Vehicle-vehicle cooperative unloading method in high-speed vehicle networking scene

Abstract

The invention relates to the technical field of vehicle networking application, in particular to a vehicle-vehicle cooperative unloading method in a high-speed vehicle networking scene, which comprises the following steps of: firstly, calculating a dynamic communication rate between vehicles by using distance information, and calculating task transmission time by using the dynamic communication rate between the vehicles; a joint optimization model giving consideration to unloading reliability and task completion time is provided, and the problems that communication links are unstable, cooperative nodes are frequently disconnected due to frequent network topology changes, and the unloading reliability is poor are solved. And finally, enhancing population diversity by combining chaotic mapping initialization and a reverse learning strategy, and introducing a task failure penalty mechanism to optimize a fitness function, so as to improve a Kepler optimization algorithm, and effectively reduce the unloading overhead of the system under the condition of ensuring the stability and reliability of task unloading, thereby obtaining an optimal unloading decision.

Description

Technical Field

[0001] The present invention relates to the field of vehicle networking application technology, and in particular to a vehicle-to-vehicle collaborative unloading method in a high-speed vehicle networking scenario. Background Art

[0002] With the deep integration of mobile communication technology and artificial intelligence, new applications such as the Internet of Vehicles (IoV), the Internet of Things, and virtual reality are constantly emerging, significantly improving people's quality of life. In particular, IoV, as a key component of intelligent transportation systems, is gradually becoming a crucial technological means to promote social and economic development. Through efficient communication and collaborative computing between vehicles and their surroundings (such as roads, pedestrians, and transportation facilities), IoV provides powerful technical support for scenarios such as autonomous driving, intelligent traffic management, and remote vehicle monitoring. IoV is demonstrating significant application value in intelligent traffic management, reducing traffic accidents, and improving road traffic efficiency.

[0003] However, in high-speed Internet of Vehicles scenarios, due to the high-speed mobility of vehicles and the complex wireless propagation environment, collaborative offloading between vehicles will face problems such as unstable communication links caused by high-speed movement of vehicles, frequent disconnection of collaborative nodes due to frequent network topology changes, and poor offloading reliability. Summary of the Invention

[0004] The purpose of the present invention is to provide a vehicle-to-vehicle collaborative unloading method in a high-speed Internet of Vehicles scenario, which effectively reduces the system's unloading overhead while ensuring the stability and reliability of task unloading.

[0005] To achieve the above objectives, the present invention provides a vehicle-to-vehicle cooperative unloading method in a high-speed vehicle networking scenario, comprising the following steps:

[0006] Step 1: Use distance information to calculate the dynamic communication rate between vehicles, use the dynamic communication rate between vehicles to calculate the task transmission time, and define a joint optimization model that takes into account both offloading reliability and task completion time;

[0007] Step 2: Design a dynamic transmission scheduling mechanism based on a pre-generated communication rate matrix. This improves the Kepler optimization algorithm by combining chaotic map initialization with a reverse learning strategy to enhance population diversity and introducing a task failure penalty mechanism to optimize the fitness function.

[0008] Step 3: Use the improved Kepler optimization algorithm to find the optimal collaborative offloading decision.

[0009] Optionally, in step 1, the communication rate between the task vehicle m and the cooperative vehicle n is:

[0010]

[0011] Where B is the channel transmission bandwidth, N0 is the noise power spectrum density, and p m,n is the transmission power, H m,n is the channel gain, expressed as:

[0012]

[0013] Among them, |g m,n | 2 represents the power gain of small-scale fast fading, which follows an exponentially distributed random variation. ε0 represents the path loss constant. γ represents the attenuation exponent of the path loss.

[0014] Optionally, the joint optimization model is expressed as:

[0015]

[0016] Among them, constraint C1 indicates that the collaborative execution time of task k must be less than or equal to its tolerated delay, constraint C2 indicates that the reliability of the collaborative vehicles is constrained, constraint C3 indicates that task k can only be executed on one vehicle, constraint C4 indicates that the computing power for executing task k is constrained, and constraint C5 indicates that the distance between the two vehicles during the unloading process does not exceed their maximum communication range.

[0017] Optionally, the process of designing a dynamic transmission scheduling mechanism based on a pre-generated communication rate matrix in step 2 includes the following steps:

[0018] Construct a communication rate matrix with vehicle number as row and column index and time series as depth dimension;

[0019] Set up a dynamic time accumulation method, retrieve the corresponding communication rate in each time slice through a sliding window, dynamically accumulate the theoretical transmission data volume and monitor the link interruption in real time.

[0020] Optionally, step 2 combines chaotic map initialization with a reverse learning strategy to enhance population diversity, and introduces a task failure penalty mechanism to optimize the fitness function to improve the Kepler optimization algorithm, including the following steps:

[0021] Planet position initialization combining chaotic mapping and reverse learning strategy;

[0022] Perform planet position updates and select fitness functions.

[0023] Optionally, the execution process of step 3 includes the following steps:

[0024] Step 3.1: Define the planetary gravity;

[0025] Step 3.2: Update celestial body positions;

[0026] Step 3.3: Update the distance from the sun;

[0027] Step 3.4: Update the optimal fitness using the elite strategy.

[0028] This paper provides a method for cooperative vehicle offloading in high-speed connected vehicle scenarios. It first uses distance information to calculate the dynamic communication rate between vehicles, and then uses this to calculate task transmission time. A joint optimization model that balances offloading reliability and task completion time is proposed to address the issues of unstable communication links, frequent disconnection of cooperative nodes due to frequent network topology changes, and poor offloading reliability. Finally, by combining chaotic map initialization with a reverse learning strategy to enhance population diversity and introducing a task failure penalty mechanism to optimize the fitness function, the Kepler optimization algorithm is improved. This effectively reduces the system's offloading overhead while ensuring the stability and reliability of task offloading, thereby achieving the optimal offloading decision. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0030] Figure 1 It is a schematic diagram of a vehicle-to-vehicle cooperative unloading system model of the present invention's vehicle-to-vehicle cooperative unloading method in a high-speed vehicle networking scenario.

[0031] Figure 2 It is a schematic flow chart of the improved Kepler optimization algorithm of the present invention. DETAILED DESCRIPTION

[0032] The following describes embodiments of the present invention in detail, examples of which are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are intended to be used to explain the present invention, and are not to be construed as limiting the present invention.

[0033] The present invention provides a vehicle-to-vehicle cooperative unloading method in a high-speed vehicle networking scenario, comprising the following steps:

[0034] Step 1: Use distance information to calculate the dynamic communication rate between vehicles, use the dynamic communication rate between vehicles to calculate the task transmission time, and define a joint optimization model that takes into account both offloading reliability and task completion time;

[0035] Step 2: Design a dynamic transmission scheduling mechanism based on a pre-generated communication rate matrix. This improves the Kepler optimization algorithm by combining chaotic map initialization with a reverse learning strategy to enhance population diversity and introducing a task failure penalty mechanism to optimize the fitness function.

[0036] Step 3: Use the improved Kepler optimization algorithm to find the optimal collaborative offloading decision.

[0037] The following is further explained with reference to specific embodiments and execution steps:

[0038] See also Figure 1 , Figure 1 This is a schematic diagram of the vehicle-to-vehicle cooperative unloading system model of the present invention, which is a scenario of dynamic mobile cooperation over time in a high-speed mobile scenario, in which the task vehicle (red / green) and the cooperative vehicle share tasks in real time through the V2V link and are wirelessly connected to base stations on both sides of the road.

[0039] In step 1, the dynamic communication rate between vehicles is calculated using distance information. The specific method is:

[0040] Assume that the task offloading start time is time 0, and the speed and direction of the vehicle at time t are constants. Use φ = {x(t), y(t), θ(t), v(t)} to represent the vehicle's mobility attributes, where x and y represent the vehicle's horizontal and vertical coordinates, θ represents the vehicle's moving direction, and v represents the vehicle's moving speed. Define the maximum communication distance between vehicles as R max When unloading tasks, it is necessary to consider the unloading time t mk,n The communication distance between the two vehicles cannot exceed the maximum communication distance R max .

[0041] x n (t) = x n (0)+t*v(t)*cosθ(t)

[0042] During the unloading time, the distance between the two vehicles can be expressed by the following formula:

[0043]

[0044] Among them, x m (t) represents the horizontal coordinate of vehicle m at time t, x n (t) represents the horizontal coordinate of vehicle n at time t, y m (t) represents the ordinate of vehicle m at time t, y n (t) represents the ordinate of vehicle n at time t, and time t is the unloading time t of the task mk,n Inner, that is, t∈[0,t mk,nIn order to ensure that the distance between the two vehicles does not exceed their maximum communication range during the unloading process, the following restrictions need to be made:

[0045] l m,n (t)≤R max

[0046] Because the data from the task vehicle needs to be transmitted through the channel for calculation when it is offloaded to the cooperative vehicle, the present invention adopts the orthogonal frequency division multiplexing (OFDM) technology to divide the spectrum resources into mutually orthogonal subcarriers and allocate them to the vehicle users in the network. The communication rate between the task vehicle and the cooperative vehicle is:

[0047]

[0048] Where B is the channel transmission bandwidth, N0 is the noise power spectrum density, and p m,n is the transmission power, H m,n The channel gain can be expressed as:

[0049]

[0050] Among them, |g m,n | 2 represents the power gain of small-scale fast fading, which follows an exponentially distributed random variation. ε0 represents the path loss constant. γ represents the attenuation exponent of the path loss.

[0051] In addition, the dynamic communication rate between vehicles is used to calculate the task transmission time. The specific method is:

[0052] The transmission time of task k is calculated by the following formula:

[0053]

[0054] In step (1), a joint optimization model that takes into account both offloading reliability and task completion time is defined. The specific method is:

[0055] We define the collaborative computation success rate as a measure of collaborative vehicle reliability. Assume that collaborative vehicle n has collaboratively computed U tasks, where the actual completion time of the μth task is within its tolerance time T. mk If the cooperative vehicle completes the task within 10 seconds, the calculation is considered successful. Otherwise, the cooperative vehicle calculation fails. Therefore, there are μ successful tasks among U tasks, and its reliability can be expressed as:

[0056]

[0057] Assume that the probability of cooperative vehicle n actually failing in cooperative calculation is p n, each collaborative vehicle is independent of each other, and the failure probability of collaborative calculation is also independent of each other. After each collaborative calculation is completed, the collaborative calculation success rate η is updated in real time according to the collaborative calculation results. n .p n is non-a priori, but when the computational tasks of cooperative vehicles tend to infinity, its reliability η n tends to 1-p n Therefore, the reliability of cooperative vehicles is represented by the set η={η1,η2,...,η n}.

[0058] In a high-speed IoV scenario, a base station connects to a series of high-speed vehicles below. In this scenario, vehicles can be divided into two types. One type is mission vehicles (MVs), which generate computationally intensive tasks, represented by the set M = {1, 2, ..., m}. While these vehicles possess a certain level of computing power, their own processing capabilities cannot meet the high-performance requirements of the tasks, and therefore they need to request that some of the computational tasks be offloaded to other vehicles. The other type is cooperative vehicles (CVs), represented by the set N = {1, 2, ..., n}, which currently have no computationally intensive tasks to process. The computing resources of these cooperative vehicles are idle, allowing them to collaborate with other mission vehicles to process computationally intensive tasks. During a certain time slot, if a mission vehicle has a computationally intensive task, the cooperative vehicles or base stations within its range can jointly process the computationally intensive task and return the processing results to the corresponding mission vehicle.

[0059] Assume that the base station can sense the information of all vehicles and deploy a controller in the base station to control the entire collaborative unloading process. In the task scheduling process, the task set of the task vehicle is represented by q = {1, 2, ..., k}, and each subtask is non-subdividable. Each subtask is represented by c k ={d k ,s k ,T k} means, d k Indicates the size (in bits) of the input data required to complete subtask k, s k Indicates the number of calculation cycles required to execute 1 bit of data, T k Indicates the tolerance time of the task. Each subtask can only be processed in two ways: one is to calculate it by the task vehicle itself, and the other is to offload the subtask to the cooperative vehicle for calculation. k,n ∈{0,1}, k∈q, n∈N represents the unloading decision of the task vehicle. If x k,0 =1 means that the task vehicle MV will perform the calculation task locally, otherwise x k,0 = 0 means that the mission vehicle MV does not perform the computation task locally.k,n =1 means that the task vehicle MV offloads the computing task to the cooperative vehicle CVn, otherwise x k,n = 0 means that the task vehicle MV does not offload the computing task to the cooperative vehicle CVn. Therefore, the subtask offloading decision set of the task vehicle MV is X = (x1, x2, ..., x k ).

[0060] Next, we will mainly discuss the task vehicle computing and collaborative vehicle computing for computationally intensive and time-sensitive application tasks. The specific computing models are as follows.

[0061] 1) Mission vehicle calculation. Assume that the mission vehicle has a calculation task k frequency f k,0 , local execution task q k The delay is:

[0062]

[0063] The time of transmitting the calculation results is ignored in this invention because the data size of the results is usually negligible compared with the input data of the task for many new applications. Indicates the time when it starts processing task k, which depends on the latest completion time of the predecessor subtask of task k and the time when it is occupied, that is,

[0064]

[0065] Among them, r i,k =1 means that task i is the direct predecessor of task k, x i,0 represents the offloading decision of task i, x i,0 =1 means task i is calculated locally, x i,0 =0 means that task i is not calculated locally. represents the local computing time of task i, represents the time for cooperative vehicle j to calculate task i, Indicates the time the vehicle is occupied. So the time it takes for the vehicle to complete the task is:

[0066]

[0067] 2) Collaborative vehicle calculation. Assume that the frequency of collaborative vehicles performing task k is f k,n , the delay of cooperative vehicle computing task k is:

[0068]

[0069] The transmission delay from task k to cooperative vehicle n is:

[0070]

[0071] The start time of cooperative vehicle calculation task k is:

[0072]

[0073] in, is the occupied time of cooperative vehicle n, and i is the latest completed predecessor subtask of task k. Therefore, the collaborative execution completion time of task k is the sum of the start time and the collaborative calculation time of cooperative vehicle n, and its expression is:

[0074]

[0075] In summary, the latency and reliability of offloading task k are:

[0076]

[0077] Taking into account the time to complete the subtask and the reliability of the cooperative vehicle, the overhead of offloading the computation task k to the task vehicle can be expressed as:

[0078]

[0079] Among them, β is the weight coefficient and β∈[0,1]. By adjusting the weight coefficient, the priority of time and reliability can be adjusted. is the normalized value of the task unloading time, Rel is the normalized value of reliability. max and Rel min The highest and lowest reliability values ​​in the vehicle set.

[0080]

[0081] Under the constraints of latency, computing power, communication distance, and reliability, the offloading decision and computing frequency are optimized to minimize the overhead of executing tasks. The final joint optimization model is:

[0082]

[0083] Among them, constraint C1 indicates that the collaborative execution time of task k must be less than or equal to its tolerated delay, constraint C2 indicates that the reliability of the collaborative vehicles is constrained, constraint C3 indicates that task k can only be executed on one vehicle, constraint C4 indicates that the computing power for executing task k is constrained, and constraint C5 indicates that the distance between the two vehicles during the unloading process does not exceed their maximum communication range.

[0084] In step 2, a dynamic transmission scheduling mechanism based on a pre-generated communication rate matrix is ​​designed. The specific method is:

[0085] To address the frequent interruptions of communication links caused by high-speed vehicle movement, this invention constructs a matrix of inter-vehicle communication rates at all times. During task scheduling, this matrix is ​​used to dynamically avoid communication interruption windows, thereby reducing the transmission failure rate. The specific process is as follows:

[0086] 1) Communication rate matrix construction

[0087] Based on the vehicle's mobility properties, φ = {x(t), y(t), θ(t), v(t)}, the instantaneous distance between any two vehicles within a continuous time window is calculated. By combining the path loss model with small-scale random fading, the variation of the channel gain is quantified. Shannon's theorem is then introduced to dynamically map the nonlinear impact of distance variation on the communication rate. Based on this, a discretized time step strategy is employed to construct a communication rate matrix with vehicle number as the row and column index and time series as the depth dimension.

[0088] 2) Dynamic time accumulation method

[0089] When a task offload request arrives, a binary search is performed to determine the nearest reference time node based on the correspondence between the transmission start time and the time slices in the rate matrix. For situations where the transmission start time may not be aligned between two adjacent time slices, the remaining data transmission amount within the transition interval is calculated based on the current channel rate. If the remaining data transmission amount is greater than or equal to the calculated task data amount, the precise segment transmission time is directly returned; otherwise, the unfinished portion is progressively accumulated in the subsequent complete time slices according to the time sequence.

[0090] During this process, the algorithm uses a sliding window to retrieve the corresponding communication rate for each time slice, dynamically accumulating the theoretical transmission data volume and monitoring link interruptions in real time. When the accumulated data approaches the task data size, the algorithm reversely estimates the transmission completion time based on the remaining demand in the current time slice and the real-time rate. This approach not only avoids the prediction error accumulation caused by time slice division in traditional discretization methods, but also effectively addresses nonlinear fluctuations in channel conditions caused by rapid vehicle movement.

[0091] Furthermore, the Kepler optimization algorithm is improved by combining chaotic map initialization with reverse learning strategy to enhance population diversity and introducing a task failure penalty mechanism to optimize the fitness function. The specific process is as follows:

[0092] 1) Planet position initialization combining chaos mapping and reverse learning strategy

[0093] The initial planetary position is represented by X=(x1,x2,...,x k ), where x k∈[1,n] represents the assigned vehicle number for task k. The initial solution of the planetary position plays a crucial role in solving the optimization algorithm. In order to improve the solution quality of the algorithm, an initialization method combining chaos mapping and reverse learning is introduced to optimize the initial solution of the planetary position.

[0094] Generate ergodic initial planets based on Tent chaos map:

[0095]

[0096] Where η∈(0,1) is the chaos parameter, which generates uniformly distributed solutions through iteration; chaos is the chaos value. The chaotic initial planet positions are further generated by combining the chaos value as shown below:

[0097]

[0098] in, Indicates the chaotic position of planet k, up bound and low bound Represent the upper and lower limits of x respectively. In order to improve the diversity of the initial positions of the planets, a reverse learning mechanism is introduced. As shown below, reverse learning generates mirror planet positions:

[0099]

[0100] Among them, r is a random value between [0,1]; x reverse is the mirror planet position, when x reverse When the value of exceeds the constraint boundary, the mirror planet position is:

[0101] x reverse =rand(low bound ,up bound )

[0102] The planet positions generated by Tent Chaos Map and reverse learning are combined to form a new population of planetary initial positions. Then, by calculating the fitness of the new population of planetary initial positions, half of the planets with the best fitness are selected as the final planetary initial solution and the corresponding sun is selected.

[0103] 2) Planet update and fitness function

[0104] First, the planet is mapped as an unloading strategy, moving around the current optimal solution. The orbital eccentricity is dynamically adjusted with iteration. The large eccentricity in the early stage promotes global exploration, while the small eccentricity in the later stage strengthens local development. New planet positions are generated through elite retention and hybrid operations. The objective function of the present invention is:

[0105] F k =C k +δ·ρ

[0106] Where δ∈{0,1} represents whether the offloading task delay meets the time limit. If δ=1, it means that the task offloading delay does not meet the time limit, and a task failure penalty is imposed on the current fitness value. ρ represents the penalty value for exceeding the time limit.

[0107] In step 3, the improved Kepler optimization algorithm is used to find the optimal collaborative offloading decision. The specific process is as follows:

[0108] 1) Definition of planetary gravity

[0109] The main reason planets orbit the sun is that gravity keeps them in orbit. Each planet has its own gravity, and according to the law of universal gravitation, the sun X Sun With any planet X pl The gravitational force is:

[0110]

[0111] The value of ε is very small, μ(t) is an exponentially decreasing function over time, and r1 randomly takes values ​​between [0,1], which can make the value of universal gravitation have greater changes during the optimization process. m pl and They represent the normalized values ​​of the solar mass, planetary mass, and Euclidean distance between the sun and the planet, respectively.

[0112] 2) Update celestial body positions

[0113] The Kepler optimization algorithm simulates the changing pattern of the distance between the sun and the planets. When the planets are close to the sun, the algorithm will focus on optimizing the development operation, while when the planets are farther away, the algorithm will optimize the detection operation. Its position update expression is:

[0114]

[0115] in, represents the position of the planet at time t+1. Indicates the current optimal sun position. F is used to control the search direction of the algorithm. Randomly select a value between [0,1].

[0116] 3) Update the distance from the sun

[0117] To further improve planetary exploration and development, this algorithm dynamically adjusts the optimization strategy by adjusting the parameter h to meet the needs of different stages. The dynamic position update expression is shown below.

[0118]

[0119] in, and is a value randomly selected from the population. h is an adaptive factor used to control the distance between the sun and the planets. This formula is used interchangeably with the update expression for updating the celestial body position in step 2).

[0120] 4) Elite Strategy

[0121] After the planet position is updated, if its fitness is smaller, the optimal fitness is updated.

[0122] The flowchart of the improved Kepler optimization algorithm in the present invention is as follows: Figure 2 As shown, the Kepler optimization algorithm has advantages such as physical intuitiveness, efficient convergence, and dynamic adaptability. However, because the algorithm randomly generates the initial positions of the planets, it is prone to obtaining local optimal solutions. Therefore, this chapter combines chaos mapping with a reverse learning strategy to improve the initial position generation of the Kepler optimization algorithm, further enhancing the global and local search capabilities in multi-objective optimization problems and finding the optimal solution in this search space.

[0123] In summary, compared with the prior art, the present invention has the following beneficial effects:

[0124] A joint optimization model that balances offloading reliability and task completion time is proposed to address the issues of unstable communication links, frequent disconnection of collaborative nodes caused by frequent network topology changes, and poor offloading reliability. While ensuring the stability and reliability of task offloading, the system's offloading overhead is effectively reduced. By combining chaotic map initialization with a reverse learning strategy to enhance population diversity and introducing a task failure penalty mechanism to optimize the fitness function, the KOA algorithm is improved to achieve optimal offloading decisions.

[0125] The above disclosure is merely one or more preferred embodiments of the present invention, and certainly cannot be used to limit the scope of the present invention. A person skilled in the art can understand that all or part of the processes of the above embodiments and equivalent changes made in accordance with the claims of the present invention still fall within the scope of the invention.

Claims

1. A vehicle-to-vehicle cooperative unloading method in a high-speed vehicle networking scenario, characterized in that: The following steps are involved: Step 1: Use distance information to calculate the dynamic communication rate between vehicles, use the dynamic communication rate between vehicles to calculate the task transmission time, and define a joint optimization model that takes into account both offloading reliability and task completion time; Step 2: Design a dynamic transmission scheduling mechanism based on a pre-generated communication rate matrix. This improves the Kepler optimization algorithm by combining chaotic map initialization with a reverse learning strategy to enhance population diversity and introducing a task failure penalty mechanism to optimize the fitness function. Step 3: Use the improved Kepler optimization algorithm to find the optimal collaborative offloading decision.

2. The vehicle-to-vehicle cooperative unloading method in a high-speed vehicle networking scenario as claimed in claim 1, characterized in that: In step 1, the communication rate between task vehicle m and cooperative vehicle n is: Where B is the channel transmission bandwidth, N0 is the noise power spectrum density, and p m,n is the transmission power, H m,n is the channel gain, expressed as: Among them, |g m,n | 2 It represents the power gain of small-scale rapid fading, which is a random variation that obeys the exponential distribution; ε0 represents the path loss constant, and γ represents the attenuation exponent of the path loss.

3. The vehicle-to-vehicle cooperative unloading method in a high-speed vehicle networking scenario as claimed in claim 2, characterized in that: The joint optimization model is expressed as: Among them, constraint C1 indicates that the collaborative execution time of task k must be less than or equal to its tolerated delay, constraint C2 indicates that the reliability of the collaborative vehicles is constrained, constraint C3 indicates that task k can only be executed on one vehicle, constraint C4 indicates that the computing power for executing task k is constrained, and constraint C5 indicates that the distance between the two vehicles during the unloading process does not exceed their maximum communication range.

4. The vehicle-to-vehicle cooperative unloading method in a high-speed vehicle networking scenario as claimed in claim 3, characterized in that: The process of designing a dynamic transmission scheduling mechanism based on a pre-generated communication rate matrix in step 2 includes the following steps: Construct a communication rate matrix with vehicle number as row and column index and time series as depth dimension; Set up a dynamic time accumulation method, retrieve the corresponding communication rate in each time slice through a sliding window, dynamically accumulate the theoretical transmission data volume and monitor the link interruption in real time.

5. The vehicle-to-vehicle cooperative unloading method in a high-speed vehicle networking scenario as claimed in claim 4, characterized in that: Step 2 combines chaos map initialization with reverse learning strategy to enhance population diversity, and introduces task failure penalty mechanism to optimize fitness function to improve Kepler optimization algorithm, including the following steps: Planet position initialization combining chaotic mapping and reverse learning strategy; Perform planet position updates and select fitness functions.

6. The vehicle-to-vehicle cooperative unloading method in a high-speed vehicle networking scenario as claimed in claim 5, characterized in that: The execution process of step 3 includes the following steps: Step 3.1: Define the planetary gravity; Step 3.2: Update celestial body positions; Step 3.3: Update the distance from the sun; Step 3.4: Update the optimal fitness using the elite strategy.

Images