CN116843135A - System, architecture and method for cluster management and scheduling of short-range vehicles - Google Patents

Abstract

The invention discloses a system, a framework and a method for managing and dispatching short vehicles, which relate to the technical field of vehicle cluster management and dispatching and comprise a data acquisition layer, a data processing layer, a data execution layer, a user interface layer and a data feedback layer.

Description

A system, architecture and method for short-distance vehicle cluster management and dispatching

Technical field

The present invention belongs to the technical field related to vehicle cluster management and dispatching, and in particular relates to a system, architecture and method for short-distance vehicle cluster management and dispatching.

Background technique

The agricultural product wholesale market is an important node for the circulation of agricultural products and an important infrastructure for ensuring and improving people's livelihood. At present, my country's agricultural product wholesale markets have problems such as high circulation costs, low circulation efficiency, chaotic circulation order, and hidden circulation safety hazards. There is an urgent need to strengthen market system construction and standardized management.

The main components of existing technical solutions:

(1) Vehicle positioning technology: Use global positioning system (GPS) and wireless communication technology to track and monitor the location and status of short-distance shuttle vehicles in real time. This technology can provide precise location information of vehicles and provide basic data for cluster management and scheduling.

(2) Data processing and analysis technology: By integrating and analyzing vehicle positioning data with other related data, such as traffic conditions, cargo information, vehicle loads, etc., the system can generate real-time vehicle dispatching plans. These technologies include data mining, machine learning, and optimization algorithms.

(3) Communication technology: Wireless communication technology is used to communicate in real time between short-distance vehicles and the system to receive dispatching instructions, transmit data and report vehicle status. This technology ensures efficient collaboration and information exchange between vehicles and systems.

(4) Cluster management algorithm: Using distributed algorithms and collaborative optimization methods, the system can automatically manage the entire short-distance vehicle cluster. These algorithms can develop optimal dispatch plans based on factors such as cargo demand, vehicle availability and routes, improving overall efficiency and reducing transportation costs.

(5) Vehicle identification device: used to identify short-distance shuttle vehicles within the market, such as license plate recognition, QR code recognition, RFID recognition, etc.;

(6) Vehicle positioning device: used to obtain real-time location information of short-distance shuttle vehicles, such as GPS positioning, base station positioning, Beidou positioning, etc.;

(7) Vehicle dispatching device: used to intelligently allocate and dispatch short-distance vehicles based on information such as cargo demand, vehicle location, vehicle status, etc., such as central console, mobile terminal, voice prompts, etc.;

(8) Vehicle monitoring device: used to monitor the operation of short-term barge vehicles, such as driving speed, driving route, driving time, etc., as well as the status of cargo, such as temperature, humidity, weight, etc.

Existing defects:

(1) Data accuracy and real-time performance: The short-distance vehicle cluster management system has high requirements for data accuracy and real-time performance. If positioning data, traffic conditions and other information are inaccurate or not updated in a timely manner, it may lead to inaccuracy and reduced efficiency of the dispatch plan.

(2) System complexity and reliability: The system needs to process a large amount of data and perform complex algorithm calculations and decisions. This can lead to increased system complexity and higher demands on computing and communication resources. At the same time, system reliability is also a challenge because factors such as signal interference, network failure, equipment damage, etc. may lead to scheduling delays or errors.

(3) Personnel training and acceptance: The introduction of such a new system and device requires training of relevant personnel to ensure that they can correctly use and understand the functions and operation methods of the system. Additionally, some personnel may have varying degrees of acceptance of automated systems, requiring appropriate training and communication to mitigate resistance.

(4) Privacy protection and network security: The short-distance vehicle cluster management system needs to collect and process sensitive data such as vehicle positioning data and cargo information. When designing your system, you need to ensure that this data is stored and transmitted securely, and that appropriate privacy protection measures are in place to prevent unauthorized access and data leakage. The system involves real-time communication between the vehicle and back-end systems and therefore needs to guard against the risks of cyberattacks and data tampering. Appropriate security measures, such as the use of encryption, authentication and access controls, must be taken to ensure the confidentiality, integrity and reliability of communications.

(5) Data backup and recovery: For important data in the system, such as vehicle dispatch records, cargo information, etc., regular data backup and recovery strategies must be implemented. This prevents data loss or corruption and enables rapid data recovery in the event of a system failure or disaster.

Contents of the invention

The technical problem to be solved by the present invention is to provide a system, architecture and method for short-distance vehicle cluster management and dispatching in view of the deficiencies of the background technology, which utilizes information technology to improve system operation stability, simplify system usage processes, and enhance system adaptability. It can uniformly dispatch and manage short-distance vehicles within the market to achieve rapid transshipment and distribution of goods, thereby improving the circulation efficiency and service level of the agricultural product wholesale market.

The present invention adopts the following technical solutions to solve the above technical problems:

A short-distance vehicle cluster management and dispatching system that constructs the following layers to implement the agricultural batch market

Intelligent management of short-term barge dispatch is as follows:

(1) The data collection layer is used to use vehicle positioning technology and sensor equipment to collect cargo demand, vehicle location, and vehicle status data within the market, and send it to the data processing layer through wireless transmission;

(2) The data processing layer is responsible for receiving the data sent by the data acquisition layer, and performing data integration, data cleaning, data analysis and processing, and using data mining and machine learning technology to extract useful information, including vehicle scheduling optimization and cargo batching. Management, thereby obtaining an optimized short-distance vehicle dispatching plan, providing decision support, such as transportation route optimization, vehicle dispatching plan, and sending it to the data execution layer through wireless transmission;

(3) Data execution layer: responsible for executing the plan sent by the data processing layer, such as the actual operation of short-distance vehicle dispatching and cargo batching, and reporting it to relevant personnel such as short-term vehicle drivers and warehouse managers through voice or graphical interfaces. Send instructions to guide them to carry out cargo transshipment and distribution according to the plan to ensure the accurate execution of the dispatch plan;

(4) User interface layer is used to provide a user-friendly interface that allows users to easily view and manage vehicle dispatching, cargo information, and market demand-related information. Users can manage orders, query transportation status, and adjust dispatching plan operations through the interface. ;

(5) The data feedback layer is responsible for collecting feedback information from short-distance vehicle drivers in executing the scheduling plan and sending it to the data processing layer through wireless transmission as a basis for the data processing layer to optimize the scheduling plan.

As a further preferred solution of the short-distance vehicle cluster management and dispatching system of the present invention, the data processing layer adopts

Reinforcement learning algorithms, optimization algorithms, real-time monitoring and prediction algorithms receive the data sent by the data collection layer, and perform data integration, data cleaning, data analysis and processing, and use data mining and machine learning technologies to extract useful information.

As a further preferred solution of the short-distance vehicle cluster management and dispatching system of the present invention, the reinforcement learning algorithm,

details as follows:

According to the reward signal provided by the data feedback layer, it continuously updates its own policy network parameters to learn the optimal or nearly optimal short-distance vehicle dispatching plan; specifically, the reinforcement learning algorithm includes the following elements:

①State space: The state space includes the characteristics of the current problem, the current solution, and the solution process; specifically, the state space consists of the following parts:

Goods demand vector: indicates the quantity of goods required for each goods demand point including the warehouse;

Vehicle position vector: represents the current location of each short-distance vehicle;

Vehicle status vector: represents the current quantity of goods carried by each short-distance vehicle;

Route length vector: represents the current route length of each shuttle vehicle;

Route sequence vector: represents the sequence of cargo demand points currently visited by each short-distance vehicle;

Remaining demand vector: Indicates that each goods demand point includes the remaining unsatisfied quantity of goods in the warehouse;

Remaining capacity vector: represents the remaining amount of cargo that each short-distance vehicle can carry;

② Action space: The action space includes all possible short-distance vehicle dispatching plans; specifically, the action space consists of the following parts:

Vehicle selection action: select a short-distance shuttle vehicle that is currently available, has not completed the task, and is not overloaded for dispatch;

Demand point selection action: Select a goods demand point that has not been visited currently and has remaining demand, that is, non-zero, as the next visit target;

Cargo transfer action: After arriving at the target demand point, determine how much goods to unload or load from that point based on the demand situation and vehicle status at that point;

③Reward function: The reward function is used to evaluate the contribution of each action to the overall objective function, that is, minimizing the total route length; specifically, the reward function consists of the following parts:

Fixed reward: means that for every visit to a goods demand point other than the warehouse, a fixed value reward will be obtained;

Penalty factor: means that every time the route length increases by one unit, a proportional reward will be deducted;

Termination reward: means that when all goods demand points are visited and satisfied, an additional value reward will be obtained;

④ Policy network: The policy network is used to output the optimal or approximately optimal action probability distribution based on the current state; specifically, the policy network consists of the following parts:

Input layer: The input layer receives each vector in the state space as input and splices it into a one-dimensional vector;

Hidden layer: The hidden layer consists of several fully connected layers or convolutional layers, and uses activation functions to add nonlinear features;

Output layer: The output layer consists of three sub-output layers, which respectively correspond to three parts of the action space, namely vehicle selection action, demand point selection action and cargo transfer action; specifically:

A. Vehicle selection sub-output layer: This sub-output layer consists of a fully connected layer and a softmax layer, and outputs a vector of length K, indicating the probability of selecting each short-distance vehicle;

B. Demand point selection sub-output layer: This sub-output layer consists of a fully connected layer and a softmax layer. It outputs a vector of length N+1, indicating the probability of selecting each goods demand point, including the warehouse, as the next access target. ;

C. Cargo transfer sub-output layer: This sub-output layer consists of a fully connected layer and a softmax layer. It outputs a vector with a length of C+1, which represents the unloading or loading of 0 to C units of goods from the target demand point. probability;

Based on the output of these three sub-output layers, a complete action is obtained, that is, which vehicle is selected, which demand point is visited, and how many goods are transferred; the reward value of the action is calculated according to the reward function, and the parameters of the policy network are updated according to the policy gradient algorithm.

As a further preferred solution of the short-term barge vehicle cluster management and dispatching system of the present invention, the vehicle dispatching algorithm: based on vehicle location, market demand and cargo information data, use an optimization algorithm to determine the optimal vehicle dispatching plan, including route planning, vehicle distribute;

Cargo batching algorithm: Based on market demand and vehicle availability data, an optimization algorithm is used to determine a reasonable batching plan for goods to improve distribution efficiency and meet market demand;

Real-time monitoring and prediction algorithm: By real-time monitoring of vehicle location, traffic conditions and other data, and combined with historical data for predictive analysis, potential problems can be discovered in a timely manner and corresponding adjustments can be made, such as route re-planning and vehicle replacement.

An architecture based on short-distance vehicle cluster management and dispatching system, including

Front-end and back-end part architecture:

The front-end part includes the user interface, which can be provided to administrators, shuttle vehicle drivers and other related personnel through web applications or mobile applications; the user interface provides user registration, task viewing, data query and statistical functions, allowing users to interact with the system interact;

The back-end part includes main modules such as cluster management, order grabbing function and order dispatching function; the cluster management module is responsible for managing the registration, status monitoring and task allocation optimization of the short-distance vehicle cluster; the order grabbing function module handles the driver's order grabbing request, filtering and matching and task allocation; the dispatch function module is responsible for task release, vehicle screening and matching, and task scheduling.

Cluster management server architecture:

The cluster management server is the core component of the system and is responsible for the cluster management, scheduling algorithm and data analysis functions of the entire system; its architecture adopts a distributed architecture and includes the following modules:

The cluster management module is used to monitor and manage the status, location and operation of short-distance vehicle clusters, and provides vehicle management and task allocation functions.

The scheduling algorithm module implements batch scheduling algorithms, generates optimal scheduling plans based on real-time vehicle and cargo information, optimizes vehicle utilization, and reduces transportation costs;

The data analysis module performs real-time analysis of vehicle and cargo data stored in the database and provides decision support, such as scheduling optimization and route planning;

The communication module is responsible for exchanging data and issuing instructions with the short-distance vehicle terminal equipment and the dispatch center terminal equipment;

Short-distance vehicle terminal equipment architecture:

Vehicle-mounted terminal equipment is a device installed on each short-distance shuttle vehicle and is used for functions such as data collection, communication and execution of dispatch instructions; its architecture includes the following components:

Sensor module: including GPS positioning sensor, speed sensor, load sensor, etc., used to collect vehicle location, status and operating data;

Communication module: used to use wireless network technology and cluster management server for real-time data transmission and instruction issuance;

The control module is used to control the loading, unloading, transportation and other operations of the vehicle according to the received dispatch instructions;

The data storage module is used to temporarily store the collected data to ensure the real-time and reliability of the data;

Dispatch center terminal equipment architecture:

The dispatch center terminal device is used by dispatchers to interact with the cluster management server to implement the issuance of dispatch instructions and data query functions; its architecture includes the following components:

User interface, used to provide an intuitive and friendly user interface through which the dispatcher can interact with the cluster management server;

The data exchange module is responsible for data exchange with the cluster management server, including the issuance of scheduling instructions and the return of data query results;

Communication module, used for real-time communication with the cluster management server using network connections;

Third-party logistics platform:

The platform is an external partner of the system and is responsible for providing business needs, including cargo name, pickup address, delivery address, pickup time, and required arrival time information; the platform is a cargo owner enterprise management system, a third-party logistics management system, or an open business Modular form; the platform is connected to the dispatch platform through an open interface to implement the functions of sending and receiving business requests, querying and updating order status, and acquiring and analyzing transportation data.

As a further preferred solution of the architecture of the short-distance vehicle cluster management and dispatching system of the present invention, it also includes a database, specifically as follows:

Vehicle information database:

The database is used to store and manage information related to short-distance shuttle vehicles, including but not limited to the following:

Vehicle number: a number that uniquely identifies each short-distance vehicle;

Location information: Record the real-time location coordinates of the vehicle for real-time monitoring and dispatching;

Status information: including vehicle operating status, load status, etc., used to determine vehicle availability and make scheduling decisions;

Cargo information database:

This database is used to store and manage cargo-related information, including but not limited to the following:

Cargo number: a number that uniquely identifies each cargo;

Cargo type: record the type of cargo for scheduling and matching with appropriate vehicles;

Quantity information: record the quantity of goods for scheduling and distribution planning;

Scheduling rule database:

This database is used to store and manage scheduling rules, including but not limited to the following:

Vehicle scheduling priority: Set the scheduling priority of vehicles according to different conditions to ensure optimal scheduling decisions;

Batch scheduling algorithm: stores the relevant parameters and rules of the batch scheduling algorithm and is used to generate the optimal batch scheduling plan;

Scheduling strategy: record dispatching rules, such as vehicle loading rules and delivery route planning, to ensure the rationality and efficiency of dispatching;

Historical data database:

This database is used to store and manage historical data for subsequent data analysis and decision support, including but not limited to the following:

Vehicle transportation records: record the transportation history of each vehicle, including starting location, destination, transportation time, etc., to analyze the transportation efficiency and quality of the vehicle;

Goods delivery time record: record the delivery time of each goods in order to predict and optimize the delivery time;

These databases can be stored and managed using relational databases or distributed databases to meet the system's requirements for real-time, reliability and scalability of data;

The database design of the short-term barge vehicle cluster management and batch dispatching system includes a vehicle information database, a cargo information database, a dispatching rule database and a historical data database, which are used to store and manage relevant real-time and historical data to support system dispatching decisions and analysis. Function.

A method based on short-distance vehicle cluster management and dispatching system, specifically including the following steps:

Cluster management algorithm, specifically as follows:

Vehicle registration: Short-distance barge vehicles are registered in the system and relevant information is provided, such as vehicle number and load capacity;

Vehicle status monitoring: The system monitors the location, availability and operating status of the vehicle in real time; obtains the vehicle's real-time data through vehicle sensors and GPS equipment, and feeds it back to the cluster management software;

Task allocation and optimization: Based on the vehicle's location, availability and real-time data, use task allocation and optimization algorithms to allocate tasks to appropriate vehicles; the algorithm takes into account the vehicle's load capacity, driving distance, and time window factors to achieve efficient task allocation ;

Task release, details are as follows:

The system dispatches tasks to the most suitable vehicle driver based on vehicle location, task type, real-time traffic and other information;

During the order dispatch process, the system needs to consider factors such as the vehicle's current task status, task priority, and the vehicle's working time to ensure that the task can be completed in a timely manner;

After dispatch, the system updates the task status and related information assigned to the vehicle to facilitate subsequent scheduling and monitoring;

The order grabbing function algorithm is as follows:

Receiving order grabbing requests: When there is a task that needs to be executed, the system will publish the task information, including task type, starting location, and destination; short-distance vehicle drivers can receive task requests through the order grabbing function software;

Driver screening and matching: The system screens and matches order grabbing requests based on task requirements and the driver's ability and availability; it evaluates the driver's location, load capacity, available time and other factors to find the most suitable driver;

Task allocation: The system allocates tasks to the most appropriate driver to ensure that the task can be executed in a timely manner; the task allocation algorithm can take into account factors such as the driver's distance, traffic conditions, task priority, etc. to achieve the best task allocation effect;

The dispatch function algorithm is as follows:

Task release: The system releases task information based on task requirements and priority, including task type, starting location, and destination;

Vehicle screening and matching: The system screens and matches vehicles for tasks based on the vehicle's location, availability and real-time data; the algorithm takes into account the vehicle's load capacity, driving distance, and time window factors to find the most suitable vehicle;

Task scheduling: The system schedules tasks to the most appropriate vehicles and generates corresponding scheduling plans; the scheduling algorithm can take into account the vehicle's route, time window, and priority factors to achieve optimal task scheduling effects;

Scheduling optimization is as follows:

The system collects and analyzes vehicle historical data, including operating trajectory, task completion time, and waiting time information;

Based on historical data and real-time information, the system can optimize dispatching strategies and improve vehicle dispatching efficiency and task completion rate;

Scheduling optimization can include route planning, task batch scheduling, and task priority adjustment technical means to minimize overall cost and time;

System monitoring and updates, as follows:

The system monitors vehicle location, task progress, driver status and other information in real time to facilitate real-time scheduling and task tracking;

The system dynamically updates task status and allocation based on the actual vehicle situation and task progress;

The system can also provide real-time data statistics and reports to facilitate decision-making and optimization by managers;

The evaluation algorithm is as follows:

Receive shuttle bus operation data, such as speed, fuel consumption, mileage, and load information;

Receive customer satisfaction feedback, such as ratings and comments;

The completion status of receiving orders, such as whether they arrived on time and whether the information is intact;

Based on the operation data of the shuttle bus, customer satisfaction feedback, order completion and other indicators, evaluate the service quality of the shuttle bus and give corresponding reward and punishment measures, such as extra points, subtracted points, bonuses, and fines;

Send evaluation results and reward and punishment measures to the shuttle bus through the wireless communication module;

The steps of the optimization algorithm are as follows:

Receive indicators such as service quality, transportation efficiency, and operating costs of shuttle buses;

According to the changing trends and target values of indicators, optimize and update the batch scheduling algorithm and order grabbing algorithm, such as adjusting parameters, adding constraints, and improving strategies;

Deploy the optimized batch scheduling algorithm and order grabbing algorithm to the central server and mobile terminals. Compared with the existing technology, the present invention adopts the above technical solution and has the following technical effects:

The invention improves efficiency: the system realizes intelligent management and dispatching of vehicle clusters through the functions of grabbing orders and dispatching orders; vehicles can actively grab orders or be dispatched according to actual needs, avoiding unnecessary waiting in traditional dispatching methods. Waste of time and resources, improved transportation efficiency;

2. The present invention optimizes resource utilization: the system uses intelligent dispatching algorithms to allocate goods in batches to the most suitable vehicles for transportation, avoiding no-load or low-load operation of vehicles and maximizing resource utilization efficiency. This can reduce transportation costs and improve operational efficiency;

The invention improves service quality: the dispatch function in the system can quickly and accurately allocate tasks according to the location, status and cargo requirements of the vehicle; this helps to improve the response speed and accuracy of transportation services, meet customer needs, and improve customer satisfaction;

4. Real-time monitoring and management of the present invention: The system has a real-time monitoring function, which can conduct real-time monitoring and management of the vehicle's location, transportation progress, driving trajectory, etc. This helps enterprises to timely track and adjust vehicle operations and improve transportation. safety and controllability;

5. Data analysis and decision support of the present invention: the system collects and analyzes vehicle operation data and provides data analysis reports on transportation efficiency, cost, customer needs, etc.; this provides data support for enterprises to help them make reasonable decisions and optimize Operational strategies to further improve transportation efficiency and profitability;

6. User-friendliness of the present invention: Our system design takes into account the needs and usage habits of users, and provides an intuitive and concise user interface, making the operation more convenient and easy to understand; users can easily grab orders, dispatch orders, and monitor And data analysis and other operations can be used without tedious training;

7. Real-time communication and collaboration of the present invention: The communication module in the system realizes real-time communication and collaboration between vehicles and dispatch centers, drivers and dispatchers; problems can be solved quickly through functions such as instant messaging, voice calls or video conferencing. Adjust transportation plans and keep information updated;

8. Powerful algorithm support of the present invention: Our system adopts advanced scheduling algorithms and can quickly generate optimal scheduling solutions while considering multiple constraints; this includes considering factors such as vehicle capacity, distance, and cargo urgency. , to minimize the total transportation cost or maximize resource utilization;

9. The present invention is highly automated: the automation function in the system greatly reduces the need for manual intervention and improves the efficiency and accuracy of operations; for example, the system can automatically generate scheduling tasks, send notifications and alarms, record transportation data, etc., reducing human error and tedious manual operations;

10. Data security and privacy protection in this invention: Our system adopts strict data encryption and access control mechanisms to ensure the security and privacy protection of transportation data and user information; at the same time, we follow relevant regulations and standards to ensure the integrity of data. compliance and confidentiality.

Detailed ways

The technical solution of the present invention will be further described in detail below:

In order to make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention are described clearly and completely below. Obviously, the described embodiments are some, but not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without making creative efforts fall within the scope of protection of the present invention.

The purpose of this invention is to provide a short-barge vehicle cluster management and batch dispatching system and device applied to smart logistics parks, using information technology to improve system operation stability, simplify system use processes, enhance system adaptability, and improve system operation stability. The short-term barge vehicles within the market are uniformly dispatched and managed to achieve rapid transshipment and distribution of goods, thereby improving the circulation efficiency and service level of the agricultural product wholesale market.

Through this invention, the following layers are constructed to realize intelligent management of short-term barge dispatch within the agricultural batch market:

(1) Data collection layer: Use vehicle positioning technology, sensors and other equipment to collect data such as cargo demand, vehicle location, vehicle status and so on within the market, and send it to the data processing layer through wireless transmission;

(2) Data processing layer: This layer is responsible for receiving the data sent by the data collection layer, and performing data integration, data cleaning, data analysis and other processing. It uses data mining, machine learning and other technologies to extract useful information, such as vehicle scheduling optimization, cargo Batch management, etc., to obtain an optimized short-distance vehicle dispatching plan, provide decision support, such as transportation route optimization, vehicle dispatching plan, etc., and send it to the data execution layer through wireless transmission;

(3) Data execution layer: This layer is responsible for executing the plans sent by the data processing layer, such as the actual operation of short-distance vehicle dispatching and cargo batching, and providing relevant information to short-term vehicle drivers, warehouse managers, etc. through voice or graphical interfaces. The personnel sends instructions to guide them to carry out cargo transshipment and distribution according to the plan to ensure the accurate execution of the dispatch plan;

(4) User interface layer: Provide a user-friendly interface so that users (such as agricultural product wholesalers, dispatchers, etc.) can easily view and manage vehicle dispatching, cargo information, market demand and other related information. Users can manage orders, check transportation status, adjust scheduling plans and other operations through the interface.

(5) Data feedback layer: This layer is responsible for collecting the feedback information of the short-distance vehicle driver's execution of the dispatch plan and sending it to the data processing layer through wireless transmission as the basis for the data processing layer to optimize the dispatch plan.

The present invention mainly uses reinforcement learning algorithms, optimization algorithms, real-time monitoring and prediction algorithms to realize the core functions of the data processing layer.

(1) Reinforcement learning algorithm: It can continuously update its own policy network parameters based on the reward signal provided by the data feedback layer, thereby learning the optimal or nearly optimal short-distance vehicle dispatching plan. Specifically, the reinforcement learning algorithm includes the following elements:

①State space: The state space includes the characteristics of the current problem, the current solution, and the solution process. Specifically, the state space consists of the following parts:

Goods demand vector: represents the quantity of goods required for each goods demand point (including warehouses);

Vehicle position vector: represents the current location of each short-distance vehicle;

Vehicle status vector: represents the current quantity of goods carried by each short-distance vehicle;

Route length vector: represents the current route length of each shuttle vehicle;

Route sequence vector: represents the sequence of cargo demand points currently visited by each short-distance vehicle;

Remaining demand vector: represents the remaining quantity of unsatisfied goods at each goods demand point (including warehouses);

Remaining capacity vector: Indicates the remaining amount of cargo that each short-distance vehicle can carry.

② Action space: The action space includes all possible short-distance vehicle dispatching plans. Specifically, the action space consists of the following parts:

Vehicle selection action: select a short-distance vehicle that is currently available (that is, has not completed the task and is not overloaded) for dispatch;

Demand point selection action: Select a goods demand point that has not been visited currently and has remaining demand (i.e. non-zero) as the next visit target;

Cargo transfer action: After arriving at the target demand point, determine how much goods to unload or load from that point based on the demand situation and vehicle status at that point.

③Reward function: The reward function is used to evaluate the contribution of each action to the overall objective function (ie, minimizing the total route length). Specifically, the reward function consists of the following parts:

Fixed reward: means that for every visit to a goods demand point (except the warehouse), a fixed value (such as 1) reward will be obtained;

Penalty factor: means that every time the route length increases by one unit, a proportional value (such as 0.01) will be deducted as a reward;

Termination reward: It means that when all goods demand points are visited and satisfied, a reward with an additional value (such as 10) will be obtained.

④ Policy network: The policy network is used to output the optimal or approximately optimal action probability distribution based on the current state. Specifically, the policy network consists of the following parts:

Input layer: The input layer receives each vector in the state space as input and splices it into a one-dimensional vector;

Hidden layer: The hidden layer consists of several fully connected layers or convolutional layers, and uses activation functions (such as ReLU) to add nonlinear features;

Output layer: The output layer consists of three sub-output layers, which respectively correspond to three parts of the action space, namely vehicle selection action, demand point selection action and cargo transfer action. specifically:

A. Vehicle selection sub-output layer: This sub-output layer consists of a fully connected layer and a softmax layer, and outputs a vector of length K, indicating the probability of selecting each short-distance vehicle;

B. Demand point selection sub-output layer: This sub-output layer consists of a fully connected layer and a softmax layer. It outputs a vector of length N+1, indicating the selection of each goods demand point (including warehouse) as the next access target. The probability;

C. Cargo transfer sub-output layer: This sub-output layer consists of a fully connected layer and a softmax layer. It outputs a vector with a length of C+1, which represents the unloading or loading of 0 to C units of goods from the target demand point. Probability.

Based on the output of these three sub-output layers, a complete action can be obtained, that is, which vehicle is selected, which demand point is visited, and how many goods are transferred. Then the reward value of the action is calculated according to the reward function, and the parameters of the policy network are updated according to the policy gradient algorithm.

(2) Vehicle scheduling algorithm: Based on data such as vehicle location, market demand and cargo information, use optimization algorithms (such as genetic algorithms, simulated annealing algorithms, etc.) to determine the optimal vehicle scheduling plan, including route planning, vehicle allocation, etc.

(3) Cargo batching algorithm: Based on data such as market demand and vehicle availability, optimization algorithms (such as dynamic programming, greedy algorithms, etc.) are used to determine a reasonable batching plan for goods to improve distribution efficiency and meet market demand.

(4) Real-time monitoring and prediction algorithm: By real-time monitoring of vehicle location, traffic conditions and other data, and combined with historical data for predictive analysis, potential problems can be discovered in a timely manner and corresponding adjustments can be made, such as route re-planning, vehicle substitution, etc.

Intelligent management of short-term barge dispatch in the agricultural batch market is realized through algorithms:

(1) Vehicle dispatching optimization: There are a large number of short-term shuttle vehicles in traditional agricultural product wholesale markets, and dispatching and managing these vehicles has become complex and difficult. The system uses cluster management algorithms, combined with vehicle positioning and cargo information, to automatically formulate optimal dispatch plans. Through the optimization of the algorithm, the empty rate of vehicles can be reduced, the waiting time can be shortened, and the transportation efficiency and reduction can be improved. low cost.

(2) Goods management in batches: Agricultural product wholesale markets usually have a large number of goods that need to be distributed in batches to different buyers. The system uses data processing and analysis technology to combine cargo demand and vehicle availability to achieve intelligent cargo batch management. Through algorithm optimization, the order and time of goods batches can be reasonably arranged to improve distribution efficiency and meet market demand.

(3) Real-time vehicle monitoring: Through vehicle positioning technology and communication technology, the system can monitor the location, status and driving conditions of short-distance shuttle vehicles in real time. By integrating real-time data, the system can promptly discover and solve potential problems, such as traffic congestion, vehicle failures, etc., improving the reliability and safety of transportation.

(4) Data analysis and decision support: The system collects and analyzes a large amount of data, including vehicle positioning data, cargo information, market demand, etc. Through data mining, machine learning and optimization algorithms, the system can provide decision support, such as predicting market demand, optimizing transportation routes and allocating vehicles in advance, to improve the accuracy and efficiency of decision-making.

The technical solution includes the following technical architecture:

(1) System purchase includes front-end and back-end parts

① The front-end part includes the user interface, which can be provided to administrators, short-distance vehicle drivers and other relevant personnel through web applications or mobile applications. The user interface provides functions such as user registration, task viewing, data query, and statistics, allowing users to interact with the system.

②The back-end part includes main modules such as cluster management, order grabbing function and order dispatching function. The cluster management module is responsible for managing the registration, status monitoring and task allocation optimization of short-distance vehicle clusters. The order-grabbing function module handles drivers’ order-grabbing requests, screening and matching, and task allocation. The dispatch function module is responsible for task release, vehicle screening and matching, and task scheduling.

(2) Cluster management server architecture:

The cluster management server is the core component of the system and is responsible for the cluster management, scheduling algorithm, data analysis and other functions of the entire system. Its architecture can adopt a distributed architecture, including the following modules:

① Cluster management module: used to monitor and manage the status, location and operation of short-distance vehicle clusters, and provide vehicle management, task allocation and other functions.

②Scheduling algorithm module: implements batch scheduling algorithm, generates the optimal scheduling plan based on real-time vehicle and cargo information, optimizes vehicle utilization, and reduces transportation costs.

③Data analysis module: Perform real-time analysis of vehicle and cargo data stored in the database to provide decision support, such as dispatch optimization, route planning, etc.

④Communication module: Responsible for data exchange and instruction issuance with the short-distance vehicle terminal equipment and dispatch center terminal equipment.

(3) Short-distance vehicle terminal equipment architecture:

Vehicle-mounted terminal equipment is a device installed on each short-distance shuttle vehicle and is used for functions such as data collection, communication and execution of dispatch instructions. Its architecture includes the following components:

① Sensor module: including GPS positioning sensor, speed sensor, load sensor, etc., used to collect the location, status and operating data of the vehicle.

② Communication module: Utilizes wireless network technology (such as 4G, 5G) and cluster management server for real-time data transmission and command issuance.

③Control module: Controls the loading, unloading and transportation operations of the vehicle according to the received dispatch instructions.

④Data storage module: Temporarily stores the collected data to ensure the real-time and reliability of the data.

(4) Dispatch center terminal equipment architecture:

The dispatch center terminal equipment is used for dispatchers to interact with the cluster management server to implement functions such as issuing dispatch instructions and data query. Its architecture includes the following components:

① User interface: Provide an intuitive and friendly user interface through which the dispatcher can interact with the cluster management server.

②Data exchange module: Responsible for data exchange with the cluster management server, including the issuance of scheduling instructions and data

The results of the query are returned.

③Communication module: Use network connection to communicate with the cluster management server in real time.

(5) Third-party logistics platform:

The platform is an external partner of the system and is responsible for providing business requirements, including cargo name, pickup address, delivery address, pickup time, requested arrival time and other information. The platform can be in the form of a cargo owner enterprise management system, a third-party logistics management system or an open business module. The platform is connected to the dispatch platform through an open interface to implement functions such as sending and receiving business requests, querying and updating order status, and acquiring and analyzing transportation data.

The system in this technical solution consists of software and hardware. Their compositions will be discussed separately below.

Software composition:

(1) Cluster management software: This is the core software module of the system and is used to manage the entire short-distance vehicle cluster. It includes vehicle registration, vehicle status monitoring, vehicle scheduling, task allocation and optimization, scheduling algorithms, data analysis and other functions. The software is implemented based on a cloud computing platform or distributed system and is responsible for coordinating vehicle operations, optimizing dispatch plans, and providing decision support.

(2) Order-grabbing function software: The system has an order-grabbing function, that is, the vehicle driver can view the tasks to be dispatched through the mobile application and voluntarily choose to accept the task. The order grabbing function can achieve real-time response and flexible scheduling, improving resource utilization. This software module should be able to receive order grabbing requests, screen and match drivers, and assign tasks to appropriate drivers.

(3) Dispatch function software: This software module is responsible for allocating tasks to appropriate short-distance vehicle drivers. It should be based on a certain scheduling algorithm, taking into account factors such as vehicle location, availability, real-time data, idle time, distance, etc., to dispatch orders in an optimal way. The software module also provides real-time task status monitoring and feedback capabilities.

(4) Data processing and analysis software: used to process and analyze the collected data, including monitoring of vehicle location and status, execution of scheduling algorithms, etc. The software can use technologies such as data mining, machine learning, and artificial intelligence to analyze data and support decision-making.

(5) Scheduling algorithm software: The software module that implements the batch scheduling algorithm generates the optimal scheduling plan based on the scheduling rules and vehicle status. The software module can consider multiple factors such as vehicle priority, cargo type, route planning, etc. to achieve efficient dispatch decisions.

Hardware composition:

(1) Short-distance vehicle terminal equipment: Each short-distance vehicle is equipped with on-board terminal equipment for real-time communication with the system, and for data interaction and instruction execution with the cluster management server. The vehicle-mounted terminal equipment can receive task dispatch, send location information, etc., to achieve seamless connection with the system.

(2) Cluster management server: A server device used to centrally manage short-distance shuttle vehicles, store and process vehicle and cargo data, and execute scheduling algorithms and decision support. Servers need to have high-performance computing and storage capabilities to handle large-scale vehicle and cargo data processing.

(3) GPS positioning system: Short-distance shuttle vehicles are equipped with a GPS positioning system, which can obtain the vehicle's location information in real time and transmit it to the system. The GPS positioning system can provide accurate vehicle location, speed, mileage and other data, providing basic data support for task scheduling.

(4) Dispatch center terminal equipment: terminal equipment used for issuing and monitoring dispatch instructions, including computers, tablets or smart phones, etc. Through this device, the dispatcher can issue dispatching instructions to the short-distance vehicle terminal equipment and monitor the location and status of the vehicle in real time.

(5) Communication network: The system needs to establish a reliable communication network for real-time data transmission and communication between the vehicle and the system. This may be a wireless network, a mobile communications network or other applicable communications technology.

This technical solution includes the design of the following databases:

(1) Vehicle information database:

The database is used to store and manage information related to short-distance shuttle vehicles, including but not limited to the following:

①Vehicle number: a number that uniquely identifies each short-distance vehicle.

②Location information: Record the real-time location coordinates of the vehicle for real-time monitoring and dispatching.

③Status information: including the vehicle's operating status, load status, etc., used to determine vehicle availability and make scheduling decisions.

(2) Cargo information database:

This database is used to store and manage cargo-related information, including but not limited to the following:

① Cargo number: A number that uniquely identifies each cargo.

②Cargo type: Record the type of cargo for scheduling and matching with appropriate vehicles.

③Quantity information: Record the quantity of goods for scheduling and distribution planning.

(3) Scheduling rule database:

This database is used to store and manage scheduling rules, including but not limited to the following:

① Vehicle scheduling priority: Set the vehicle scheduling priority according to different conditions (such as vehicle type, idle time, etc.) to ensure optimal scheduling decisions.

②Batch scheduling algorithm: stores the relevant parameters and rules of the batch scheduling algorithm and is used to generate the optimal batch scheduling plan.

③Scheduling strategy: Record dispatching rules, such as vehicle loading rules, delivery route planning, etc., to ensure the rationality and efficiency of dispatching.

(4) Historical data database:

This database is used to store and manage historical data for subsequent data analysis and decision support, including but not limited to the following:

① Vehicle transportation records: Record the transportation history of each vehicle, including starting location, destination, transportation time, etc., used to analyze the transportation efficiency and quality of the vehicle.

② Goods delivery time record: Record the delivery time of each goods in order to predict and optimize the delivery time.

These databases can be stored and managed using relational databases (such as MySQL) or distributed databases (such as Hadoop) to meet the system's requirements for real-time, reliability and scalability of data.

To sum up, the database design of short-term barge vehicle cluster management and batch dispatching system includes vehicle information database, cargo information database, dispatching rules database and historical data database, which are used to store and manage relevant real-time and historical data to support the system. scheduling decision-making and analysis functions.

Contains a series of algorithm steps. The following is a detailed description of its algorithm steps:

(1) Cluster management algorithm steps:

a. Vehicle registration: The short-distance shuttle vehicle is registered in the system and relevant information is provided, such as vehicle number, load capacity, etc.

b. Vehicle status monitoring: The system monitors the location, availability and operating status of the vehicle in real time. This can obtain real-time data of the vehicle through on-board sensors, GPS and other equipment, and feed it back to the cluster management software.

c. Task allocation and optimization: Use task allocation and optimization algorithms to allocate tasks to appropriate vehicles based on the vehicle’s location, availability and real-time data. The algorithm takes into account the vehicle's load capacity, driving distance, time window and other factors to achieve efficient task allocation.

(2) Task release:

a. The system dispatches tasks to the most suitable vehicle driver based on vehicle location, task type, real-time traffic and other information.

b. During the order dispatch process, the system needs to consider factors such as the vehicle's current task status, task priority, and vehicle working hours to ensure that the task can be completed in time.

c. After dispatch, the system updates the task status and related information assigned to the vehicle to facilitate subsequent scheduling and monitoring.

(3) Algorithm steps of order grabbing function:

a. Receiving order grabbing requests: When there is a task that needs to be executed, the system will publish the task information, including task type, starting location, destination, etc. Short-distance vehicle drivers can receive task requests through the order grabbing function software.

b. Driver screening and matching: The system screens and matches order grabbing requests based on task requirements and driver capabilities and availability. This can be assessed based on factors such as the driver's location, load capacity, time available, etc. to find the most suitable driver.

c. Task allocation: The system assigns tasks to the most appropriate driver to ensure that the task can be executed in a timely manner. The task allocation algorithm can take into account the driver's distance, traffic conditions, task priority and other factors to achieve the best task allocation effect.

(4) Order dispatch function algorithm steps:

a. Task release: The system releases task information based on task requirements and priority, including task type, starting location, destination, etc.

b. Vehicle screening and matching: The system screens and matches vehicles for tasks based on the location, availability and real-time data of the vehicle. The algorithm takes into account the vehicle's load capacity, travel distance, time window and other factors to find the most suitable vehicle.

c. Task scheduling: The system schedules tasks to the most suitable vehicles and generates corresponding scheduling plans. The scheduling algorithm can take into account the vehicle's route, time window, priority and other factors to achieve the optimal task scheduling effect.

(5) Scheduling optimization:

a. The system collects and analyzes vehicle historical data, including operating trajectory, task completion time, waiting time and other information.

b. Based on historical data and real-time information, the system can optimize dispatching strategies and improve vehicle dispatching efficiency and task completion rate.

c. Scheduling optimization can include route planning, task batch scheduling, task priority adjustment and other technical means to minimize overall cost and time.

(6) System monitoring and updating:

a. The system monitors vehicle location, task progress, driver status and other information in real time to facilitate real-time scheduling and task tracking.

b. The system dynamically updates task status and allocation based on the actual situation of the vehicle and task progress.

c. The system can also provide real-time data statistics and reports to facilitate decision-making and optimization by managers.

(7) Steps of evaluation algorithm:

a. Receive the operation data of the shuttle bus, such as speed, fuel consumption, mileage, load and other information.

bReceive customer satisfaction feedback, such as ratings, comments and other information.

c. Receive the completion status of the order, such as whether it arrived on time, whether it was in good condition, etc.

d. Evaluate the service quality of the shuttle bus based on the operation data of the shuttle bus, customer satisfaction feedback, order completion and other indicators, and give corresponding reward and punishment measures, such as extra points, subtracted points, bonuses, and fines. wait.

e. Send the evaluation results and reward and punishment measures to the shuttle bus through the wireless communication module.

(5) Steps of optimization algorithm:

a. Receive indicators such as service quality, transportation efficiency, and operating costs of short shuttle buses.

b According to the changing trend and target value of the indicator, optimize and update the batch scheduling algorithm and order grabbing algorithm, such as adjusting parameters, adding constraints, improving strategies, etc.

c Deploy the optimized batch scheduling algorithm and order grabbing algorithm to the central server and mobile terminals.

The invention improves efficiency: the system realizes intelligent management and dispatching of vehicle clusters through the functions of grabbing orders and dispatching orders; vehicles can actively grab orders or be dispatched according to actual needs, avoiding unnecessary waiting in traditional dispatching methods. Time and resources are wasted and transportation efficiency is improved; the invention optimizes resource utilization: the system uses intelligent scheduling algorithms to allocate goods in batches to the most suitable vehicles for transportation, avoiding no-load or low-load operation of vehicles and maximizing optimization improve resource utilization efficiency. This can reduce transportation costs and improve operational efficiency; the invention improves service quality: the dispatch function in the system can quickly and accurately allocate tasks according to the location, status of the vehicle and the requirements of the goods; this helps to improve the response speed of transportation services and accuracy to meet customer needs and improve customer satisfaction; real-time monitoring and management of the present invention: the system has a real-time monitoring function and can conduct real-time monitoring and management of the vehicle's location, transportation progress, driving trajectory, etc.; this helps enterprises Carry out timely tracking and adjustment of vehicle operation conditions to improve the safety and controllability of transportation; data analysis and decision support of the present invention: the system collects and analyzes vehicle operation data and provides data analysis on transportation efficiency, cost, customer needs, etc. Report; This provides data support for enterprises to help them make reasonable decisions, optimize operational strategies, and further improve transportation efficiency and profitability; User-friendliness of the invention: Our system design takes into account the needs and usage habits of users, and provides The intuitive and concise user interface makes the operation more convenient and easy to understand; users can easily perform operations such as order grabbing, order dispatching, monitoring and data analysis, and can get started without tedious training; the real-time communication and collaboration system of the present invention The communication module in the system realizes real-time communication and collaboration between vehicles and dispatch centers, drivers and dispatchers; through functions such as instant messaging, voice calls or video conferencing, problems can be quickly solved, transportation plans can be adjusted, and information can be kept updated in a timely manner ; Powerful algorithm support of the present invention: Our system adopts advanced scheduling algorithms, which can quickly generate the optimal scheduling plan while considering multiple constraints; this includes considering factors such as vehicle capacity, distance, cargo urgency, etc. To minimize the total transportation cost or maximize resource utilization; the invention is highly automated: the automation function in the system greatly reduces the need for manual intervention and improves the efficiency and accuracy of operations; for example, the system can automatically generate scheduling tasks, send Notifications and alarms, recording transportation data, etc., reduce human errors and tedious manual operations; data security and privacy protection of the present invention: Our system adopts strict data encryption and access control mechanisms to ensure the security of transportation data and user information and privacy protection; at the same time, we follow relevant regulations and standards to ensure data compliance and confidentiality.

Claims (7)

1. A cluster management and dispatch system for short-range vehicles, characterized by: the following layers are constructed to realize the agricultural market

Intelligent management of short-cut scheduling, specifically as follows:

(1) The data acquisition layer is used for acquiring the goods demand, the vehicle position and the vehicle state data in the market by utilizing a vehicle positioning technology and sensor equipment and transmitting the data to the data processing layer in a wireless transmission mode;

(2) The data processing layer is used for receiving the data sent by the data acquisition layer, carrying out data integration, data cleaning and data analysis processing, extracting useful information by utilizing data mining and machine learning technologies, and carrying out vehicle dispatching optimization and cargo batch management, so that an optimized short-distance vehicle dispatching scheme is obtained, decision support such as transportation route optimization and vehicle dispatching scheme is provided, and the decision support is sent to the data execution layer in a wireless transmission mode;

(3) Data execution layer: the system is used for executing a scheme sent by a data processing layer, such as actual operation of dispatching of short-range vehicles and batched goods, sending instructions to related personnel of drivers and warehouse administrators of the short-range vehicles in a voice or graphical interface mode, guiding the drivers and the warehouse administrators to transfer and distribute the goods according to the scheme, and ensuring accurate execution of a dispatching plan;

(4) The user interface layer is used for providing a user-friendly interface, so that a user can conveniently view and manage vehicle dispatching, cargo information and market demand related information, and the user can perform order management, inquire transportation state and adjust dispatching plan operation through the interface;

(5) The data feedback layer is used for collecting feedback information of the dispatching scheme executed by the driver of the short-range vehicle, and sending the feedback information to the data processing layer in a wireless transmission mode to serve as a basis for optimizing the dispatching scheme by the data processing layer.

2. The short-range vehicle cluster management and dispatch system of claim 1, wherein: the data place

The processing layer receives the data sent by the data acquisition layer by adopting a reinforcement learning algorithm, an optimization algorithm and a real-time monitoring and predicting algorithm, integrates the data, cleans the data and analyzes the data, and extracts useful information by utilizing data mining and machine learning technologies.

3. The short-range vehicle cluster management and dispatch system of claim 2, wherein: the chemical enhancement is

Xi Suanfa, specifically the following:

continuously updating the strategy network parameters of the self according to the reward signals provided by the data feedback layer, so as to learn an optimal or nearly optimal short-distance vehicle scheduling scheme; specifically, the reinforcement learning algorithm includes the following elements:

(1) State space: the state space comprises the characteristics of the current problem, the current solution and the solving process; specifically, the state space is composed of the following parts:

cargo demand vector: representing that each cargo demand point includes a number of cargo required by the warehouse;

vehicle position vector: representing the current position of each short-cut vehicle;

vehicle state vector: representing the number of cargo currently carried by each short-range vehicle;

route length vector: representing a current traveled route length of each short vehicle;

route sequence vector: representing a sequence of cargo demand points that each short-range vehicle has currently visited;

residual demand vector: representing that each cargo demand point comprises the number of unsatisfied cargoes remaining in the warehouse;

residual capacity vector: representing the number of cargo remaining per short-range vehicle;

(2) action space: the action space includes all possible short-barge vehicle scheduling schemes; specifically, the action space is composed of the following parts:

vehicle selection actions: selecting a currently available short-range vehicle which is not completed in tasks and is not overloaded for scheduling;

demand point selection action: selecting a cargo demand point which is not accessed currently and has residual demand, namely non-zero, as a next access target;

And (3) cargo transferring action: after reaching the target demand point, determining how much goods are unloaded or loaded from the point according to the demand condition of the point and the state condition of the vehicle;

(3) bonus function: the reward function is used to evaluate the contribution of each action to the overall objective function, i.e. minimizing the total route length; specifically, the reward function consists of the following parts:

fixed rewards: indicating that a fixed value of rewards are obtained every time a goods demand point is accessed except a warehouse;

penalty factor: deducting a proportional value of the reward every time the route length of one unit length is increased;

terminating the bonus: indicating that when all cargo demand points are accessed and satisfied, an extra value is awarded;

(4) policy network: the strategy network is used for outputting optimal or near optimal action probability distribution according to the current state; specifically, the policy network consists of the following parts:

input layer: the input layer receives each vector in the state space as input and splices the vectors into a one-dimensional vector;

hidden layer: the hidden layer is composed of a plurality of full-connection layers or convolution layers, and nonlinear characteristics are added by using an activation function;

output layer: the output layer consists of three sub-output layers, which respectively correspond to three partial action spaces, namely a vehicle selection action, a demand point selection action and a goods transfer action; specifically:

A. Vehicle selection sub-output layer: the sub-output layer consists of a full connection layer and a softmax layer, outputs a vector with the length of K, and represents the probability of selecting each short vehicle;

B. the demand point selection sub-output layer: the sub-output layer consists of a full connection layer and a softmax layer, outputs a vector with the length of n+1, and represents the probability of selecting each goods demand point to comprise a warehouse as the next access target;

C. cargo transfer sub-output layer: the sub-output layer consists of a full connection layer and a softmax layer, and outputs a vector with the length of C+1, which represents the probability of unloading or loading 0 to C units of goods from a target demand point;

according to the output of the three sub-output layers, a complete action is obtained, namely which vehicle is selected, which demand point is accessed and how much goods are transferred; and calculating the rewarding value of the action according to the rewarding function, and updating the parameters of the strategy network according to the strategy gradient algorithm.

4. The short-range vehicle cluster management and dispatch system of claim 2, wherein:

vehicle scheduling algorithm: according to the vehicle position, market demand and cargo information data, an optimal vehicle dispatching scheme is determined by utilizing an optimization algorithm, wherein the optimal vehicle dispatching scheme comprises route planning and vehicle distribution;

Cargo batching algorithm: based on market demand and vehicle availability data, an optimization algorithm is adopted to determine a reasonable batching scheme of goods so as to improve distribution efficiency and meet market demand;

real-time monitoring and prediction algorithm: the method comprises the steps of monitoring vehicle position and traffic condition data in real time, carrying out prediction analysis by combining historical data, finding potential problems in time and carrying out corresponding adjustment, such as route re-planning and vehicle replacement.

5. An architecture based on the short-barge cluster management and scheduling system according to any one of claims 1 to 4, characterized in that: comprises

Front-rear end section architecture:

the front end portion includes a user interface that can be provided for use by administrators, short-range vehicle drivers, and other related personnel through a web application or mobile application; the user interface provides the functions of user registration, task viewing, data query and statistics, so that the user can interact with the system;

the back end part comprises a main module of cluster management, a single robbing function and a single dispatching function; the cluster management module is responsible for managing registration, state monitoring and task allocation optimization of the short-range vehicle cluster; the order-robbing functional module processes order-robbing requests of drivers, screening matching and task allocation; the dispatch function module is responsible for task release, vehicle screening matching and task scheduling;

Cluster management server architecture:

the cluster management server is a core component of the system and is responsible for cluster management, a scheduling algorithm and a data analysis function of the whole system; the architecture adopts a distributed architecture and comprises the following modules:

the cluster management module is used for monitoring and managing the state, the position and the running condition of the short-distance vehicle cluster and providing vehicle management and task allocation functions;

the scheduling algorithm module is used for realizing a batch scheduling algorithm, generating an optimal scheduling scheme according to real-time vehicle and cargo information, optimizing the vehicle utilization rate and reducing the transportation cost;

the data analysis module is used for analyzing the vehicle and cargo data stored in the database in real time and providing decision support such as scheduling optimization and route planning;

the communication module is responsible for carrying out data exchange and instruction issuing with the short-distance vehicle terminal equipment and the dispatching center terminal equipment;

short-range vehicle terminal equipment architecture:

the vehicle-mounted terminal equipment is equipment installed on each short-range vehicle and is used for data acquisition, communication and execution of scheduling instruction functions; the framework comprises the following components:

a sensor module: the system comprises a GPS positioning sensor, a speed sensor and a load sensor, wherein the GPS positioning sensor, the speed sensor and the load sensor are used for acquiring the position, the state and the operation data of a vehicle;

And a communication module: the system is used for transmitting real-time data and issuing instructions with the cluster management server by utilizing a wireless network technology;

the control module is used for controlling loading, unloading and transportation operations of the vehicle according to the received scheduling instruction;

the data storage module is used for temporarily storing the acquired data and ensuring the real-time performance and reliability of the data;

scheduling center terminal equipment architecture:

the dispatching center terminal equipment is used for interacting a dispatcher with the cluster management server to realize the functions of issuing dispatching instructions and data query; the framework comprises the following components:

the user interface is used for providing an intuitively friendly user interface, and a dispatcher can interact with the cluster management server through the interface;

the data exchange module is used for carrying out data exchange with the cluster management server and comprises the issuing of a scheduling instruction and the return of a data query result;

the communication module is used for carrying out real-time communication with the cluster management server by utilizing network connection;

third party logistics platform:

the platform is an external partner of the system and is responsible for providing service requirements, including goods names, goods picking addresses, delivery addresses, goods picking time and time information of arrival requirements; the platform is in the form of a cargo owner enterprise management system, a third party logistics management system or an open business module; the platform is connected with the dispatching platform through an open interface to realize the functions of sending and receiving service requests, inquiring and updating order states and acquiring and analyzing transportation data.

6. The architecture of a short-range vehicle cluster management and dispatch system of claim 5, wherein: the system also comprises a database, and specifically comprises the following steps:

vehicle information database:

the database is used for storing and managing information related to short-range vehicles, including but not limited to the following:

vehicle number: a number uniquely identifying each short-range vehicle;

position information: recording real-time position coordinates of the vehicle so as to monitor and schedule in real time;

status information: the method comprises the steps of judging the availability of the vehicle and carrying out scheduling decision, wherein the running state and the loading state of the vehicle are used for judging the availability of the vehicle;

cargo information database:

the database is used to store and manage information about the goods, including but not limited to the following:

cargo number: a number uniquely identifying each good;

the cargo category: recording the type of goods so as to schedule and match proper vehicles;

quantity information: recording the quantity of goods for scheduling and delivery plans;

scheduling rules database:

the database is used to store and manage scheduling rules including, but not limited to, the following:

vehicle dispatch priority: according to different conditions, setting the scheduling priority of the vehicle to ensure the optimal scheduling decision;

Batch scheduling algorithm: storing relevant parameters and rules of the batch scheduling algorithm for generating an optimal batch scheduling scheme;

scheduling policy: recording scheduling rules, such as vehicle loading rules and delivery route planning, so as to ensure the rationality and the high efficiency of scheduling;

historical data database:

the database is used to store and manage historical data for subsequent data analysis and decision support, including but not limited to the following:

vehicle transportation record: recording a transportation history of each vehicle, including a start place, a destination and a transportation time, for analyzing the transportation efficiency and quality of the vehicle;

and (5) cargo delivery time record: recording the delivery time of each cargo so as to predict and optimize the delivery time;

the databases can be stored and managed by adopting a relational database or a distributed database so as to meet the requirements of the system on the real-time property, the reliability and the expansibility of the data;

the database design of the short-refuted vehicle cluster management and batch dispatching system comprises a vehicle information database, a cargo information database, a dispatching rule database and a historical data database, and is used for storing and managing relevant real-time and historical data so as to support the dispatching decision and analysis functions of the system.

7. A method based on the short-barge cluster management and scheduling system according to any one of claims 1 to 6, characterized in that: the method specifically comprises the following steps:

the cluster management algorithm specifically comprises the following steps:

vehicle registration: the short-range vehicles are registered in the system and provide relevant information such as vehicle numbers and loading capacity;

vehicle condition monitoring: the system monitors the position, availability and running state of the vehicle in real time; acquiring real-time data of a vehicle through a vehicle-mounted sensor and GPS equipment, and feeding the real-time data back to cluster management software;

task allocation and optimization: assigning tasks to the appropriate vehicles using task assignment and optimization algorithms based on the location, availability, and real-time data of the vehicles; the algorithm considers the load capacity, the driving distance and the time window factors of the vehicle so as to realize efficient task allocation;

task release is specifically as follows:

the system distributes the task to the most suitable vehicle driver according to the position of the vehicle, the task type and the real-time traffic information;

in the dispatch process, the system needs to consider the current task condition, task priority and working time factors of the vehicle so as to ensure that the task can be completed in time;

after dispatch, the system updates the task state and related information assigned to the vehicle for subsequent scheduling and monitoring;

The robbery single function algorithm specifically comprises the following steps: :

and (5) receiving a robbery order request: when a task needs to be executed, the system issues task information, including a task type, a starting place and a destination; the driver of the short-range vehicle can receive a task request through the robbery-to-order function software;

driver screening and matching: the system screens and matches the robbery request according to the task requirement and the capability and availability of the driver; evaluating according to the position, the loading capacity and the available time factors of the driver to find the most suitable driver;

task allocation: the system distributes the task to the most suitable driver to ensure that the task can be executed in time; the task allocation algorithm can consider the distance of a driver, traffic conditions and task priority factors so as to realize the optimal task allocation effect;

the dispatch function algorithm specifically comprises the following steps:

task release: the system issues task information according to task demands and priorities, wherein the task information comprises task types, starting places and destinations;

vehicle screening and matching: the system screens and matches the vehicles for the tasks according to the positions, the availability and the real-time data of the vehicles; the algorithm considers the load capacity, the driving distance and the time window factors of the vehicle so as to find the most suitable vehicle;

Task scheduling: the system dispatches the task to the most suitable vehicle and generates a corresponding dispatch plan; the scheduling algorithm can consider the route, time window and priority factors of the vehicle so as to realize the optimal task scheduling effect;

the scheduling optimization is specifically as follows:

the system collects and analyzes historical data of the vehicle, including running track, task completion time and waiting time information;

based on the historical data and the real-time information, the system can optimize the scheduling strategy, and the scheduling efficiency and the task completion rate of the vehicle are improved;

scheduling optimization may include route planning, task batch scheduling, task priority adjustment techniques to minimize overall cost and time;

the system is monitored and updated as follows:

the system monitors the position, task progress and driver state information of the vehicle in real time so as to perform real-time scheduling and task tracking;

the system dynamically updates the task state and allocation according to the actual condition of the vehicle and the task progress;

the system can also provide real-time data statistics and reports so that management personnel can make decisions and optimize the data statistics and reports;

the evaluation algorithm specifically comprises the following steps:

receiving operation data of the short-distance vehicle, such as speed, oil consumption, mileage and load information;

Receiving satisfaction feedback of the clients, such as scoring and comment information;

receiving the completion condition of the order, if yes, arriving on time or not, and if not, obtaining information on whether the order is intact;

according to the running data of the short-distance vehicle, the satisfaction feedback of the customer and the completion condition index of the order, the service quality of the short-distance vehicle is evaluated, and corresponding rewarding and punishing measures such as adding, subtracting, rewarding and punishing are given;

the method comprises the steps of sending an evaluation result and rewarding and punishing measures to a short barge through a wireless communication module;

the steps of the optimization algorithm are as follows:

receiving service quality, transport efficiency and operation cost indexes of the short barge;

optimizing and updating the partitioned scheduling algorithm and the robbery algorithm according to the change trend and the target value of the index, such as parameter adjustment, constraint increase and strategy improvement;

and deploying the optimized batch scheduling algorithm and the optimized robbery scheduling algorithm to the central server and the mobile terminal.