Centralized management method and system for smart vehicles

Abstract

The invention relates to the field of logistics vehicle management, in particular to a centralized management method and system for intelligent vehicles. The invention discloses a centralized management system for smart vehicles. The centralized management system comprises a task receiving module, a portrait construction module, a vehicle screening module, a sign-in verification module and a queue management module. According to the method, the vehicle multi-dimensional portrait model is constructed, and the transportation cost, the transportation efficiency and the like are subjected to multi-dimensional matching calculation based on the multi-objective optimization algorithm, so that intelligent vehicle dispatching decision-making from experience driving to data driving is realized; according to the method, the historical operation data, the real-time state and the driver behavior characteristics of the vehicle are comprehensively considered, a global optimal vehicle dispatching scheme can be output, the vehicle dispatching accuracy and the resource utilization efficiency are remarkably improved, and the limitation that a traditional method only depends on a single factor for dispatching is overcome.

Description

Technical Field

[0001] This invention relates to the field of logistics vehicle management, and more specifically to a centralized management method and system for intelligent vehicles. Background Technology

[0002] With the rapid development of modern logistics, especially in the transportation of large materials such as minerals and castables, the requirements for transportation efficiency, cost control and process coordination are increasing. Traditional logistics vehicle management models usually rely on manual dispatching, telephone communication and simple information systems, which are difficult to cope with complex transportation scenarios with large scale, multiple nodes and high real-time requirements.

[0003] Currently, common vehicle management systems typically possess basic functions such as vehicle information management, task assignment, and status tracking. However, these systems have several significant drawbacks: First, in the vehicle dispatching stage, most systems rely solely on single or limited static factors such as vehicle availability and load matching for dispatching decisions. They lack in-depth integration and intelligent analysis of multi-dimensional information, including historical vehicle performance and driver behavior preferences, resulting in dispatching schemes that are often not globally optimal and vehicle resources cannot be utilized efficiently. Second, in the vehicle entry management stage, a rigid "first-come, first-served" queuing model is commonly adopted. This fails to incorporate factors such as the urgency of the task, the real-time location of the vehicle, and the busy status of resources within the plant into the dynamic adjustment criteria. This leads to unreasonable phenomena such as high-priority task vehicles queuing and the coexistence of idle or congested plant resources, seriously affecting overall logistics efficiency.

[0004] Therefore, this invention proposes a centralized management method and system for intelligent vehicles, which enables integrated centralized control of the entire logistics transportation chain, thereby comprehensively improving the intelligence level and operational efficiency of logistics transportation. Summary of the Invention

[0005] This invention constructs a multi-dimensional vehicle profile model and performs multi-dimensional matching calculations on transportation costs and efficiency based on a multi-objective optimization algorithm, realizing intelligent vehicle dispatching decisions from experience-driven to data-driven. By comprehensively considering historical vehicle operation data, real-time status, and driver behavior characteristics, it can output the globally optimal dispatching plan, significantly improving the accuracy of vehicle dispatching and resource utilization efficiency, and overcoming the limitations of traditional methods that rely on only a single factor for dispatching.

[0006] A centralized management method for intelligent vehicles, comprising: Receive transportation tasks and obtain a transportation task list, wherein the transportation tasks include cargo information, destination factory area, required arrival time and batch type; A multi-dimensional profile model is constructed for any vehicle, which includes the vehicle's static attributes, dynamic performance tags, and historical behavior tags; after any vehicle completes a transportation task, the multi-dimensional profile model of that vehicle is updated. For any given transportation task, based on the transportation task and the multi-dimensional profile model of currently available vehicles, candidate vehicles that meet preset conditions are selected for the transportation task; a multi-objective optimization function is constructed with transportation cost, transportation efficiency, time matching degree, route familiarity and resource utilization rate as the core, and the demand of the transportation task is matched with the multi-dimensional profile model of the candidate vehicles in a multi-dimensional way to calculate the vehicle dispatch plan and assign the transportation task to the corresponding driver. Receive check-in requests initiated by dispatched vehicles in designated areas, verify the vehicle's financial audit status and task validity, and after verification, add the vehicle to the dynamic entry queue and assign a queue number to the vehicle. The system monitors the real-time status of each vehicle in the dynamic entry queue, dynamically adjusts the priority order of vehicles in the queue, and issues a departure command to the vehicle ranked first. After a vehicle enters the site, its status is automatically updated, and subsequent weighing, loading and unloading, and factory exit processes are triggered.

[0007] Preferably, a multi-dimensional vehicle profile model is constructed, and the specific operations are as follows: Collect basic static attributes of vehicles and drivers, including vehicle license plate number, vehicle type, approved load capacity, driver identity information, and driver's license type; The process of acquiring historical vehicle operation data and generating dynamic performance tags includes: extracting cargo types and transportation frequencies from historical transportation records, calculating the proportion of transportation times for each cargo type, marking cargo types with proportions exceeding a preset frequency threshold as frequently transported cargo for the vehicle, and generating frequently transported cargo tags; calculating the ratio of travel time to distance in historical tasks to generate average travel speed tags; recording the start and end times of each loading and unloading operation, calculating the operation duration, calculating the average loading and unloading time by cargo type, and generating loading and unloading efficiency tags; comparing the required arrival time and actual arrival time of historical tasks, statistically analyzing the proportion of tasks completed on time, and generating on-time rate scoring tags. Based on preset behavioral rules and historical vehicle operation data, historical behavior tags are generated. The generation process of historical behavior tags includes: calculating credit rating tags using a credit scoring algorithm based on information such as the historical task completion quality, violation records, and customer complaint records of the vehicle and driver; recording and classifying violations in the vehicle's historical operation, including speeding, deviating from the route, and failure to operate according to procedures, and generating corresponding violation tags for each violation; extracting high-frequency path segments and their distribution areas from the driver's historical driving trajectory to generate path familiarity tags. The basic static attributes, dynamic performance tags, and historical behavior tags are integrated to form a multi-dimensional profile model of the vehicle.

[0008] Preferably, based on the transportation task and the multi-dimensional profile model of currently available vehicles, candidate vehicles that meet preset conditions are selected for the transportation task. The specific operation is as follows: Extract preliminary screening criteria from the current transportation task. The preliminary screening criteria include: the type of goods to be transported, the total weight of the goods, the address of the destination factory area, and the delivery deadline. Based on the preliminary screening criteria, the static attributes of each vehicle are checked, and vehicles that simultaneously meet all of the following basic conditions are selected to form a candidate vehicle set: a) The vehicle's rated load capacity is greater than or equal to the total weight of the goods in this transportation task. This condition is limited to the current transportation task being a batch of one vehicle per batch. b) The physical characteristics of the vehicle type must match the requirements of the cargo type; The candidate vehicle set and its corresponding multi-dimensional profile data are output for subsequent multi-objective optimization function calculation and vehicle dispatch decision-making.

[0009] Preferably, a multi-objective optimization function is constructed with total transportation cost, overall efficiency, and vehicle queuing time as the core. The demand for the transportation task is matched and calculated with the real-time status, real-time location, and multi-dimensional profile data of the candidate vehicles in a multi-dimensional manner, and a vehicle dispatching plan is output. The specific operation is as follows: A multi-objective optimization function is established, comprising the following five core evaluation dimensions. For any given transportation task, multi-dimensional calculations are performed on several candidate vehicles corresponding to that task. The specific calculation methods for each dimension are as follows: Transportation cost calculation: The shortest path distance from the pickup plant to the destination plant is calculated using a route planning algorithm. Combined with the historical average energy consumption data per 100 kilometers of this vehicle under the same road conditions and the current energy price, the estimated transportation cost is calculated. Transportation efficiency calculation: The total task time is broken down into three parts: pickup travel time, delivery travel time, and loading / unloading time. The estimated pickup travel time is calculated based on the vehicle's real-time location, distance to the pickup area, and average vehicle speed. The delivery travel time is calculated based on the shortest path distance and average vehicle speed. The average loading / unloading time for the current cargo type is directly used as the estimated loading / unloading time based on the loading / unloading efficiency tags in the vehicle's multi-dimensional profile. The total estimated task time is the sum of the pickup travel time, delivery travel time, and loading / unloading time. Time matching dimension calculation: Based on the estimated total task time, the estimated arrival time is obtained, and the absolute time difference between the estimated arrival time and the delivery deadline is calculated; Path familiarity dimension calculation: The trajectory similarity algorithm based on the proportion of common road segments is used to calculate the overlap between the task-planned path and the driver's historical driving path, and the ratio of the length of the longest common subsequence of the two paths to the total path length is calculated as the path familiarity. Resource utilization rate calculation: Calculate the ratio of the weight of the goods to the vehicle's rated load capacity to obtain the load utilization rate value; For each evaluation dimension, the maximum and minimum values ​​of the current candidate vehicle set in that dimension are extracted; the original values ​​of each evaluation dimension are converted into standardized scores in the range of 0 to 1 using a linear scaling transformation method. The weight coefficients of each dimension are dynamically configured according to the characteristics of the transportation task. The comprehensive score of each candidate vehicle is calculated by weighted summation. All candidate vehicles are sorted in descending order according to the comprehensive score. If the batch type of the current transportation task is one vehicle per batch, the vehicle with the highest comprehensive score is selected as the optimal dispatch plan. If the batch type of the current transportation task is sequential batching, vehicles are selected in descending order of comprehensive score until the sum of the approved load capacity of the selected vehicles is not less than the total weight of the goods. Output a vehicle dispatch plan, which includes recommended vehicles and corresponding multi-dimensional calculation results.

[0010] Preferably, the vehicle's financial audit status and task validity are verified. Once the verification is successful, the vehicle is added to the dynamic entry queue and assigned a queue number. The specific operation is as follows: When a vehicle enters the preset electronic fence geographical area, the driver manually initiates a check-in request. This request must include at least the vehicle identification information, waybill number, and real-time location coordinates. Upon receiving the check-in request, the following verification process is executed: Initiate a real-time query to the financial management system to obtain the current financial review status corresponding to the waybill number; verify whether the binding relationship between the waybill and the vehicle is valid, and confirm that the waybill status is dispatched and awaiting execution; If the financial audit status is "audited" and the waybill is valid, the verification is considered successful; if the financial status is "not audited" or "audit rejected", a check-in failure response containing the specific reason for rejection will be immediately returned to the vehicle terminal. For vehicles that pass the verification, the task priority is calculated based on the following objective factors: Obtain the current time, the required arrival time for the task, and the estimated arrival time based on the vehicle's performance of the current transportation task; Calculate the difference between the estimated arrival time and the required arrival time, then proportionally calculate the time urgency value by dividing the difference by the total available time, which is used as the task priority score. Vehicle information is added to a dynamic entry queue, which is implemented using a data structure based on task priority. The vehicle's ranking in the queue is determined by its task priority score; the higher the score, the higher the ranking.

[0011] Preferably, the order of the dynamic entry queue is dynamically updated based on real-time data, and the specific operation is as follows: Real-time location coordinates and speed of each vehicle in the dynamic entry queue are collected, and loading and unloading efficiency labels are extracted from the multi-dimensional vehicle profile model. Establish a dynamic ranking adjustment mechanism, making comprehensive decisions based on the following three core factors: Task priority factor: Directly use task priority scoring; Real-time distance factor: Based on the vehicle's real-time location and speed, calculate the estimated time it will take for the vehicle to reach the pickup area; Loading and unloading efficiency factors: Obtain the average loading and unloading time of vehicles for the current type of goods; Set a fixed time interval, and at regular intervals, perform the following sorting adjustment operation: Based on the maximum and minimum values ​​of each factor, a linear transformation method is used to convert the original values ​​of each vehicle on each factor into a standardized score between 0 and 1. The weighted calculation is performed using fixed weight coefficients, with task priority having the highest weight, real-time distance having the second highest weight, and loading / unloading efficiency having the lowest weight; finally, a dynamic ranking score is obtained for each vehicle; and all vehicles in the queue are re-ranked based on the dynamic ranking score.

[0012] A centralized management system for intelligent vehicles, comprising: The task receiving module is used to receive transportation tasks and obtain a transportation task list. The profile building module is used to build a multi-dimensional profile model for any vehicle; after any vehicle completes a transportation task, the multi-dimensional profile model of that vehicle is updated. The vehicle screening module includes a preliminary screening unit and an optimization calculation unit. The preliminary screening unit is used to screen candidate vehicles that meet preset conditions for any given transportation task, based on the transportation task and the multi-dimensional profile model of currently available vehicles. The optimization calculation unit is used to construct a multi-objective optimization function with transportation cost, transportation efficiency, time matching degree, route familiarity, and resource utilization rate as the core, to perform multi-dimensional matching calculations between the demand of the transportation task and the multi-dimensional profile model of the candidate vehicles, output a dispatch plan, and assign the transportation task to the corresponding driver. The check-in verification module is used to receive check-in requests initiated by dispatched vehicles in designated areas, verify the vehicle's financial audit status and task validity, and add the vehicle to the dynamic entry queue after the verification is passed, and assign a queue number to the vehicle. The queue management module is used to monitor the real-time status of each vehicle in the dynamic entry queue, dynamically adjust the priority order of vehicles in the queue, and issue a departure command to the vehicle with the highest priority. After a vehicle enters the site, its status is automatically updated, and subsequent weighing, loading and unloading and factory exit processes are triggered.

[0013] The present invention has the following advantages: 1. This invention constructs a multi-dimensional vehicle profile model and performs multi-dimensional matching calculations on transportation costs and efficiency based on a multi-objective optimization algorithm, realizing intelligent vehicle dispatching decisions from experience-driven to data-driven. It comprehensively considers historical vehicle operation data, real-time status, and driver behavior characteristics, and can output the globally optimal dispatching plan, significantly improving the accuracy of vehicle dispatching and resource utilization efficiency, and overcoming the limitations of traditional methods that rely on only a single factor for dispatching.

[0014] 2. This invention introduces an entry queuing mechanism based on real-time data updates, incorporating multiple factors such as task priority, real-time vehicle location, and loading / unloading efficiency into the queuing algorithm. This achieves a transformation from static "first-come, first-served" queuing to dynamic scheduling based on "intelligent priority." The system can automatically adjust the queuing order according to the urgency of the task and changes in the real-time location of the vehicle. This ensures the rapid processing of high-priority tasks while effectively reducing vehicle waiting time and factory congestion, significantly improving the overall collaborative efficiency and emergency response capabilities of logistics transportation. Attached Figure Description

[0015] Figure 1 This is a schematic diagram of the structure of the centralized management system for intelligent vehicles used in an embodiment of the present invention. Detailed Implementation

[0016] To enable those skilled in the art to better understand the technical solutions of this invention, the technical solutions of this invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of this invention.

[0017] Example 1: A centralized management method for intelligent vehicles, comprising: The system receives transportation tasks and obtains a transportation task list. The transportation task includes cargo information, destination factory area, required arrival time, and batching type. Specifically, receiving transportation tasks refers to receiving standardized electronic waybill data from Enterprise Resource Planning (ERP) systems, Transportation Management Systems (TMS), or manual input terminals in real time via a system interface. The transportation task list is a structured data set. Cargo information includes not only cargo name and category but also detailed cargo weight, volume, physical characteristics (such as whether it is hazardous material, temperature control requirements, etc.), and special loading and unloading requirements. Destination factory area information includes a unique factory area code, detailed geographical coordinates, designated loading and unloading area location, and factory area operation time limits. The required arrival time clearly specifies the latest delivery time or expected time window range. The batching type field defines the batching method required for the transportation task. For example, "one vehicle per batch" means a single vehicle can independently complete the entire transportation volume of the batch, while "sequential batching" means the batch needs to be transported by multiple vehicles in a specific order, and the system needs to coordinate the scheduling of multiple vehicles according to this type. A multi-dimensional profile model is constructed for any vehicle. The multi-dimensional profile model includes the vehicle's static attributes, dynamic performance tags, and historical behavior tags. After any vehicle completes a transportation task, the multi-dimensional profile model of the vehicle is updated. Through this continuous update mechanism, the vehicle's profile model can always keep up with its latest operating status, thereby providing more accurate and reliable data support for subsequent task allocation and optimized scheduling. For any given transportation task, based on the multi-dimensional profile model of the transportation task and currently available vehicles, candidate vehicles that meet preset conditions are selected. A multi-objective optimization function is constructed, with transportation cost, transportation efficiency, time matching degree, route familiarity, and resource utilization rate as its core. This optimization function not only helps the scheduling system make scientific decisions under multi-dimensional conditions, but also flexibly adjusts the weights of each objective according to the characteristics of different tasks, further improving the accuracy and efficiency of task completion. The requirements of the transportation task are matched with the multi-dimensional profile model of the candidate vehicles in a multi-dimensional matching calculation, and a vehicle dispatch plan is output. The transportation task is then assigned to the corresponding driver. The system receives check-in requests from dispatched vehicles in designated areas, verifies the vehicle's financial audit status and task validity, and adds the vehicle to the dynamic entry queue and assigns a queue number to the vehicle after verification. This step prevents unapproved or invalid task vehicles from entering the factory area, avoiding resource waste and potential risks. The system monitors the real-time status of each vehicle in the dynamic entry queue, dynamically adjusts the priority order of vehicles in the queue, and issues a departure command to the vehicle ranked first. Unlike traditional fixed-order queuing, this invention dynamically adjusts vehicle priority by monitoring the running status, position, and progress of vehicles in the queue in real time. This mechanism aims to avoid vehicles waiting in vain for a long time and to achieve flexible sorting based on the urgency of the task, arrival time, and actual processing capacity of the factory. After a vehicle enters the factory, its status is automatically updated, triggering subsequent weighing, loading and unloading, and outgoing processes.

[0018] The specific steps for constructing a multi-dimensional vehicle profile model are as follows: Collect basic static attributes of vehicles and drivers, including vehicle license plate number, vehicle type, approved load capacity, driver identity information, and driver's license type; The process of acquiring historical vehicle operation data and generating dynamic performance tags includes: extracting cargo types and transportation frequencies from historical transportation records, calculating the proportion of transportation times for each cargo type, marking cargo types with proportions exceeding a preset frequency threshold as frequently transported cargo for the vehicle, and generating frequently transported cargo tags; calculating the ratio of travel time to distance in historical tasks to generate average travel speed tags; recording the start and end times of each loading and unloading operation, calculating the operation duration, calculating the average loading and unloading time by cargo type, and generating loading and unloading efficiency tags; comparing the required arrival time and actual arrival time of historical tasks, statistically analyzing the proportion of tasks completed on time, and generating on-time rate scoring tags. Based on preset behavioral rules and historical vehicle operation data, historical behavior tags are generated. The generation process of historical behavior tags includes: calculating credit rating tags using a credit scoring algorithm based on information such as the historical task completion quality, violation records, and customer complaint records of the vehicle and driver; recording and classifying violations in the vehicle's historical operation, including speeding, deviating from the route, and failure to operate according to procedures, and generating corresponding violation tags for each violation; extracting high-frequency path segments and their distribution areas from the driver's historical driving trajectory to generate path familiarity tags. The basic static attributes, dynamic performance tags, and historical behavior tags are integrated to form a multi-dimensional profile model of the vehicle.

[0019] Based on the transportation task and the multi-dimensional profile model of currently available vehicles, candidate vehicles that meet preset conditions are selected for the transportation task. The specific operation is as follows: Extract preliminary screening criteria from the current transportation task. The preliminary screening criteria include: the type of goods to be transported, the total weight of the goods, the address of the destination factory area, and the delivery deadline. Based on the preliminary screening criteria, the static attributes of each vehicle are checked, and vehicles that simultaneously meet all of the following basic conditions are selected to form a candidate vehicle set: a) The vehicle's rated load capacity is greater than or equal to the total weight of the goods in this transportation task. This condition is limited to the current transportation task being a batch of one vehicle per batch. b) The vehicle type must match the physical characteristics of the cargo. For example, liquid cargo may require tank trucks, while mineral materials may require dump trucks. The candidate vehicle set and its corresponding multi-dimensional profile data are output for subsequent multi-objective optimization function calculation and vehicle dispatch decision-making.

[0020] A multi-objective optimization function is constructed, with total transportation cost, overall efficiency, and vehicle queuing time as the core. The demand for the transportation task is matched and calculated with the real-time status, real-time location, and multi-dimensional profile data of the candidate vehicles in a multi-dimensional manner, and a vehicle dispatch plan is output. The specific operation is as follows: A multi-objective optimization function is established, comprising the following five core evaluation dimensions. For any given transportation task, multi-dimensional calculations are performed on several candidate vehicles corresponding to that task. The specific calculation methods for each dimension are as follows: Transportation cost calculation: The shortest path distance from the pickup plant to the destination plant is calculated using a path planning algorithm. Combined with the historical average energy consumption data of this vehicle model under the same road conditions and the current energy price, the estimated transportation cost is calculated according to the formula: transportation cost = path distance × energy consumption per unit distance × energy price. Transportation efficiency calculation: The total task time is broken down into three parts: pickup travel time, delivery travel time, and loading / unloading time. The estimated pickup travel time is calculated based on the vehicle's real-time location, distance to the pickup area, and average vehicle speed. The delivery travel time is calculated based on the shortest path distance and average vehicle speed. The average loading / unloading time for the current cargo type is directly used as the estimated loading / unloading time based on the loading / unloading efficiency tags in the vehicle's multi-dimensional profile. The total estimated task time is the sum of the pickup travel time, delivery travel time, and loading / unloading time. Time matching dimension calculation: Based on the estimated total task time, the estimated arrival time is obtained, and the absolute time difference between the estimated arrival time and the delivery deadline is calculated; Path familiarity dimension calculation: The trajectory similarity algorithm based on the proportion of common road segments is used to calculate the overlap between the task-planned path and the driver's historical driving path, and the ratio of the length of the longest common subsequence of the two paths to the total path length is calculated as the path familiarity. Resource utilization rate calculation: Calculate the ratio of the weight of the goods to the vehicle's rated load capacity to obtain the load utilization rate value; For each evaluation dimension, the maximum and minimum values ​​of the current candidate vehicle set in that dimension are extracted; the original values ​​of each evaluation dimension are converted into standardized scores in the range of 0 to 1 using a linear scaling transformation method. The system dynamically configures the weight coefficients for each dimension based on the characteristics of the transportation task. It adjusts weights according to the task's urgency, transportation cost control requirements, cargo characteristics, and historical transportation data. For example, for urgent tasks, the system prioritizes time-matching and transportation efficiency to ensure timely completion. For tasks with limited budgets, the system focuses more on cost control, increasing the weight of transportation costs. Furthermore, for certain special goods, the system prioritizes time and efficiency, while for regular goods, it emphasizes resource utilization and cost optimization. In this way, the system can flexibly adjust weights and optimize dispatching plans based on the characteristics of different tasks, thereby improving transportation efficiency and resource utilization, and ensuring efficient task execution. The weighted summation method is used to calculate the comprehensive score of each candidate vehicle. All candidate vehicles are sorted in descending order according to the comprehensive score. If the current transportation task is a batch type of one vehicle per batch, the vehicle with the highest comprehensive score is selected as the optimal dispatch plan. If the current transportation task is a batch type of sequential batching, vehicles are selected in descending order of comprehensive score until the sum of the approved load capacity of the selected vehicles is not less than the total weight of the goods. Output a vehicle dispatch plan, which includes recommended vehicles and corresponding multi-dimensional calculation results.

[0021] Verify the vehicle's financial audit status and task validity. Once the verification is successful, add the vehicle to the dynamic entry queue and assign a queue number to it. The specific steps are as follows: When a vehicle enters the preset electronic fence geographical area, the driver manually initiates a check-in request. This request must include at least the vehicle identification information, waybill number, and real-time location coordinates. Upon receiving the check-in request, the following verification process is executed: Initiate a real-time query to the financial management system to obtain the current financial review status corresponding to the waybill number; Verify that the binding relationship between the waybill and the vehicle is valid, and confirm that the waybill status is "dispatched and awaiting execution"; If the financial audit status is "audited" and the waybill is valid, the verification is considered successful; if the financial status is "not audited" or "audit rejected", a check-in failure response containing the specific reason for rejection will be immediately returned to the vehicle terminal. For vehicles that pass the verification, the task priority is calculated based on the following objective factors: Obtain the current time, the required arrival time for the task, and the estimated arrival time based on the vehicle's performance of the current transportation task; Calculate the difference between the estimated arrival time and the required arrival time, then proportionally calculate the time urgency value by dividing this difference by the total available time (required arrival time minus the current time), which is used as the task priority score. Vehicle information is added to a dynamic entry queue, which is implemented using a data structure based on task priority. The vehicle's ranking in the queue is determined by its task priority score; the higher the score, the higher the ranking.

[0022] The order of the dynamic entry queue is dynamically updated based on real-time data. The specific operation is as follows: Real-time location coordinates and speed of each vehicle in the dynamic entry queue are collected, and loading and unloading efficiency labels are extracted from the multi-dimensional vehicle profile model. Establish a dynamic ranking adjustment mechanism, making comprehensive decisions based on the following three core factors: Task priority factor: Directly use task priority scoring; Real-time distance factor: Based on the vehicle's real-time location and speed, calculate the estimated time it will take for the vehicle to reach the pickup area; Loading and unloading efficiency factors: Obtain the average loading and unloading time of vehicles for the current type of goods; Set a fixed time interval, and at regular intervals, perform the following sorting adjustment operation: Based on the maximum and minimum values ​​of each factor, a linear transformation method is used to convert the original values ​​of each vehicle on each factor into a standardized score between 0 and 1. The weighted calculation is performed using fixed weight coefficients, with task priority having the highest weight, real-time distance having the second highest weight, and loading / unloading efficiency having the lowest weight; finally, a dynamic ranking score is obtained for each vehicle; and all vehicles in the queue are re-ranked based on the dynamic ranking score.

[0023] Example 2: A centralized management system for intelligent vehicles, such as... Figure 1 As shown, it includes: The task receiving module is used to receive transportation tasks and obtain a transportation task list. The transportation tasks include cargo information, destination factory area, required arrival time and batch type. The profile building module is used to build a multi-dimensional profile model for any vehicle. The multi-dimensional profile model includes the vehicle's static attributes, dynamic performance tags, and historical behavior tags. After any vehicle completes a transportation task, the multi-dimensional profile model of that vehicle is updated. The vehicle screening module includes a preliminary screening unit and an optimization calculation unit. The preliminary screening unit is used to screen candidate vehicles that meet preset conditions for any given transportation task, based on the transportation task and the multi-dimensional profile model of currently available vehicles. The optimization calculation unit is used to construct a multi-objective optimization function with transportation cost, transportation efficiency, time matching degree, route familiarity, and resource utilization rate as the core, to perform multi-dimensional matching calculations between the demand of the transportation task and the multi-dimensional profile model of the candidate vehicles, output a dispatch plan, and assign the transportation task to the corresponding driver. The check-in verification module is used to receive check-in requests initiated by dispatched vehicles in designated areas, verify the vehicle's financial audit status and task validity, and add the vehicle to the dynamic entry queue after the verification is passed, and assign a queue number to the vehicle. The queue management module is used to monitor the real-time status of each vehicle in the dynamic entry queue, dynamically adjust the priority order of vehicles in the queue, and issue a departure command to the vehicle with the highest priority. After a vehicle enters the site, its status is automatically updated, and subsequent weighing, loading and unloading and factory exit processes are triggered.

[0024] It should be understood that those skilled in the art can make improvements or modifications based on the above description, and all such improvements and modifications should fall within the protection scope of the appended claims. Parts not described in detail in this specification are prior art known to those skilled in the art.

Claims

1. A centralized management method for intelligent vehicles, characterized in that, include: Receive transportation tasks and obtain a transportation task list; Build a multi-dimensional profile model for any vehicle; After any vehicle completes a transportation task, the multi-dimensional profile model of that vehicle is updated. For any given transportation task, based on the transportation task and the multi-dimensional profile model of the currently available vehicles, candidate vehicles that meet preset conditions are selected for the transportation task; a multi-objective optimization function is constructed to perform multi-dimensional matching calculations between the requirements of the transportation task and the multi-dimensional profile model of the candidate vehicles, output a vehicle dispatching plan, and assign the transportation task to the corresponding driver. Receive check-in requests initiated by dispatched vehicles in designated areas, verify the vehicle's financial audit status and task validity, and after verification, add the vehicle to the dynamic entry queue and assign a queue number to the vehicle. The system monitors the real-time status of each vehicle in the dynamic entry queue, dynamically adjusts the priority order of vehicles in the queue, and issues a departure command to the vehicle ranked first. After a vehicle enters the site, its status is automatically updated, and subsequent weighing, loading and unloading, and factory exit processes are triggered.

2. The centralized management method for intelligent vehicles according to claim 1, characterized in that, The specific steps for constructing a multi-dimensional vehicle profile model are as follows: The multi-dimensional profile model includes the vehicle's basic static attributes, dynamic performance tags, and historical behavior tags; Collect basic static attributes of vehicles and drivers, including vehicle license plate number, vehicle type, approved load capacity, driver identity information, and driver's license type; The process of acquiring historical vehicle operation data and generating dynamic performance tags includes: extracting cargo types and transportation frequencies from historical transportation records, calculating the proportion of transportation times for each cargo type, marking cargo types with proportions exceeding a preset frequency threshold as frequently transported cargo for the vehicle, and generating frequently transported cargo tags; calculating the ratio of travel time to distance in historical tasks to generate average travel speed tags; recording the start and end times of each loading and unloading operation, calculating the operation duration, calculating the average loading and unloading time by cargo type, and generating loading and unloading efficiency tags; comparing the required arrival time and actual arrival time of historical tasks, statistically analyzing the proportion of tasks completed on time, and generating on-time rate scoring tags. Based on preset behavioral rules and historical vehicle operation data, historical behavior tags are generated. The generation process of historical behavior tags includes: calculating credit rating tags using a credit scoring algorithm based on information such as the historical task completion quality, violation records, and customer complaint records of the vehicle and driver; recording and classifying violations in the vehicle's historical operation, including speeding, deviating from the route, and failure to operate according to procedures, and generating corresponding violation tags for each violation; extracting high-frequency path segments and their distribution areas from the driver's historical driving trajectory to generate path familiarity tags. The basic static attributes, dynamic performance tags, and historical behavior tags are integrated to form a multi-dimensional profile model of the vehicle.

3. The centralized management method for intelligent vehicles according to claim 2, characterized in that, Based on the transportation task and the multi-dimensional profile model of currently available vehicles, candidate vehicles that meet preset conditions are selected for the transportation task. The specific operation is as follows: The transportation task includes cargo information, destination factory area, required arrival time, and batch type; Extract preliminary screening criteria from the current transportation task. The preliminary screening criteria include: the type of goods to be transported, the total weight of the goods, the address of the destination factory area, and the delivery deadline. Based on the preliminary screening criteria, the static attributes of each vehicle are checked, and vehicles that simultaneously meet all of the following basic conditions are selected to form a candidate vehicle set: a) The vehicle's rated load capacity is greater than or equal to the total weight of the goods in this transportation task. This condition is limited to the current transportation task being a batch of one vehicle per batch. b) The physical characteristics of the vehicle type must match the requirements of the cargo type; The candidate vehicle set and its corresponding multi-dimensional profile data are output for subsequent multi-objective optimization function calculation and vehicle dispatch decision-making.

4. The centralized management method for intelligent vehicles according to claim 3, characterized in that, A multi-objective optimization function is constructed to perform multi-dimensional matching calculations between the transportation task requirements and the real-time status, real-time location, and multi-dimensional profile data of candidate vehicles, and outputs a vehicle dispatch plan. The specific operation is as follows: A multi-objective optimization function is established, comprising the following five core evaluation dimensions. For any given transportation task, multi-dimensional calculations are performed on several candidate vehicles corresponding to that task. The specific calculation methods for each dimension are as follows: Transportation cost calculation: The shortest path distance from the pickup plant to the destination plant is calculated using a route planning algorithm. Combined with the historical average energy consumption data per 100 kilometers of this vehicle under the same road conditions and the current energy price, the estimated transportation cost is calculated. Transportation efficiency calculation: The total task time is broken down into three parts: pickup travel time, delivery travel time, and loading and unloading time; the estimated pickup travel time is calculated based on the distance from the vehicle's real-time location to the pickup area and the vehicle's average speed. Calculate the vehicle's delivery travel time based on the shortest path distance and the vehicle's average speed. Based on the loading and unloading efficiency labels in the multi-dimensional vehicle profile, the average loading and unloading time corresponding to the current cargo type is directly used as the estimated loading and unloading time. The estimated total task time is the sum of the pickup travel time, delivery travel time, and loading / unloading time. Time matching dimension calculation: Based on the estimated total task time, the estimated arrival time is obtained, and the absolute time difference between the estimated arrival time and the delivery deadline is calculated; Path familiarity dimension calculation: The trajectory similarity algorithm based on the proportion of common road segments is used to calculate the overlap between the task-planned path and the driver's historical driving path, and the ratio of the length of the longest common subsequence of the two paths to the total path length is calculated as the path familiarity. Resource utilization rate calculation: Calculate the ratio of the weight of the goods to the vehicle's rated load capacity to obtain the load utilization rate value; For each evaluation dimension, extract the maximum and minimum values ​​of the current candidate vehicle set in that dimension; The original values ​​of each evaluation dimension were converted into standardized scores in the range of 0 to 1 using a linear scaling transformation method. The weight coefficients of each dimension are dynamically configured according to the characteristics of the transportation task. The comprehensive score of each candidate vehicle is calculated by weighted summation. All candidate vehicles are sorted in descending order according to the comprehensive score. If the batch type of the current transportation task is one vehicle per batch, the vehicle with the highest comprehensive score is selected as the optimal dispatch plan. If the batch type of the current transportation task is sequential batching, vehicles are selected in descending order of comprehensive score until the sum of the approved load capacity of the selected vehicles is not less than the total weight of the goods. Output a vehicle dispatch plan, which includes recommended vehicles and corresponding multi-dimensional calculation results.

5. A centralized management method for intelligent vehicles according to claim 4, characterized in that, Verify the vehicle's financial audit status and task validity. Once the verification is successful, add the vehicle to the dynamic entry queue and assign a queue number to it. The specific steps are as follows: When a vehicle enters the preset electronic fence geographical area, the driver manually initiates a check-in request. This request must include at least the vehicle identification information, waybill number, and real-time location coordinates. Upon receiving the check-in request, the following verification process is executed: Initiate a real-time query to the financial management system to obtain the current financial review status corresponding to the waybill number; verify whether the binding relationship between the waybill and the vehicle is valid, and confirm that the waybill status is dispatched and awaiting execution; If the financial audit status is "audited" and the waybill is valid, the verification is considered successful; if the financial status is "not audited" or "audit rejected", a check-in failure response containing the specific reason for rejection will be immediately returned to the vehicle terminal. For vehicles that pass the verification, the task priority is calculated based on the following objective factors: Obtain the current time, the required arrival time for the task, and the estimated arrival time based on the vehicle's performance of the current transportation task; Calculate the difference between the estimated arrival time and the required arrival time, then proportionally calculate the time urgency value by dividing the difference by the total available time, which is used as the task priority score. Vehicle information is added to a dynamic entry queue, which is implemented using a data structure based on task priority. The vehicle's ranking in the queue is determined by its task priority score; the higher the score, the higher the ranking.

6. A centralized management method for intelligent vehicles according to claim 5, characterized in that, The order of the dynamic entry queue is dynamically updated based on real-time data. The specific operation is as follows: Real-time location coordinates and speed of each vehicle in the dynamic entry queue are collected, and loading and unloading efficiency labels are extracted from the multi-dimensional vehicle profile model. Establish a dynamic ranking adjustment mechanism, making comprehensive decisions based on the following three core factors: Task priority factor: Directly use task priority scoring; Real-time distance factor: Based on the vehicle's real-time location and speed, calculate the estimated time it will take for the vehicle to reach the pickup area; Loading and unloading efficiency factors: Obtain the average loading and unloading time of vehicles for the current type of goods; Set a fixed time interval, and at regular intervals, perform the following sorting adjustment operation: Based on the maximum and minimum values ​​of each factor, a linear transformation method is used to convert the original values ​​of each vehicle on each factor into a standardized score between 0 and 1. The weighted calculation is performed using fixed weight coefficients, with task priority having the highest weight, real-time distance having the second highest weight, and loading / unloading efficiency having the lowest weight; finally, a dynamic ranking score is obtained for each vehicle; and all vehicles in the queue are re-ranked based on the dynamic ranking score.

7. A centralized management system for intelligent vehicles, characterized in that, The system is applied to a centralized management method for intelligent vehicles as described in any one of claims 1-6, comprising: The task receiving module is used to receive transportation tasks and obtain a transportation task list. The profile building module is used to build a multi-dimensional profile model for any vehicle; after any vehicle completes a transportation task, the multi-dimensional profile model of that vehicle is updated. The vehicle screening module includes a preliminary screening unit and an optimization calculation unit. The preliminary screening unit is used to screen candidate vehicles that meet preset conditions for any given transportation task, based on the transportation task and the multi-dimensional profile model of currently available vehicles. The optimization calculation unit is used to construct a multi-objective optimization function with transportation cost, transportation efficiency, time matching degree, route familiarity, and resource utilization rate as the core, and to perform multi-dimensional matching calculations between the demand of the transportation task and the multi-dimensional profile model of the candidate vehicles, output a dispatch plan, and assign the transportation task to the corresponding driver. The check-in verification module is used to receive check-in requests initiated by dispatched vehicles in designated areas, verify the vehicle's financial audit status and task validity, and add the vehicle to the dynamic entry queue after the verification is passed, and assign a queue number to the vehicle. The queue management module is used to monitor the real-time status of each vehicle in the dynamic entry queue, dynamically adjust the priority order of vehicles in the queue, and issue a departure command to the vehicle with the highest priority. After a vehicle enters the site, its status is automatically updated, and subsequent weighing, loading and unloading and factory exit processes are triggered.

Images