicates the maintenance quote for cellular base station j^ for autonomous vehicle i^;

[0036] (6b) Based on the above two types of total social welfare U, U -i And the repair quote for cellular base station J P j,i The maintenance cost P for driverless vehicle i i,j Determine the maintenance price that autonomous vehicle i pays to cellular base station j.

[0037]

[0038] (7) Based on the matching results, the service provider of the cellular base station agent will provide maintenance services to the matched autonomous vehicles, and the autonomous vehicles will pay the agreed maintenance price to the service provider after the service is completed.

[0039] Compared with the prior art, the present invention has the following advantages:

[0040] 1. Improved the service efficiency of service providers.

[0041] The digital twin network system for handling accident vehicles proposed in this invention, by considering two types of digital twins—digital twins of autonomous vehicles and digital twins of cellular base stations—achieves real-time acquisition of all information from both autonomous vehicles and service providers. Therefore, in the event of a large-scale vehicle accident, the affected vehicles can promptly interact with service providers within the digital twin network to make decisions, thereby reducing service upload delays and improving the service efficiency of service providers.

[0042] 2. It achieves optimal accident response for autonomous vehicles.

[0043] The optimal accident response strategy for autonomous vehicles designed in this invention takes into account both the impact of the accident on the autonomous vehicle and its benefits in the network. This approach can fully utilize the available computing resources of the autonomous vehicle after the accident, thereby achieving the best response to the accident.

[0044] 3. It improved the vehicle accident handling rate and social welfare.

[0045] This invention models the two-way selection problem between accident vehicles and service providers as a dynamic matching game, thereby matching the best service provider to the autonomous vehicle involved in the accident and determining the repair price to be paid by the autonomous vehicle according to the VCG pricing mechanism, thus improving the vehicle accident handling rate and social welfare. Attached Figure Description

[0046] Figure 1 This is a flowchart illustrating the implementation of the present invention;

[0047] Figure 2 This is a framework diagram of post-accident vehicle repair empowered by digital twins in this invention;

[0048] Figure 3 This is a comparison chart of the average processing rates of accident vehicles by the present invention and existing technologies;

[0049] Figure 4 This is a comparison chart of the average revenue obtained by service providers by the present invention and existing technologies;

[0050] Figure 5 This is a comparison chart of the average social welfare of the present invention and the prior art. Detailed Implementation

[0051] The embodiments and effects of the present invention will be further described in detail below with reference to the accompanying drawings and specific examples.

[0052] Reference Figure 1 The implementation steps for this example are as follows:

[0053] Step 1: Construct the task model and network model under the digital twin network.

[0054] Reference Figure 2 The implementation of this step is as follows:

[0055] 1.1) Construct a task set comprising N computational tasks:

[0056]

[0057] Where: each computational task n can be represented as: {a n ,l n ,w n}, and w n =wa n l n a n l represents the computing resources required to execute computational task n. n w represents the time required to rent computing resources for computation task n. n represents the achievable reward for performing computational task n, and w represents the incentive coefficient;

[0058] 1.2) Construct a digital twin network system comprising I autonomous vehicles i, J cellular base stations j, I autonomous vehicle digital twin i', and J cellular base station digital twins j', wherein:

[0059] The network state information of each autonomous vehicle's digital twin i' at time t is represented as:

[0060]

[0061] In the formula, This represents the position coordinates of the autonomous vehicle i at time t; This represents the speed of the driverless vehicle i at time t; This represents the maximum speed of driverless vehicle i. A represents the computing resources available to autonomous vehicle i at time t; i This represents the total computing resources possessed by autonomous vehicle i;

[0062] The information of the maintenance service provider represented by each cellular base station digital twin j' is represented as follows:

[0063]

[0064] In the formula, (x j ,y j ) represents the geographical location of the service provider that cellular base station j acts as an agent for; This indicates the start and end times of the maintenance service provided by the service provider acting as an agent for cellular base station j; s j P represents the average speed at which the service provider acting as the agent of cellular base station j travels to the vehicle to provide on-site repair services; j This represents the unit time maintenance cost provided by the service provider acting as an agent for cellular base station j.

[0065] Step 2: Publish the repair information for each service provider.

[0066] Each cellular base station's digital twin j' publishes the service providers it represents within the digital twin network, excluding the unit time maintenance cost P. j External maintenance information is represented as follows:

[0067]

[0068] Step 3: Establish a maintenance response strategy for an unmanned vehicle after an accident at time t.

[0069] The extent of the accident's impact can be expressed as:

[0070]

[0071] In the formula, This indicates the degree of impact of the accident on the movement state of the autonomous vehicle i. This indicates the extent to which the accident affects the available computing resources of autonomous vehicle i.

[0072] The response strategy of autonomous vehicle i to an accident is represented as:

[0073]

[0074] In the formula, P i r This represents the maintenance budget considered for driverless vehicle i; This indicates the time when driverless vehicle i entered the accident handling market; This represents the maintenance time required for driverless vehicle i; s i The average speed of the autonomous vehicle i after the accident is represented by the following parameters:

[0075] 3.1) Based on the degree of impact of an accident involving an unmanned vehicle at time t Calculate the speed s of the unmanned vehicle i after the accident. i Available computing resources a i Repair time and the maintenance budget P for autonomous vehicle i i r :

[0076]

[0077]

[0078]

[0079]

[0080] In the formula, This represents the duration assessment parameter indicating the impact of the accident on the moving state of the autonomous vehicle i; i p represents the duration assessment parameter indicating the impact of an accident on the available computing resources of autonomous vehicle i; i This represents the maintenance cost per unit of time for driverless vehicle i.

[0081] 3.2) Based on the results of 3.1), calculate the deadline t for autonomous vehicle i to choose to enter the maintenance market. i,max The formula is as follows:

[0082]

[0083] In the formula, This indicates the longest deadline for providing repair services among all service providers; d max This indicates the farthest distance between the autonomous vehicle and the nearest repair point; This represents the minimum acceptable driving speed for driverless vehicle i, when... The autonomous vehicle i expects to be repaired at the accident site; otherwise, the autonomous vehicle i will... i Drive at the selected service provider at the designated speed for repairs; min(s) j (j∈J) represents the minimum average driving speed among all service providers;

[0084] 3.3) According to a i The value of i is calculated for autonomous vehicles. The cost of time loss

[0085] when a i =0, the computing resources of autonomous vehicle i are completely destroyed, requiring the cancellation of all computing tasks already undertaken. This loss incurs a cost. Represented as:

[0086]

[0087] In the formula, n represents the set of computational tasks performed by autonomous vehicle i at time t; n^ represents the task set. One of the computational tasks; This represents the default constant for computation task n^; This represents the reward obtained from executing computational task n^;

[0088] when a i ≠0, the autonomous vehicle i will handle some computational tasks in a way that violates the rules, as specifically implemented below:

[0089] 3.3.1) For autonomous vehicle i, the deadline for the computational task is greater than... The computational task involves default processing, and the resulting losses incur costs. Represented as:

[0090]

[0091] In the formula, This indicates that the deadline undertaken by autonomous vehicle i is greater than [a certain time]. The set of computational tasks; n' represents the task set. One of the computational tasks; k n'w represents the default constant for calculating task n'; n' This represents the reward obtained from performing computational task n';

[0092] 3.3.2) Calculate the available computing resources for driverless vehicle i and the deadline for undertaking is less than or equal to The computing resources required for the computing task

[0093]

[0094]

[0095] In the formula, a n' This represents the computing resources required to execute computation task n'. This indicates that the autonomous vehicle i is responsible for excluding the deadline being greater than [a certain time]. The set of computational tasks; n * Represents a set of tasks One of the computational tasks; Indicates the execution of computational task n * Required computing resources;

[0096] 3.3.3) According to and Calculate the autonomous vehicle i in Total loss cost at all times

[0097]

[0098] In the formula, k n ' represents the default constant for computation task n'; w n ' represents the reward obtained from executing computational task n';

[0099] Indicates that the autonomous vehicle i selects a set of tasks. The minimum default cost for partial computational tasks in the process of defaulting, and satisfying the condition of autonomous vehicle i in... The available computing resources at any given time are greater than or equal to the execution time. The amount of computing resources required for tasks that do not violate the agreement; Represents the computation task n * The default constant; Indicates the execution of computational task n * The gains obtained.

[0100] 3.4) Based on the cost of loss Calculate the autonomous vehicle i in Repair quotes at all times

[0101] when a i =0, the unmanned vehicle i immediately enters the accident repair market, at this time:

[0102]

[0103] when a i ≠0. The autonomous vehicle i will utilize its spare computing resources to select new computing tasks in the digital twin network to obtain benefits. The specific implementation is as follows:

[0104] 3.4.1) Calculate the autonomous vehicle i in Real-time idle computing resources and expected idle computing resources

[0105]

[0106]

[0107] In the formula, Indicates that driverless vehicle i is in t i 'The set of computational tasks that a given moment needs to complete; m represents the task set.' One of the computational tasks; a m This represents the computing resources required to execute computation task m;

[0108] 3.4.2) Determine if driverless vehicle i is in A new set of computational tasks available at any time Represented as:

[0109]

[0110] In the formula, T(n) represents the deadline for calculating task n;

[0111] 3.4.3) Calculate the autonomous vehicle i at time t i 'Maximum actual gain P obtained at time ' i x (t') and maximum expected return P i y (t'):

[0112] and

[0113] and

[0114] In the formula, Indicates that driverless vehicle i is at t′ iFrom the task set at all times The actual maximum benefit gained by undertaking a new task m' in the process; w m′ a represents the benefit gained from performing computational task m'; m′ This represents the computing resources required to execute computation task m'; Indicates that driverless vehicle i is at t′ i From the task set at all times China undertakes new tasks * The expected maximum return; Indicates the execution of computation task m * The gains obtained; Indicates the execution of computation task m * Required computing resources;

[0115] 3.4.4) Calculate the autonomous vehicle i in Maximum actual profit obtained within the period and maximum expected return

[0116]

[0117]

[0118] 3.4.5) Based on the cost of loss Calculate the autonomous vehicle i in Repair quotes at all times

[0119]

[0120] In the formula, Indicates that driverless vehicle i is The revenue earned from completing the computational tasks assigned to you within a given time period;

[0121] 3.5) Calculate the autonomous vehicle i in [t,t i,max The best time to enter the accident handling market within the specified time period. and repair quotes

[0122]

[0123]

[0124] In the formula, This indicates that the driverless vehicle i is in [t,t] i,max The moment with the highest repair price within the specified time period; This indicates that the driverless vehicle i is in [t,t] i,max The highest repair price during the specified time period.

[0125] Step 4: Determine the start time for maintenance of each autonomous vehicle at each cellular base station. and end time Repair quote P i,j And the maintenance quote P for cellular base station j for autonomous vehicle i j,i .

[0126] 4.1) Calculate the start time for maintenance of the autonomous vehicle i by selecting the service provider acting as an agent of cellular base station j. and end time

[0127]

[0128]

[0129] In the formula, mind i,j Represents the position coordinates of autonomous vehicle i at time t. Location coordinates (x, y) of the service provider proxies the network edge device j j ,y j The shortest distance between ( );

[0130] 4.2) Calculate the maintenance cost P of autonomous vehicle i for cellular base station j. i,j :

[0131]

[0132] In the formula, q i This represents the cost per unit time for autonomous vehicle i to travel;

[0133] 4.3) Each cellular base station is maintained according to the start time of maintenance by the service provider to which the autonomous vehicle is represented at each cellular base station. and end time Calculate the maintenance cost P for cellular base station j on autonomous vehicle i. j,i :

[0134]

[0135] Step 5: Complete the matching between the cellular base station and the autonomous vehicle.

[0136] 5.1) Determine the preference list for each autonomous vehicle:

[0137] Each autonomous vehicle digital twin selects an acceptable set of cellular base stations. i ={j:j∈J,P i,j -P j,i ≥0}, based on P i,j-P j,i The descending order of the preference list L(J) for generating cellular base stations i );

[0138] 5.2) Determine the preference list for each cellular base station:

[0139] Each cellular base station digital twin selects an acceptable set of autonomous vehicles. i ={i:i∈I,P i,j -P j,i ≥0}, based on P i,j -P j,i The preferred list L(V) for autonomous vehicles is generated in descending order. i );

[0140] 5.3) Each autonomous vehicle digital twin i' has a preference list L(J) i The first cellular base station sends the maintenance request;

[0141] 5.4) Generate a new matching maintenance vehicle for each cellular base station digital twin:

[0142] The digital twin of the cellular base station takes into account the autonomous vehicles that send maintenance requests to it. Autonomous vehicles matched with digital twins of cellular base stations Based on the following formula, a new matching repair vehicle will be generated.

[0143]

[0144] In the formula, express An autonomous vehicle that is matched with a digital twin of a cellular base station j', which meets the requirements of The maintenance times for any of the autonomous vehicles in the system do not overlap, and are consistent with... The sum of the repair costs for cellular base station j from other autonomous vehicles is the highest.

[0145] If the autonomous vehicle's digital twin i' does not match the cellular base station that sent the maintenance request, it will remove that cellular base station from its preference list L(J). i Remove from the list and reselect the top-ranked cellular base station in the preference list to send a maintenance request;

[0146] 5.5) Complete the matching of network edge devices with autonomous vehicles:

[0147] Repeat steps (5.2) and (5.4) until the preference list of unmatched autonomous vehicles is exhausted, then end the matching process to obtain the set of matching pairs between autonomous vehicles and cellular base stations.

[0148] Step 6: Determine the maintenance price payable for the autonomous vehicle based on the Vickrey-Clark-Groves VCG pricing mechanism.

[0149] 6.1) Based on the set of matching pairs Calculate the total social welfare U resulting from autonomous vehicle i selecting cellular base station j, and the total social welfare U resulting from matching autonomous vehicle i with cellular base station j in the network when autonomous vehicle i is not present. -i :

[0150]

[0151]

[0152] In the formula, express The set excluding the matching pair of autonomous vehicle i is included; (i^,j^) represents A set of matching pairs between autonomous vehicles and cellular base stations; This represents the repair price offered by the autonomous vehicle i^ for the cellular base station j^. This indicates the maintenance quote for cellular base station j^ for autonomous vehicle i^;

[0153] 6.2) Based on the above two types of total social welfare U and U -i And the repair quote for cellular base station J P j,i The maintenance cost P for driverless vehicle i i,j Determine the maintenance price that autonomous vehicle i pays to cellular base station j.

[0154]

[0155] Step 7: Repair and pay the price.

[0156] Based on the matching results, the service provider acting as the agent for the cellular base station will provide maintenance services to the matched autonomous vehicles. After the maintenance is completed, the autonomous vehicles will pay the agreed maintenance price to the service provider.

[0157] The technical effects of the present invention will be further explained below with reference to simulation experiments.

[0158] 1. Simulation conditions

[0159] This invention uses a 3x3km map, sets the accident handling time to [1, 2000]s, and uses Dijkstra's algorithm to calculate the shortest distance between two points.

[0160] The simulation platform consisted of a Windows 10 operating system and MATLAB 2020b.

[0161] The simulation experiment parameter settings are shown in Table 1:

[0162] Table 1

[0163]

[0164]

[0165] Where J represents the number of cellular base stations; I represents the number of autonomous vehicles in the network; This represents the position coordinates of the autonomous vehicle i at time t; This represents the speed of the driverless vehicle i at time t; This represents the maximum speed of driverless vehicle i. This represents the computing resources available to autonomous vehicle i at time t; (x j ,y j ) represents the geographical location of the service provider that cellular base station j acts as an agent for; This indicates the start and end times of the maintenance service provided by the service provider acting as an agent for cellular base station j; s j P represents the average speed at which the service provider acting as the agent of cellular base station j travels to the vehicle to provide on-site repair services for the accident vehicle; j represents the unit time maintenance cost provided by the service provider acting as an agent for cellular base station j; w represents the incentive coefficient. and l i These represent the duration-based evaluation parameters for mobility and computing resources, respectively; p i This represents the maintenance cost per unit of time for driverless vehicle i.

[0166] 2. Simulation Content

[0167] In the above scenario, the accident vehicle processing rate, average revenue of service providers, and average social welfare of the present invention and existing immediate response, random response, and nearest-location processing solutions are simulated respectively, wherein:

[0168] The immediate response scheme refers to determining the response strategy as soon as the accident occurs and selecting a service provider through a designed dynamic matching game.

[0169] The random response scheme refers to the random response of the accident vehicle within the time from the occurrence of the accident to the optimal response time, and the selection of service providers is carried out through a designed dynamic matching game.

[0170] The nearby handling plan refers to determining the best response strategy for the accident vehicle and then selecting the nearest service provider for repairs;

[0171] Simulation 1: In the above scenario, this invention, by changing the number of cellular base stations, simulates the average processing rate of accident vehicles using this invention and existing immediate response, random response, and nearest-neighbor processing schemes. The results are as follows: Figure 3 .

[0172] from Figure 3 It can be seen that the average handling rate of accident vehicles increases with the increase of the number of cellular base stations. This is because with the increase of the number of cellular base stations, the number of service providers that each autonomous vehicle can choose for repair also increases. The performance of the proximity scheme is different from the other three schemes because this scheme always selects the nearest repair service provider. The present invention can achieve a higher accident handling rate than other schemes.

[0173] Simulation 2: In the above scenario, this invention, by changing the number of cellular base stations, simulates the average revenue of service providers using this invention and existing immediate response, random response, and nearest-neighbor processing solutions. The results are as follows: Figure 4 .

[0174] from Figure 4 It can be seen that as the number of cellular base stations increases, the average revenue of the service provider decreases. This is because the vehicles generated in the simulation follow a Poisson distribution, and the parameters of the Poisson distribution are fixed. Therefore, the number of accident vehicles is fixed, and the total revenue that can be brought to the service provider is limited. Furthermore, as can be seen from the figure, the average revenue of the service provider in this invention is consistently the highest.

[0175] Simulation 3: In the above scenario, this invention simulates the average social welfare of the present invention and existing immediate response, random response, and nearest-neighbor processing schemes by changing the number of cellular base stations. The results are as follows: Figure 5 .

[0176] from Figure 5 It can be seen that as the number of cellular base stations increases, the average social welfare of both the present invention and the existing solutions increases. This is because the increase in the number of network edge devices intensifies competition among service providers, thereby bringing higher social welfare to accident vehicles. However, compared with the existing solutions, the average social welfare obtained by the present invention is always the highest.

[0177] In summary, compared with the existing immediate response, random response, and localized processing solutions, the present invention can bring about a higher average processing rate for accident vehicles, higher average revenue for service providers, and higher average social welfare.

Claims

1. A digital twin-enabled method for post-accident repair of unmanned vehicles, characterized in that, Includes the following steps: (1) Constructing the task model and network model under the digital twin network: (1a) Construction including A set of computational tasks; (1b) Construction including One driverless vehicle , Cellular base stations , A digital twin of an autonomous vehicle and Digital twin of a cellular base station Digital twin network system; (2) Digital twin of each cellular base station In a digital twin network, publish the service providers you represent, excluding unit time maintenance costs. External repair information; (3) Establish autonomous vehicles in Repair response strategy after an accident: (3a) According to The extent of the impact of an accident involving an autonomous vehicle is calculated at any given moment, and the post-accident impact on the autonomous vehicle is determined. driving speed Available computing resources Repair time and driverless vehicles Repair budget ; (3b) Calculate driverless vehicles You can choose the deadline for entering the repair market. ; (3c) According to The value of the calculation for autonomous vehicles exist The cost of time loss The formula is as follows: ； In the formula, Indicates driverless vehicles exist The set of tasks that perform computational tasks at any given time; Represents a set of tasks One of the computational tasks; Represents computational task The default constant; Indicates the execution of a computational task The gains obtained; Indicates driverless vehicles The deadline for undertaking is greater than The set of computational tasks; Represents a set of tasks One of the computational tasks; Represents computational task The default constant; Indicates the execution of a computational task The gains obtained; Indicates driverless vehicles Select task set The minimum cost of defaulting on partial computational tasks within the system, and to meet the requirements of autonomous vehicles. exist The available computing resources at any given time are greater than or equal to the execution time. The amount of computing resources required for tasks that do not violate the agreement; Indicates driverless vehicles Bearing the responsibility except for the deadline greater than The set of computational tasks; Represents a set of tasks One of the computational tasks; Represents computational task The default constant; Indicates the execution of a computational task The gains obtained; Indicates that there was no driverless vehicle after the accident. Available computing resources; Indicates driverless vehicles exist The amount of computing resources available at any given time; Indicates driverless vehicles exist Execute at all times The amount of computing resources required for all tasks in the process; (3d) Based on the cost of loss Calculate driverless vehicles exist Repair quotes at all times The formula is as follows: ； In the formula, Indicates driverless vehicles Maintenance budget; Indicates driverless vehicles exist The cost of losing time; Indicates that there was no driverless vehicle after the accident. Available computing resources; Indicates driverless vehicles exist The maximum actual benefit gained from undertaking new computing tasks within a given time period; Indicates driverless vehicles exist The maximum expected benefit to be gained from undertaking new computing tasks within a given time period; Indicates driverless vehicles The deadline for undertaking is greater than The set of computational tasks; Represents a set of tasks One of the computational tasks; Represents computational task The default constant; Indicates the execution of a computational task The gains obtained; Indicates driverless vehicles exist The revenue earned from completing the computational tasks assigned to you within a given time period; (3e) Calculate driverless vehicles exist The best time to enter the accident handling market within a given timeframe and repair quotes ; (4) Determine the maintenance strategy for each autonomous vehicle at each cellular base station and the pricing for each autonomous vehicle at each cellular base station: (4a) Calculate driverless vehicles Select Cellular Base Station The start time of repairs by the authorized service provider. and end time : (4b) Calculate driverless vehicles For cellular base stations Repair quote : (4c) Each cellular base station calculates the maintenance strategy determined by the autonomous vehicle. For driverless vehicles Repair quote ; (5) Complete the matching between cellular base stations and autonomous vehicles: (5a) Digital twin of each driverless vehicle Select an acceptable set of cellular base stations: ,based on A preference list for generating cellular base stations in descending order. ; (5b) Digital twin of each cellular base station Select an acceptable set of vehicles: ,based on In descending order, generate a list of vehicle preferences. ; (5c) Digital twin of each autonomous vehicle To the preference list The leading cell tower sends a maintenance request; (5d) Digital twin of cellular base station Taking into account the set of vehicles that submitted repair requests and the set of vehicles it matches Regenerate a new set of matching repair vehicles. And vehicles that fail to find a match will have their repair requests resubmitted; (5e) Repeat (5c)-(5d) until the preference list of unmatched autonomous vehicles is exhausted, then end the matching process and obtain the set of matching pairs between autonomous vehicles and cellular base stations. ; (6) Determine the maintenance costs payable for autonomous vehicles using the Vickrey-Clark-Groves VCG pricing mechanism: (6a) Based on the set of matching pairs Calculate the autonomous vehicles separately Select Cellular Base Station Total social welfare brought about and driverless vehicles Total social welfare resulting from matching vehicles with cellular base stations in the absence of the network : ； ； in, express Excluding driverless vehicles A set of matching pairs; express A set of matching pairs between autonomous vehicles and cellular base stations; Indicates driverless vehicles For cellular base stations Repair quote; Indicates cellular base station For driverless vehicles Repair quote; (6b) Based on the above two types of total social welfare , and cellular base stations Repair quote driverless vehicles Repair quote Determine driverless vehicles to cellular base stations Repair price paid : ； (7) Based on the matching results, the service provider of the cellular base station agent will provide maintenance services to the matched autonomous vehicles, and the autonomous vehicles will pay the agreed maintenance price to the service provider after the service is completed.

2. The method according to claim 1, characterized in that, According to (3a) The extent of the impact of an accident involving an autonomous vehicle is calculated at any given moment, and the post-accident impact on the autonomous vehicle is determined. driving speed Available computing resources Repair time and driverless vehicles Repair budget The formula is as follows: ， ， ， ， In the formula, This indicates the accident affected the driverless vehicle. The degree of influence of the movement state ; Indicates driverless vehicles Maximum driving speed; This indicates the accident affected the driverless vehicle. The extent of the impact of available computing resources ; Indicates driverless vehicles Total computing resources available; This indicates the accident affected the driverless vehicle. Duration assessment parameters under the influence of mobility status; Indicates driverless vehicles exist The speed at any given moment; This indicates the accident affected the driverless vehicle. Duration assessment parameters under the influence of available computing resources; Indicates driverless vehicles Maintenance expenses per unit of time.

3. The method according to claim 1, characterized in that, The calculation of autonomous vehicles in (3b) You can choose the deadline for entering the repair market. The formula is as follows: ， In the formula, This indicates the longest deadline for providing repair services among all service providers; Indicates cellular base station The deadline for the authorized repair service provider to provide repair services; This represents the collection of cellular base stations belonging to various service providers; Indicates driverless vehicles Repair time; This indicates the farthest distance between the autonomous vehicle and the nearest repair point; Indicates that there was no driverless vehicle after the accident. The speed of travel; Indicates driverless vehicles The minimum acceptable driving speed; This represents the lowest average driving speed among all service providers; Indicates cellular base station The average speed at which the authorized repair service provider travels to the accident site to provide on-site repair services for the driverless vehicle.

4. The method according to claim 3, characterized in that, The calculation of driverless vehicles in (3e) exist The best time to enter the accident handling market within a given timeframe and repair quotes The formula is as follows: ， ， In the formula, Indicates driverless vehicles exist The moment with the highest repair price within the specified time period; Indicates driverless vehicles exist Real-time repair quotes; Indicates that there was no driverless vehicle after the accident. Available computing resources; Indicates driverless vehicles Maintenance budget; Indicates driverless vehicles exist The cost of losing time; Indicates driverless vehicles exist The highest repair price within the specified time period.

5. The method according to claim 4, characterized in that, The calculation of unmanned vehicles in (4a) Select network edge devices The start time of repairs by the authorized service provider. and end time The formula is as follows: ， ， In the formula, Indicates driverless vehicles The best time to enter the accident handling market; Indicates driverless vehicles Position coordinates With cellular base stations Location coordinates of the agent's service provider The shortest distance between them; Indicates that there was no driverless vehicle after the accident. The speed of travel; Indicates driverless vehicles The minimum acceptable driving speed; Indicates driverless vehicles Repair time.

6. The method according to claim 5, characterized in that, The calculation of unmanned vehicles in (4b) For cellular base stations Repair quote The formula is as follows: ， In the formula, Indicates driverless vehicles Select Cellular Base Station The start time of repairs by the authorized service provider; Indicates driverless vehicles Select Cellular Base Station The end time of repairs by the authorized service provider; Indicates cellular base station The service provider on behalf of the agent provides the start time for repair services; Indicates cellular base station The deadline for the authorized repair service provider to provide repair services; Indicates driverless vehicles exist Repair bids are constantly being offered in the accident handling market; Indicates driverless vehicles Cost per unit time of travel; Indicates driverless vehicles exist Position coordinates at time With network edge devices Location coordinates of the agent's service provider The shortest distance between them; Indicates that there was no driverless vehicle after the accident. The speed of travel; Indicates driverless vehicles The minimum acceptable driving speed.

7. The method according to claim 6, characterized in that, The computing network edge device in (4c) For driverless vehicles Repair quote The formula is as follows: ， In the formula, Indicates driverless vehicles Select Cellular Base Station The start time of repairs by the authorized service provider; Indicates driverless vehicles Select Cellular Base Station The end time of repairs by the authorized service provider; Indicates cellular base station The repair cost per unit of time for the service provider you represent.

8. The method according to claim 7, characterized in that, The cellular base station mentioned in (5d) Regenerate a new set of matching repair vehicles , means as follows: ， In the formula, express China and cellular base stations A matching driverless vehicle that meets the requirements of The maintenance times for any of the autonomous vehicles in the system do not overlap, and are consistent with... Other autonomous vehicles in China use cellular base stations The sum of the repair quotes is the highest; Indicates to cellular base stations A collection of driverless vehicles that have sent maintenance requests; Indicates connection with cellular base stations The set of matched driverless vehicles.

Images