Charging guiding method based on car networking data driving

Abstract

The invention discloses a charging guiding method based on Internet of Vehicles data driving, which belongs to the field of Internet of Vehicles and comprises five steps of data acquisition, path optimization, reputation management, double-layer iteration and guiding execution. Aiming at the problems that the traditional static path planning (such as the Dijkstra algorithm) does not consider actual traffic factors and the path search efficiency is low, dynamic path optimization is realized by improving an A * algorithm: a known item G (i) introduces a traffic light mechanism, traffic light waiting time is taken as a node weight to be incorporated into actual cost calculation, and the path search efficiency is improved. The problem of time consumption exceeding estimation caused by traffic light waiting is avoided; the pre-estimation item H (i) is added into a return road penalty item, the actual preference of'little turning around and no return road 'of a vehicle owner is met, and invalid path search is reduced; and finally, an optimal path is preferentially selected through a decision function, roads having no influence on an optimal result are avoided, the path searching efficiency is remarkably improved, and the total consumed time of'road driving time + charging station waiting time 'of the user is shortened.

Description

Technical Field

[0001] This invention relates to the field of vehicle networking technology, and in particular to a charging guidance method based on vehicle networking data. Background Technology

[0002] With the increasing popularity of electric vehicles and the development of vehicle-to-everything (V2X) technology, "charging anxiety" and "imbalance in the quality of V2X services" have become core pain points in the industry. Traditional charging guidance relies on static path planning (such as Dijkstra's algorithm), which does not take into account actual traffic factors such as traffic light waiting time and U-turn behavior, resulting in low path search efficiency and users' actual time consumption far exceeding the estimate. During the unloading process of vehicle-to-everything (V2X) tasks, edge servers may have untrust issues (such as refusal to cooperate or calculation errors). Existing solutions lack effective reputation management mechanisms, making it difficult to guarantee cooperation enthusiasm and task reliability. Charging guidance (transportation network side) and task offloading (vehicle network side) are independent of each other, and the coupling relationship between the two is not considered. Changes in traffic flow will change the task distribution, and task offloading will affect traffic flow scheduling, ultimately leading to high vehicle network operating costs or poor service quality. Existing charging plans do not fully integrate vehicle-to-everything (V2X) data such as user profiles (e.g., charging time preferences) and battery degradation (e.g., remaining lifespan), making it impossible to accurately match user needs and fundamentally solve charging anxiety.

[0003] Therefore, a charging guidance method that integrates vehicle-to-everything (V2X) data, route optimization, reputation management, and dual-network collaboration is needed to improve guidance efficiency and service quality. Summary of the Invention

[0004] To address the shortcomings of existing technologies, this invention provides a data-driven, multi-module integrated, and dual-network collaborative charging guidance scheme, solving the problems of "low path planning efficiency, unreliable task unloading, lack of dual-network collaboration, and prominent charging anxiety" in existing technologies.

[0005] Technical Solution: To solve the above-mentioned technical problems, according to one aspect of the present invention, more specifically, a charging guidance method based on vehicle network data, comprising a five-step process of "data acquisition - path optimization - reputation management - two-layer iteration - guidance execution", as detailed below: S1. Vehicle-to-Everything (V2X) Data Acquisition and Preprocessing Data sources: On-Board Unit (OBU), Roadside Unit (RSU), 5G Macro Base Station, Charging Station Management System, Electric Vehicle Battery Management System (BMS); Data type: User dimension: User profile (charging time preference, historical charging record); Vehicle-level data: battery degradation data (based on BMS charge / discharge cycle count statistics), real-time battery range, and vehicle's current location; Traffic dimensions: road congestion status, traffic light timing (updated by time period), and charging station waiting queue length; Communication dimension: Computational load, transmission rate, and equipment operating status of RSU / 5G macro base stations; Preprocessing methods: Statistical data analysis methods are used to form a standardized dataset by removing outliers (such as battery life data that exceeds a reasonable range), standardizing data (unifying the data format of different dimensions), and aligning time series (synchronizing the timestamps of various data sources).

[0006] S2. Improved A* algorithm for charging path planning To address the issue of traditional A* algorithms not being well-suited for urban traffic scenarios, improvements are made from two dimensions: "known cost" and "estimated cost." The core formula is as follows: ; Improvement of known term G(i): In the scenario where traffic lights are widely present on urban roads, a traffic light mechanism is introduced, and the traffic light waiting time is set as the node weight and added to the actual cost from the initial node to the current node i, that is, G(i) = travel time from the initial node to i + traffic light waiting time; Improved prediction term H(i): The traditional A* algorithm uses the Euclidean geometric distance from the current point to the target point as H(i), but in reality, drivers tend to choose the path with fewer U-turns (avoiding backtracking). Therefore, a backtracking penalty term is added to the Euclidean distance, i.e., H(i) = Euclidean geometric distance + backtracking penalty value (triggered when there is a U-turn). Path output logic: The node priority is determined by the F(i) value, and the path with the smallest F(i) is selected as the initial charging path to avoid paths that have no impact on the optimal path and improve search efficiency.

[0007] S3. Construction and Update of Two-Dimensional Reputation Value Model To address the issues of untrusted edge servers and vehicle behavior norms, a dual-reputation system of "server-vehicle" is constructed: Edge server reputation value model: Model definition: The set represents the edge servers, with each server corresponding to a triple. ;in, (Direct Reputation Score) is calculated statistically based on task completion performance (such as task completion rate and calculation accuracy). (Indirect reputation score) is calculated based on the server's own status (such as response latency, communication interruption frequency), and both scores range from 0 to 1; Dynamic update strategy: Basic update: via formula Implementation, in which , It is a weighting factor. + =1; Time slice weight adjustment: A formula is designed to quickly identify untrusted servers. , Represents the update rate. The larger the value, the greater the impact on the reputation value within the current time slice; the vehicle reputation value model refers to the edge server reputation value model, and incorporates a mutual evaluation mechanism between vehicles, and the edge server reputation value determines its weight in vehicle reputation value management.

[0008] Vehicle reputation value model: The basic framework is based on the edge server reputation value model, and a new "mutual evaluation mechanism between vehicles" (such as mutual evaluation of charging order and data upload authenticity) has been added to improve the vehicle reputation calculation logic; Weighting rules: The higher the reputation value of the edge server, the greater its evaluation weight in vehicle reputation value management, ensuring the credibility of vehicle reputation statistics; Incentive mechanisms: Edge servers with higher reputation scores are given priority in responding to collaborative offloading requests to enhance their participation; vehicles with higher reputation scores are provided with benefits such as priority queuing at charging stations to regulate vehicle behavior.

[0009] S4, Solving the two-level optimization model To address the coupling relationship between the transportation network and the vehicle-to-everything (V2X) network, a two-layer optimization model is constructed, and collaborative optimization is achieved through alternating iterations: Upper layer: Active charging guidance model on the transportation network side: Core logic: Replace the traditional Dijkstra algorithm and use the improved A* algorithm in step S2 to calculate the shortest path, with the goal of minimizing the total time of "road travel time + charging station waiting time + charging time"; Constraints: Road selection constraints: In the set of visited road nodes V, the initial node must have an electric vehicle departing from it; in the set of target charging station nodes C, the target node D must have a vehicle arriving at it; for any intermediate node or initial node, the number of vehicles entering equals the number of vehicles leaving. Arrival time constraint: The vehicle arrival time Tarr must match the departure time Tdep to meet the user's expected charging time requirements; Lower layer: Vehicle-to-everything (V2X) side task offloading model: Core logic: By coordinating tasks through mobile edge computing, the effective allocation of computing and communication tasks is achieved, with the goal of minimizing the operating costs of the Internet of Vehicles (computing costs + communication costs). Constraints: The optimization objective of minimizing the operating cost of the Internet of Vehicles can be expressed as: The constraints include: the unloading decision is a 0-1 constraint, and the constraints are as follows: The task unloading constraint, where the ratio should be in the range of 0-1, and the sum of the unloading ratios is 1, is as follows: Let the transmission from the electric vehicle to the roadside unit be v2r, and the transmission from the roadside unit to the 5G macro base station be r2b. The constraint that the transmission rates of v2r and r2b are not less than the minimum transmission rate is: ; Iterative solution process (reference) Figure 2 ): Initialization: Determine the convergence coefficients of the upper and lower layers, obtain the charging station waiting time and the initial operating status of the vehicle network equipment, and determine the charging destination; Upper-level calculation: Electric vehicles perform path planning based on road resistance, charging station waiting time, and the operation status of vehicle-to-everything (V2X) devices, solve the active charging guidance model, and obtain the traffic flow information matrix; Lower-level calculation: The traffic flow information matrix is ​​input into the task unloading model, the model is solved to obtain the task unloading plan, and the operation status of the vehicle network equipment is updated; Convergence judgment: If the current result meets the convergence coefficient requirement, output the traffic flow information matrix, task unloading plan and vehicle network equipment load status; if not, return to the upper layer to recalculate until convergence.

[0010] S5, Charging Guide Execution The converged optimal charging route (including real-time traffic lights and congestion alerts), estimated charging station waiting time, and task offloading status (such as the target server for task offloading and transmission progress) are pushed to the user terminal (in-vehicle screen, mobile APP) to guide the vehicle to the target charging station accurately. At the same time, the vehicle network automatically completes the allocation of computing tasks according to the offloading plan (such as offloading navigation data processing tasks to the nearest RSU) to ensure service quality.

[0011] The beneficial effects of the charging guidance method based on vehicle network data driven by the present invention are as follows: (1) This invention addresses the problem that traditional static path planning (such as Dijkstra's algorithm) does not consider actual traffic factors and has low path search efficiency by improving the A* algorithm to achieve dynamic path optimization: Given that term G(i) introduces a traffic light mechanism, the traffic light waiting time is included as a node weight in the actual cost calculation, thus avoiding the problem of time exceeding the estimate due to traffic light waiting time; The estimated term H(i) incorporates a backtracking penalty term, which aligns with drivers' actual preference for "fewer U-turns and less backtracking," thereby reducing invalid path searches. Ultimately, the optimal path is selected first by the decision function F(i)=G(i)+H(i), avoiding roads that have no impact on the optimal result, which significantly improves the path search efficiency and shortens the total time spent by users "road travel time + charging station waiting time".

[0012] (2) To address the untrustworthiness issues such as refusal to cooperate and calculation errors of edge servers in the unloading of vehicle networking tasks, this invention constructs a two-dimensional reputation value model of "edge server-vehicle": The edge server reputation value model dynamically evaluates server trustworthiness through direct reputation value (based on task completion rate and calculation accuracy) and indirect reputation value (based on response latency and communication interruption frequency). Combined with a time slice weight adjustment strategy (current behavior has a more significant impact on reputation value), it can quickly identify untrustworthy servers. A supporting incentive mechanism (edge ​​servers with high reputation scores respond to collaborative offload requests first) effectively enhances the enthusiasm of edge servers to participate in collaboration and reduces denial-of-service behavior; The vehicle reputation value model incorporates a mutual evaluation mechanism between vehicles, and assigns evaluation weights based on the reputation value of edge servers to regulate vehicle behavior (such as charging order and data upload authenticity), further ensuring the reliability and security of vehicle-to-everything (V2X) task offloading. (3) This invention addresses the problem that traditional charging guidance and task offloading are independent and do not consider the coupling relationship between the two networks (traffic flow affects task distribution, and task load affects traffic scheduling). It constructs a two-layer optimization model and solves it through alternating iterations: The upper-level transportation network-side active charging guidance model (based on the improved A* algorithm) outputs a traffic flow information matrix, providing accurate traffic scenario basis for the lower-level task offloading; The lower-level vehicle network task offloading model aims to minimize computational and communication costs, outputs a task offloading plan, updates the vehicle network operation status, and feeds back to the upper-level path planning. The dual-network collaborative iteration until convergence effectively balances the operating costs and service quality of the Internet of Vehicles, avoiding excessive costs or service quality imbalances caused by the independence of the two networks. Attached Figure Description

[0013] The present invention will now be described in further detail with reference to the accompanying drawings and specific implementation methods.

[0014] Figure 1 This is a flowchart of the improved A* algorithm dynamic path planning of the present invention; Figure 2 This is a schematic diagram of the structure of the present invention. Detailed Implementation

[0015] The present invention will be described in detail below with reference to the accompanying drawings and embodiments. It should be noted that, unless otherwise specified, the embodiments and features described in the present application can be combined with each other.

[0016] To make the technical solution of the present invention clearer, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0017] Reference Figures 1-2 A charging guidance method based on vehicle network data is implemented as follows: I. Implementation Prerequisites and Infrastructure Preparation Before implementing charging guidance, the deployment and debugging of core vehicle-to-everything (V2X) infrastructure must be completed to ensure the stability of data transmission, computing, and interaction. Specifically, this includes: Terminal equipment deployment: Equip electric vehicles participating in the guidance with on-board units (OBU) to ensure that they can collect vehicle battery status and location information in real time; deploy roadside units (RSU) at key nodes on urban roads (such as intersections and around charging stations), and activate 5G macro base stations within the area coverage to ensure smooth communication links between vehicles and roads (V2R) and roads and base stations (R2B).

[0018] System integration: Complete the integration of the charging station management system, electric vehicle battery management system (BMS), traffic signal control system and vehicle network cloud platform to achieve cross-system interoperability of data such as user charging records, battery cycle data, traffic light timing, and road congestion status.

[0019] Edge computing node configuration: Deploy edge servers on the RSU and 5G macro base station sides, complete the initial parameter configuration (such as initial reputation value weight, task offloading basic threshold) to ensure that the edge nodes have the ability to perform task computing, data temporary storage and reputation assessment.

[0020] II. Step-by-step implementation process (I) S1: Implementation of Vehicle Network Data Collection and Preprocessing Real-time collection of multi-dimensional data Based on the OBU, vehicle and user data are collected, including the vehicle's current location, real-time battery range, battery degradation-related charge and discharge cycle records, as well as user profile information such as historical charging time preferences and frequently used charging stations. The collection frequency is synchronized with the vehicle's driving status (e.g., location data is updated every 3-5 seconds).

[0021] Traffic data is collected by traffic signal control system and roadside sensors: real-time road congestion level (such as average vehicle speed on the road section), traffic light sequence (updated in time periods, such as 1 minute / time during morning peak and 5 minutes / time during off-peak), and charging station waiting queue length (obtained through charging station entrance cameras or queuing system).

[0022] Based on RSU and 5G macro base stations, collect communication and equipment data, including the computing load of RSU / macro base stations (such as the number of currently processed tasks), V2R and R2B transmission rates, and equipment operating status (such as whether there are communication interruption records).

[0023] Data preprocessing operations Statistical methods were used to clean the collected data: outliers (such as battery life data that exceeds the reasonable range or invalid transmission rates caused by momentary disconnection) were removed, and missing data was properly supplemented (such as by interpolation based on historical data from the same period).

[0024] Data standardization and time-series alignment: Convert data from different dimensions into a unified format (e.g., quantify congestion levels into fixed interval values), synchronize data sources based on timestamps (e.g., ensure that the vehicle location at the same time matches the traffic light status of the corresponding road segment), and finally form a standardized dataset that can be directly used for subsequent models.

[0025] (II) S2: Implementation of charging path planning for the improved A* algorithm Initial path planning triggered When an electric vehicle user initiates a charging request (such as by inputting the target charging type through the in-vehicle screen or APP), the system automatically calls the standardized dataset generated by S1, takes the vehicle's current location as the initial node and the charging station that meets the user's needs (such as a fast charging station or a station with a distance that matches the range) as the target node, and starts the improved A* algorithm for initial path planning.

[0026] Path cost calculation adaptation Given the cost G(i) calculation: Combine real-time traffic light timing data, convert the traffic light waiting time at each intersection into node weights, and add the actual travel time from the initial node to the current node (estimated based on the road congestion status) to obtain the actual value of G(i).

[0027] The estimated cost H(i) is calculated as follows: Based on the Euclidean geometric distance from the current node to the target charging station, the vehicle's driving direction is used to determine whether there is a U-turn behavior (such as the current road segment driving direction being opposite to the previous road segment). If the backtracking condition is triggered, a penalty term is added to obtain the actual value of H(i).

[0028] Initial path output The system selects the path with the smallest F(i) through a decision function as the initial charging path, avoids roads that have no impact on the optimal result (such as detours that do not alleviate congestion), and temporarily stores the initial path for final confirmation after collaboration with the reputation model and the two-layer optimization model.

[0029] (III) S3: Construction and Update Implementation of Two-Dimensional Reputation Value Model Edge server reputation management Initial calculation of reputation value: For each edge server, a direct reputation value is calculated based on its task completion status (such as on-time completion rate and calculation result accuracy). An indirect reputation value is obtained by weighted summation based on its own operating status (such as response latency, communication interruption frequency, and calculation error rate). The initial weights of the two are set according to the regional vehicle network load (such as increasing the weight of the indirect reputation value during high load periods to prioritize equipment stability).

[0030] Dynamic update execution: Reputation values ​​are updated at fixed time slices (e.g., 5 minutes / slice). First, the direct and indirect reputation values ​​are merged through the basic update formula, and then the update rate parameter is introduced to adjust the influence weight of the current time slice. If the server makes multiple calculation errors within a certain time slice, its reputation value will drop rapidly by increasing the update rate, making it easier for the system to quickly identify untrusted nodes.

[0031] Vehicle credit score management Model Construction: Referring to the edge server reputation value model framework, a new inter-vehicle mutual evaluation mechanism is added. After a vehicle completes charging or uploads data, it can evaluate the charging order of other vehicles (such as whether they cut in line) and the authenticity of the data uploaded (such as whether they falsely report the range). The evaluation results are included in the calculation of the vehicle's direct reputation value.

[0032] Weighting: Based on the reputation value of the edge server, assign its weight in vehicle reputation management (e.g., the evaluation weight of a high-reputation server is higher than that of a low-reputation server) to ensure the credibility of vehicle reputation statistics.

[0033] Incentive mechanism implementation For edge servers with high reputation scores: When receiving collaborative unloading requests, the system prioritizes assigning tasks to these servers, reducing their idle time and increasing their enthusiasm for participating in collaboration.

[0034] For vehicles with high credit scores: their priority is marked through the charging station management system, which can shorten waiting time when queuing at charging stations or give priority to using available charging piles, thus regulating vehicle behavior.

[0035] (iv) S4: Implementation of the two-level optimization model solution Model initialization Set the convergence coefficients for the upper and lower layers (determined based on the fluctuation range of regional traffic flow and the load stability of vehicle networking equipment), extract the current charging station waiting time and the initial operating status of the edge server / 5G macro base station from the standardized dataset of S1, determine the target charging station in combination with user needs, and complete the initial parameter configuration of the model.

[0036] Upper layer: Calculation of active charging guidance model on the transportation network side With the goal of "minimizing road travel time + charging station waiting time + charging time", the improved A* algorithm of S2 is used to replace the traditional algorithm. Combined with the reputation model of S3, high-reliability charging stations are selected (such as prioritizing stations with high reputation of surrounding edge servers and short waiting time) to generate a traffic flow information matrix (recording the number of vehicles expected to pass through each road segment).

[0037] Perform constraint verification: ensure that there are vehicles departing from the initial node, vehicles arriving at the target node, and the number of vehicles entering / exiting the intermediate nodes is balanced; at the same time, verify whether the vehicle arrival time matches the user's expectations (e.g., if the user sets "arrive within 1 hour", then filter the paths that meet this time constraint).

[0038] Lower layer: Calculation of task offloading model on the vehicle-to-everything (V2X) side The traffic flow information matrix output from the upper layer is used as input, and the task offloading scheme for each vehicle is allocated with the goal of "minimizing the computational cost + communication cost" (such as determining whether the navigation data processing task of a certain vehicle is offloaded to the nearest RSU or uploaded to the 5G macro base station).

[0039] Execution constraint verification: Ensure that the unloading decision is a 0-1 choice (i.e., a task is either unloaded or processed locally), the unloading ratio is within the range of 0-1 and the sum is 1; at the same time, ensure that the V2R and R2B transmission rates are not lower than the minimum threshold to avoid data transmission interruption.

[0040] Iterative solution and convergence judgment After the lower layer outputs the task offloading plan, it updates the operating status of the edge server and 5G macro base station (e.g., a certain RSU experiences a load increase due to receiving a large number of offloading tasks), and sends the updated status back to the upper layer.

[0041] Compare the differences in traffic flow information and task unloading costs between two consecutive iterations: if both meet the preset convergence coefficient requirements (e.g., the difference is less than 5%), the model is considered to have converged; otherwise, return to the upper layer to recalculate the path until convergence.

[0042] (v) S5: Charging guidance implementation Information push After the model converges, the system will push the optimal charging route (including real-time traffic light prompts, road congestion warnings, and charging station locations) and the task unloading plan (such as task unloading target nodes and transmission progress) simultaneously through the in-vehicle screen and the user's mobile APP to ensure that the user can keep track of the driving and task status in real time.

[0043] Dynamic adjustment and task coordination During vehicle operation, the system checks whether the route needs to be adjusted at a preset update frequency (e.g., every 10 minutes). If the vehicle is in an unstable period (e.g., during morning or evening rush hours) or a road accident occurs, the system immediately updates the road segment weights (e.g., the weight of the accident-affected road segment increases), calls the improved A* algorithm to replan the route, and pushes the updated route.

[0044] The vehicle-to-everything (V2X) network automatically completes task allocation according to the offloading plan: for example, offloading the real-time navigation data processing task of the vehicle to a high-reputation RSU along the route, and uploading the charging reservation data to the 5G macro base station to ensure efficient task processing, while avoiding excessive local computing load on the vehicle.

[0045] III. Implementation Process Monitoring and Safeguards Status monitoring: Throughout the boot process, the system monitors changes in the reputation value of edge servers, the transmission rate of vehicle networking devices, and road traffic conditions in real time. If any abnormality occurs (such as a sudden drop in the reputation value of a server or a sudden traffic jam on a road section), an early warning mechanism is triggered and relevant links are adjusted in priority (such as switching the unloading node or updating the path).

[0046] Data backup and traceability: Real-time backup of collected raw data, model calculation results, and path adjustment records facilitates subsequent analysis of guidance effects (such as the difference between the actual and estimated time of a certain path), and provides data support for the optimization of reputation models and algorithm parameters.

[0047] The embodiments described above are merely illustrative of several implementations of the present invention, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of the present invention. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the present invention, and these modifications and improvements all fall within the scope of protection of the present invention. Therefore, the scope of protection of this patent should be determined by the appended claims.

Claims

1. A charging guidance method based on vehicle network data, characterized in that, Includes the following steps: S1. Vehicle-to-Everything (V2X) Data Acquisition and Preprocessing: Collect data on electric vehicle user profiles, battery degradation data, real-time battery range, road traffic conditions (including traffic light timings), roadside units (RSUs), and 5G macro base station operation status. Clean and integrate the data using statistical data analysis methods to form a standardized dataset. S2. Charging Path Planning with Improved A* Algorithm: Based on the dataset from step S1, a dynamic path planning model is constructed, and the optimal charging path is calculated using the improved A* algorithm. The decision function of the improved A* algorithm is: Where G(i) is the actual cost from the initial node to the current node i, the traffic light mechanism is introduced to improve the cost, and the traffic light waiting time is set as the node weight and added to G(i); H(i) is the estimated cost from the current node i to the target charging station, and a backtracking penalty term is added to the traditional Euclidean geometric distance. S3. Construction and Update of a Two-Dimensional Reputation Value Model: Constructing an edge server reputation value model and a vehicle reputation value model; the edge server reputation value model is represented by a set. , Defined as a triple, , It is an edge server The logo, It is an edge server The reputation score, ranging from 0 to 1, is calculated based on the task completion status of the edge server. It is an edge server Indirect reputation value (range 0~1, calculated based on its own status); The formula for updating reputation score is: ,in , It is a weighting factor. ; Simultaneously, a dynamic update strategy is designed, with the formula as follows: , This represents the reputation score calculated within the current time slice. Represents the update rate. The larger the value, the greater the impact on the reputation value within the current time slice; the vehicle reputation value model refers to the edge server reputation value model, and incorporates a mutual evaluation mechanism between vehicles, and the edge server reputation value determines its weight in vehicle reputation value management; S4. Solving the two-layer optimization model: Establishing a two-layer optimization model consisting of an active charging guidance model on the transportation network side (upper layer) and a task offloading model on the vehicle network side (lower layer): Upper-level model: The traditional Dijkstra algorithm is replaced with the improved A* algorithm from step S2, with constraints including:

1. Road selection constraints In EV path selection, the initial node must have an electric vehicle departing from it, the target node D must have a vehicle arriving at it, and the number of vehicles entering and exiting intermediate nodes must be the same; therefore, the mathematical expression for the path selection constraint is: In the formula: The set of road nodes visited. The set of charging station nodes being visited; 2. Arrival Time Constraints Electric vehicle user arrival time constraints: In the formula: For departure time The expected arrival time; Lower-level model: The objective is to minimize the operating cost of the vehicle-to-everything (V2X) network. Constraints include: The offloading decision is a 0-1 constraint. The task unloading constraint, where the ratio should be in the range of 0-1, and the sum of the unloading ratios is 1, is as follows: Let the transmission from the electric vehicle to the roadside unit be v2r, and the transmission from the roadside unit to the 5G macro base station be r2b. The constraint that the transmission rates of v2r and r2b are not less than the minimum transmission rate is: ； Iterative solution: The traffic flow information matrix output by the upper layer is used as the input of the lower layer. The task unloading plan output by the lower layer updates the vehicle network operation status and is sent back to the upper layer. The iteration is repeated until the convergence condition is met. S5. Charging guidance execution: Push the optimal charging path and task offloading plan converged in step S4 to the electric vehicle user terminal, guide the vehicle to the target charging station, and simultaneously complete the vehicle network computing task offloading.

2. The charging guidance method based on vehicle network data as described in claim 1, characterized in that, The triggering condition for the backtracking penalty item in step S2 is: the current travel direction is opposite to the previous road segment (i.e., there is a U-turn behavior), and the penalty method is to add a penalty value with a fixed proportional coefficient on the basis of Euclidean distance.

3. The charging guidance method based on vehicle network data as described in claim 1, characterized in that, The calculation parameters of the edge server indirect reputation value Rindir mentioned in step S3 include server response latency, communication interruption frequency, and calculation error rate, which are obtained by weighted summation (the sum of the weights of each parameter is 1).

4. The charging guidance method based on vehicle network data as described in claim 1, characterized in that, The objective function of the upper-level model in step S4 is to minimize the total time consumption of "road travel time + charging station waiting time + charging time", and the objective function of the lower-level model is to minimize the total operating cost of "computation cost + vehicle network communication cost".

5. The charging guidance method based on vehicle network data as described in claim 1, characterized in that, The convergence condition for the iterative solution in step S4 is determined by the preset convergence coefficients between the upper and lower layers. When the difference in traffic flow information and the difference in task unloading cost between two consecutive iterations both meet the convergence coefficient requirements, it is determined to be converged.

6. The charging guidance method based on vehicle network data as described in claim 1, characterized in that, The incentive mechanism of the reputation value model in step S3 is as follows: edge servers with higher reputation values ​​are responded to first when requesting collaborative unloading; vehicles with higher reputation values ​​can obtain higher priority for charging station services.

Images