A freight dispatch information management method, system and storage medium

Abstract

The present invention relates to the field of freight dispatching technology, and in particular to a freight dispatching information management method, system and storage medium. The method comprises the following steps: collecting unfinished orders on a freight platform to obtain an unfinished order set, wherein the unfinished order set contains a number of unfinished orders, and each unfinished order at least includes an order ID, order release time, target cargo attribute data, publisher identity data, origin, destination, order status identifier, order transportation task schedule, cargo stacking attributes, estimated delivery time and cargo expected arrival time; classifying and marking the order status of the unfinished order set to obtain a classified order set. The present invention can reduce the empty and half-loaded conditions of the vehicle by optimizing the loading space utilization of the vehicle, thereby reducing fuel consumption and transportation costs.

Description

Technical Field

[0001] The present invention relates to the technical field of freight dispatching, and in particular to a freight dispatching information management method, system and storage medium. Background Art

[0002] In the field of modern freight logistics, the demand for freight transportation is diversified and instant. However, the existing freight dispatching information management system often finds it difficult to adapt to this rapidly changing market demand. On the one hand, the system lacks flexible dispatching strategies when dealing with instant delivery and scheduled delivery orders, resulting in inefficient resource allocation. On the other hand, due to the diversity of transport vehicle types and cargo volumes, the existing system often fails to achieve the optimal loading solution when matching vehicles and cargo, resulting in a waste of transportation resources. In actual operations, the loading space utilization rate of transport vehicles is a key efficiency indicator. The current freight dispatching system often fails to effectively utilize the loading space of the vehicle when dealing with small cargo. For example, when a large truck receives an order for a small item, most of its huge loading space is idle. This situation not only increases transportation costs, but also leads to safety hazards in the transportation of goods, because small items are easily displaced or damaged during transportation. Summary of the invention

[0003] Based on this, it is necessary for the present invention to provide a freight scheduling information management method, system and storage medium to solve at least one of the above technical problems.

[0004] To achieve the above object, a freight dispatch information management method comprises the following steps:

[0005] Step S1: Collect unfinished orders on the freight platform to obtain an unfinished order set, wherein the unfinished order set includes a number of unfinished orders, and each unfinished order at least includes an order ID, order release time, target goods attribute data, publisher identity data, origin, destination, order status identifier, order transportation task schedule, estimated delivery time and expected arrival time of goods;

[0006] Step S2: classifying and marking the order status of the unfinished order set to obtain a classified order set, wherein the classified order set includes a subset of orders in transit, a subset of orders interrupted in transit, a subset of orders to be transported, and a subset of orders not received;

[0007] Step S3: collecting real-time vehicle information of transport vehicles through the freight platform to obtain a transport vehicle data set; screening the loading capacity of transport vehicles according to the transport vehicle data set to obtain a vehicle data set to be allocated;

[0008] Step S4: According to the to-be-allocated vehicle data set, the transportation interruption order subset and the unaccepted order subset are scheduled and matched to obtain an order-vehicle matching table, and the order transportation status of the transportation interruption order subset is updated according to the order-vehicle matching table, and the order transportation status is marked as rescheduled.

[0009] The present invention can fully grasp the order status of the freight platform, including detailed information of the order and the progress of the transportation task, by collecting unfinished order information in real time. Comprehensive information collection enables the scheduling strategy to more accurately adapt to the diversified and instant needs of the market. Compared with the traditional scheduling system, this can more flexibly handle instant delivery and scheduled delivery orders, thereby significantly improving the efficiency of resource allocation. By classifying and marking the status of unfinished orders, orders in different states can be clearly distinguished, such as orders in transportation, transportation interruption, to be transported and unaccepted orders. This classification not only helps to identify and prioritize urgent and critical transportation tasks, but also can adopt corresponding scheduling strategies for orders in different states, thereby improving the flexibility and response speed of scheduling. By collecting transportation vehicle information in real time and screening the loading capacity of the vehicle, the reasonable allocation of vehicle resources is ensured. This can accurately match the actual loading capacity of the vehicle with the volume and weight of the goods, thereby improving the utilization rate of the vehicle's loading space. This not only reduces the waste of transportation resources, but also reduces transportation costs. Through scheduling matching, the vehicles to be allocated are effectively connected with the orders with transportation interruption or unaccepted orders, thereby realizing the dynamic optimization configuration of transportation resources. This can not only quickly respond to emergencies such as transportation interruptions, but also make adjustments based on real-time changing transportation needs and vehicle status to ensure that the goods can be delivered to the destination in a timely and safe manner. In summary, the present invention not only improves the efficiency and accuracy of freight scheduling, but also reduces transportation costs and enhances the flexibility and adaptability of the system.

[0010] Preferably, the present invention further provides a freight dispatching information management system for executing the freight dispatching information management method as described above, the freight dispatching information management system comprising:

[0011] The order collection module is used to collect unfinished orders on the freight platform to obtain an unfinished order set, wherein the unfinished order set contains a number of unfinished orders, and each unfinished order at least includes the order ID, order release time, target goods attribute data, publisher identity data, origin, destination, order status identification, order transportation task schedule, estimated delivery time and expected arrival time of goods;

[0012] An order classification module is used to classify and mark the order status of the unfinished order set to obtain a classified order set, wherein the classified order set includes a subset of orders in transit, a subset of orders interrupted in transit, a subset of orders to be transported, and a subset of orders not received;

[0013] The vehicle screening module is used to collect real-time vehicle information of transport vehicles through the freight platform to obtain a transport vehicle data set; the loading capacity of the transport vehicles is screened according to the transport vehicle data set to obtain a vehicle data set to be allocated;

[0014] The scheduling module is used to perform intelligent multi-dimensional parameter scheduling optimization on the subset of transportation interrupted orders and the subset of unaccepted orders based on the data set of vehicles to be allocated, obtain the order-vehicle matching table, and update the order status of the freight platform based on the subset of orders being transported and the order-vehicle matching table.

[0015] The present invention collects detailed information of unfinished orders in real time through the order collection module, ensuring that the scheduling decision is based on the latest and most complete data. This reduces the possibility of information lag and error, thereby improving the efficiency and accuracy of scheduling. The order classification module can quickly classify the status of a large number of orders and divide the orders into different categories such as being transported, transport interrupted, to be transported and unaccepted. This can give priority to key orders according to the urgency and status of the orders, thereby optimizing resource allocation and improving transportation efficiency. The vehicle screening module collects information of transport vehicles in real time and screens them according to the loading capacity of the vehicles, ensuring the rational use of vehicle resources. This can flexibly adapt to changes in vehicle status, such as vehicle failure, maintenance, etc., and adjust the scheduling plan in time. The scheduling module ensures that the goods can be transported in a timely and safe manner by accurately matching the vehicles to be allocated with the orders of transport interrupted or unaccepted orders. This reduces the delay and loss of goods during transportation and improves the safety of cargo transportation. By optimizing the utilization rate of the vehicle's loading space, the empty and half-loaded conditions of the vehicle can be reduced, thereby reducing fuel consumption and transportation costs. At the same time, the efficient scheduling of the system reduces the residence time of goods in the warehouse and reduces storage costs.

[0016] Preferably, a computer-readable storage medium stores a freight scheduling information management method as described above that can be loaded and executed by a processor. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Other features, objects and advantages of the present invention will become more apparent from reading the detailed description with reference to the following drawings:

[0018] Figure 1 A schematic flow chart of the steps of a freight dispatch information management method according to an embodiment is shown.

[0019] Figure 2 A schematic diagram of a vehicle load constraint matrix according to an embodiment is shown.

[0020] Figure 3 A schematic diagram of an order time window constraint set according to an embodiment is shown.

[0021] Figure 4 A schematic diagram of a cargo stacking rule matrix according to an embodiment is shown.

[0022] Figure 5 A schematic diagram of a vehicle load balancing indicator matrix according to an embodiment is shown.

[0023] Figure 6 A schematic diagram of a path continuity indicator matrix according to an embodiment is shown.

[0024] Figure 7 A schematic diagram of a customer priority weight matrix according to an embodiment is shown. DETAILED DESCRIPTION

[0025] The technical method of the present invention is described clearly and completely below in conjunction with the accompanying drawings. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by technicians in this field without creative work are within the scope of protection of the present invention.

[0026] In addition, the accompanying drawings are only schematic illustrations of the present invention and are not necessarily drawn to scale. The same reference numerals in the figures represent the same or similar parts, and their repeated description will be omitted. Some of the block diagrams shown in the accompanying drawings are functional entities and do not necessarily correspond to physically or logically independent entities. The functional entities can be implemented in software form, or implemented in one or more hardware modules or integrated circuits, or implemented in different networks and / or processor methods and / or microcontroller methods.

[0027] It should be understood that, although the terms "first", "second", etc. may be used herein to describe various units, these units should not be limited by these terms. These terms are used only to distinguish one unit from another unit. For example, without departing from the scope of the exemplary embodiments, the first unit may be referred to as the second unit, and similarly the second unit may be referred to as the first unit. The term "and / or" used herein includes any and all combinations of one or more of the listed associated items.

[0028] To achieve this, please refer to Figures 1 to 7 The present invention provides a method for managing freight dispatch information, comprising the following steps:

[0029] Step S1: Collect unfinished orders on the freight platform to obtain an unfinished order set, wherein the unfinished order set includes a number of unfinished orders, and each unfinished order at least includes an order ID, order release time, target goods attribute data, publisher identity data, origin, destination, order status identifier, order transportation task schedule, estimated delivery time and expected arrival time of goods;

[0030] Step S2: classifying and marking the order status of the unfinished order set to obtain a classified order set, wherein the classified order set includes a subset of orders in transit, a subset of orders interrupted in transit, a subset of orders to be transported, and a subset of orders not received;

[0031] Step S3: collecting real-time vehicle information of transport vehicles through the freight platform to obtain a transport vehicle data set; screening the loading capacity of transport vehicles according to the transport vehicle data set to obtain a vehicle data set to be allocated;

[0032] Step S4: According to the to-be-allocated vehicle data set, the transportation interruption order subset and the unaccepted order subset are scheduled and matched to obtain an order-vehicle matching table, and the order transportation status of the transportation interruption order subset is updated according to the order-vehicle matching table, and the order transportation status is marked as rescheduled.

[0033] In this embodiment, first, an order management system (OMS) is used to collect unfinished orders on the platform and generate an unfinished order set. The OMS extracts detailed information of all unfinished orders, including order ID, release time, goods attributes, publisher identity, origin, destination, order status, schedule, estimated delivery time and expected arrival time, by executing a database query, and stores these data in the unfinished order set. Next, the unfinished order set is classified and marked by status to generate a classified order set. By analyzing the order status identifier, the order is assigned to the corresponding subset, such as the subset of orders in transit, interrupted transportation, to be transported and unaccepted orders. At the same time, the transportation management system (TMS) collects information on transport vehicles in real time, including the location, status, capacity, etc. of the vehicle, and creates a transport vehicle data set. The TMS receives real-time data of the vehicle through GPS and the vehicle communication system and updates the database. According to the maximum loading weight and volume of the vehicle, the transport vehicle data set is screened to generate a vehicle data set to be allocated. Finally, the scheduling matching tool matches the vehicle data set to be allocated with the interrupted order subset and the unaccepted order subset to generate an order-vehicle matching table. Finally, the best order-vehicle matching solution is determined by considering the real-time location of the vehicle, the priority of the goods, and the estimated time window. Once the matching is completed, the transportation status of the relevant orders is automatically updated and marked as "rescheduled". For example, suppose there are 50 unfinished orders and 30 vehicles. The OMS extracts the details of these orders from the database through SQL queries and stores them in the unfinished order set. The order classification tool then analyzes these orders and assigns them to four subsets. TMS collects vehicle information and uses a screening algorithm to determine 15 available vehicles. The scheduling matching tool matches these 15 vehicles with the orders in the subset, generates a matching table, and updates the order status.

[0034] Preferably, step S2 comprises the following steps:

[0035] Step S21: extracting order status identifiers from the unfinished order set to obtain an order status identifier set, wherein the order status identifier set includes a plurality of order status identifiers, and the order status identifier is any one of a pending order and an accepted order;

[0036] Specifically, an order management software can be used, which interacts with the freight platform's database interface. First, the software will execute an SQL query to select the order ID and current status of all uncompleted orders. The query results will be stored in an order status identifier set, where each data item is a dictionary containing an order ID and status.

[0037] Step S22: if the order status is marked as pending order, the corresponding order is recorded as an unaccepted order, thereby obtaining an unaccepted order subset;

[0038] Specifically, the order status identifier set can be traversed and the status of each order can be checked. For each order with the status of "pending order", the software will perform another database query to obtain the detailed information of the order. These details are then stored in a new collection, namely the subset of unaccepted orders.

[0039] Step S23: if the order status is marked as received, the corresponding order is recorded as an received order, thereby obtaining a subset of received orders;

[0040] Specifically, the order status identification set can be traversed to filter out all orders with a status identification of "accepted". Using conditional statements, such as if statements in Python, the software will check the status of each order. When encountering an order with a status of "accepted", the software will obtain the detailed information of the order through another database query, and the obtained information will be added to the "accepted order subset". For example, if there are three orders in the order status identification set with a status of "accepted", the software will perform three queries and store the detailed information of these orders in the accepted order subset to form a list.

[0041] Step S24: extracting the transport task schedule for the received order subset to obtain a plurality of order transport task schedules, wherein each order transport task schedule includes at least one transport stage event;

[0042] Specifically, order management software can be used to perform SQL queries on each order in the subset of received orders. This query will return the shipping task schedule associated with each order, including events for each shipping stage. The software will then parse these query results and create a shipping task schedule for each order, which includes at least one shipping stage event. For example, for order 100004, the shipping task schedule is as follows: [{'order_id': 100004, 'stage': 'Loaded', 'event_time': '2024-10-15 08:00'}, {'order_id': 100004, 'stage': 'In transit', 'event_time': '2024-10-15 09:00'}]. These schedules will be collected and stored in a collection.

[0043] Step S25: classifying the transportation progress status of the received order subset according to the order transportation task schedule to obtain a subset of orders being transported, a subset of orders interrupted in transportation, and a subset of orders to be transported;

[0044] Specifically, please refer to the sub-steps of step S25 for the detailed implementation process of this embodiment.

[0045] Step S26: Integrate the subset of unaccepted orders, the subset of orders in transit, the subset of orders with interrupted transit, and the subset of orders to be transported to obtain a classified order set.

[0046] Specifically, the data processing function of the order management software can be used to merge the subset of unaccepted orders, the subset of orders in transit, the subset of orders with interrupted transportation, and the subset of orders to be shipped. This function will create a new data structure, such as a list, and add the order information of each subset to this list in turn. When adding information, the software will ensure that the information of each order is complete, including order ID, status and other relevant details.

[0047] The present invention can quickly identify different types of orders by extracting and classifying the status identifiers of unfinished orders. This can quickly understand the overall situation of the current orders, thereby reducing processing time and improving work efficiency. By classifying orders into different categories such as unreceived orders, received orders, in transit, transport interrupted, and to be transported, it is possible to more reasonably allocate transportation resources, ensure that key orders are given priority, and optimize resource utilization. By extracting and classifying the progress of transportation tasks for received orders, the transportation status can be monitored in real time and potential transportation interruption problems can be discovered in a timely manner. This helps to take measures in advance to reduce the risk of cargo loss and delays and improve transportation safety.

[0048] Preferably, step S25 comprises the following steps:

[0049] Step S251: extracting the current transportation event from the order transportation task schedule to obtain the current transportation event, wherein the current transportation event includes the transportation event ID, transportation event description, transportation event start time and identity information of the transportation event person in charge;

[0050] Specifically, a SQL statement can be used to extract current shipping events. The SQL statement is designed to select all relevant information about the current shipping event with a specific order ID from the event table, including the event ID, description, start time, and identity information of the person in charge. The query result will return one or more records of current shipping events for each order. The software then stores these records in a data structure, such as a dictionary or object, where each order ID corresponds to a list containing current shipping event information.

[0051] Step S252: acquiring the corresponding origin and the corresponding destination of the order transportation task schedule based on the received order subset, and taking the origin as the first origin and the destination as the first destination;

[0052] Specifically, the order table and the associated transport task schedule can be accessed to execute a SQL query that is intended to select the origin and destination information in the transport task schedule associated with a specific order ID. For each order, the software will extract this information and mark it as the first origin and first destination, respectively. This information will then be stored in a data structure, such as a list or dictionary, where each order ID corresponds to an object containing the origin and destination information. For example, if the origin of order 100002 is "Shanghai" and the destination is "Beijing", create a record as follows: {'order_id': 100002, 'first_origin': 'Shanghai', 'first_destination': 'Beijing'}. This process can be repeated for all received orders.

[0053] Step S253: collecting completed orders from the freight platform to obtain a completed order set, wherein the completed order set includes a number of completed orders;

[0054] Specifically, you can use a SQL query to retrieve all order records marked as completed in the database. The query will return detailed information about all completed orders in the database, including order ID, product details, shipping time, customer information, etc. After the query is executed, the results will be exported to a dataset called the completed order set. This set contains multiple order records, each of which represents a successful shipment of goods.

[0055] Step S254: performing historical path matching on the completed order set according to the first origin and the first destination to obtain a matched completed order set, wherein the matched completed order set includes a plurality of matched completed orders;

[0056] Specifically, the origin and destination information of each order can be extracted from the completed order set. Then, for the first origin and first destination of each current order, orders with the same or similar origin and destination in the completed order set will be searched. For example, if the origin of the current order is Shanghai and the destination is Beijing, then all completed orders from Shanghai to Beijing will be found. The matching process involves calculating geographical distances, comparing route characteristics, or analyzing transportation time. After a successful match, a new set will be created, namely the matching completed order set, which contains all completed orders with similar transportation paths as the current order. For example, if five completed orders from Shanghai to Beijing are found, then these orders will be added to the matching completed order set.

[0057] Step S255: extracting historical transport task schedules from the matched completed order set to obtain a number of historical transport task schedules;

[0058] Specifically, a list of order IDs matching the completed order set can be determined. Then, using these order IDs, a series of SQL queries are executed, such as: SELECT * FROM shipment_progress WHERE order_id IN (order ID in the list). This query will retrieve all the shipping task progress records related to these completed orders. A historical shipping task schedule is generated for each order, which contains the events of each stage of the order during the transportation process. For example, for order ID 200001, the historical shipping task schedule includes the following information: loading date and time, departure date and time, date and time of each checkpoint on the way, expected arrival date and time, and actual arrival date and time. This information will be extracted and stored in a corresponding data structure, such as a dictionary or object, where each order ID corresponds to a list containing its historical shipping task progress information. This process can be repeated for each order in the matching completed order set, and finally a number of historical shipping task schedules are obtained.

[0059] Step S256: classifying the transport progress status of the received orders according to the historical transport task schedule and the current transport event, and obtaining an order status classification identifier;

[0060] Specifically, please refer to the sub-steps of step S256 for the detailed implementation process of this embodiment.

[0061] Step S257: Traverse the subset of received orders, and for each received order, execute steps S251-S256, that is, obtain several order status classification identifiers, and classify and integrate the progress status of the subset of received orders according to the order status classification identifiers to obtain the subset of orders in transit, the subset of orders with interrupted transportation, and the subset of orders to be transported.

[0062] Specifically, the received order subset can be traversed, and for each received order, steps S251-S256 are executed. After completing these operations, the status of each order is classified according to the acquired information, and an order status classification identifier is generated. For example, if the current transportation event of an order indicates that it has started transportation but has not yet reached the first checkpoint, it is classified as the "in-transit" state. If the transportation time spent on an order exceeds the warning time calculated based on historical data, it is classified as the "transport interruption" state. If an order has not yet started transportation, it is classified as the "pending transportation" state. Finally, the received order subset is classified and integrated according to these classification identifiers to form three new subsets: the in-transit order subset, the transportation interruption order subset, and the pending transportation order subset. For example, if the current transportation event of order 100007 is "departed" and the transportation time spent matches the warning time, it is classified into the in-transit order subset.

[0063] By extracting detailed information of current transportation events, the present invention can accurately grasp the specific situation of each transportation task, which helps to make more accurate scheduling decisions. By providing detailed transportation progress for each order, including the origin and destination, the transparency of the transportation process is increased. By matching historical routes and extracting historical transportation task schedules, historical data can be used to optimize the current transportation route and scheduling plan, thereby reducing transportation time and cost. By classifying the transportation progress status according to historical data and current events, potential transportation interruptions or delays can be discovered in time, so that measures can be taken in advance to reduce risks. By traversing and classifying the status of a subset of received orders, changes in the transportation process can be responded to quickly, and orders that require special attention (such as transportation interruptions or orders to be transported) can be processed quickly.

[0064] Preferably, step S256 includes the following steps:

[0065] Step S2561: if the current transport event is an empty set, the corresponding received order is marked as a waiting-for-transportation state, thereby obtaining an order state classification identifier, otherwise, proceed to the next step;

[0066] Specifically, you can connect to the database and execute a SQL query, such as: SELECT event_id,description FROM events WHERE order_id = [specific order ID] AND is_current = 'true'. The query obtains the description of the current shipping event for a specific order ID. After the query is executed, the software will check the returned result set. If the result set is empty, that is, no current shipping event is found, the software will mark the corresponding order as waiting for shipping based on this information. The software will update the status of the order in the order database and change its status field to "waiting for shipping". This can be done by executing another SQL update statement, such as: UPDATE orders SET status = 'waiting for shipping' WHERE order_id = [specific order ID]. Once an order is marked as waiting for shipping, the software will generate an order status classification identifier, for example, it can be a text label "waiting for shipping", including the order ID and its new status. For example, if the current shipping event query result for order 100009 is empty, the transportation management software will update its status to "waiting for shipping" and create a status classification identifier, such as: "Order 100009 - Waiting for Shipping".

[0067] Step S2562: for each historical transportation task schedule, extract the transportation events in the historical transportation task schedule that are consistent with the transportation event description in the current transportation event, and obtain a plurality of matching transportation events;

[0068] Specifically, a data analysis tool capable of processing and comparing large amounts of data can be used to extract transport events that match the current transport event description from the historical transport task schedule. The specific operation is as follows: the data analysis tool will access the database that stores the historical transport task progress and execute a query to obtain the historical transport event records for a specific order, for example: SELECT event_id, description FROM shipment_progress WHERE order_id = [specific order ID]. Then, obtain the description of the current transport event, which can be obtained from step S2561. Next, compare the historical transport event description with the current transport event description. This is achieved through a string matching algorithm, in which the tool checks whether each historical event description contains keywords or phrases of the current event description. For each matching event, the tool will extract relevant information, such as event ID and description, and store it in a new data structure, such as a list or dictionary. Ultimately, the tool will generate a set of matching transport events that are consistent with the current transport event in description. For example, if the description of the current transport event is "loading cargo", and there are three events in the historical transport task schedule whose descriptions also contain "loading cargo", extract these events and create a set of matching transport events, such as: [{'event_id': 1, 'description': 'Loading cargo at Shanghai warehouse'}, {'event_id': 5, 'description': 'Loading cargo at Beijing warehouse'},{'event_id': 9, 'description': 'Loading cargo at Guangzhou warehouse'}].

[0069] Step S2563: Determine the transportation time length for each matching transportation event based on the historical transportation task schedule to obtain a number of historical transportation durations;

[0070] Specifically, a SQL query may be executed to obtain the start time and end time of each matching transport event. For each returned event record, the transport duration is calculated, that is, the difference between the end time and the start time. For example, if the start time is "2024-09-01 08:00" and the end time is "2024-09-01 12:00", the transportation time is 4 hours. The transportation time of each matching event is recorded and stored in a new data structure, such as a list or dictionary, where each event ID corresponds to a transportation time value. This process will be repeated for all matching transportation events, and finally a set of historical transportation time data will be obtained. For example, if there are three matching transportation events, their start time and end time are: event 1 ("2024-09-01 08:00", "2024-09-01 12:00"), event 2 ("2024-09-02 09:00", "2024-09-02 13:30"), event 3 ("2024-09-03 10:00", "2024-09-03 15:45"), the logistics analysis tool will calculate that the transportation times for these three events are 4 hours, 4.5 hours, and 5.75 hours respectively.

[0071] Step S2564: Screen the historical transportation time for the maximum duration to obtain the historical maximum transportation time, add the preset safety time to the historical maximum transportation time, and obtain the transportation event warning time;

[0072] Specifically, the historical transportation time list obtained in step S2563 can be loaded, and a statistical analysis method can be used to find the maximum time. This involves finding the maximum value in the list. For example, if the historical transportation time list is [4, 4.5, 5.75, 5.5], the maximum value is 5.75 hours. Next, the software will add a safety time according to the predefined safety standard. For example, if the preset safety time is 1 hour, the software will add the maximum historical transportation time to this safety time to obtain the warning time of the transportation event. This warning time will be used as a benchmark for evaluating whether the current transportation event is delayed. For example, if the maximum historical transportation time is 5.75 hours and the preset safety time is 1 hour, the warning time of the transportation event is 6.75 hours.

[0073] Step S2565: obtaining the current system time, and calculating the time spent on the current transportation event according to the current system time and the transportation event start time in the current transportation event, to obtain the current transportation time spent;

[0074] Specifically, the time spent calculation can be performed using a real-time shipping monitoring system that has access to the order database and obtains the current date and time information. The specific operation is as follows: the shipping monitoring system will first obtain the current system time through its built-in date and time functions. This can be achieved through the API or function provided by the system. For example, in many programming environments, the datetime.now() function can be used to obtain the current date and time. Next, the system will query the current shipping event table in the order database to obtain the start time of the shipping event for the specific order. This can be done by executing a SQL query, such as: SELECT start_time FROM events WHERE event_id = [current event ID] AND is_current = 'true'. Once the start time is obtained, the system will calculate the time spent in shipping. This usually involves comparing the current system time with the event start time and calculating the difference between them. For example, if the event start time is "2024-10-15 08:00" and the current system time is "2024-10-15 11:30", the time spent in shipping is 3 hours and 30 minutes. The calculated elapsed transportation time will be recorded in the system. For example, if the current system time is "2024-10-15 15:45" and the start time of a transportation event is "2024-10-15 08:00", the transportation monitoring system will calculate the elapsed transportation time of the event as 7 hours and 45 minutes.

[0075] Step S2566: If the current transport time is less than or equal to the transport event warning time, the corresponding received order is marked as being in transport status, and the order status classification identifier is obtained;

[0076] Specifically, the spent shipping time calculated in step S2565 and the shipping event warning time determined in step S2564 can be loaded. The two time values: the spent shipping time and the warning time. If the spent shipping time is less than or equal to the warning time, this indicates that the shipping event has not exceeded the expected maximum shipping time. In this case, the tool will update the status of the order to "in transit". This usually involves executing a SQL update statement to change the status field in the order database, such as: UPDATE orders SET status = 'in transit' WHERE order_id= [specific order ID]. After the update, an order status classification identifier will be generated to reflect the new status of the order. This identifier can be a simple text label, such as "in transit", or a more complex data structure containing the order ID and its new status information. For example, if the spent shipping time for an order is 7 hours and the warning time is 8 hours, the order status evaluation tool will determine that the order is still within the normal shipping time. Therefore, the order is marked as 'in transit' and a status classification identifier is generated, such as: Order 100010 - In transit.

[0077] Step S2567: If the current transport time is longer than the transport event warning time, the corresponding received order is marked as a transport interruption status, thereby obtaining an order status classification identifier.

[0078] Specifically, the current spent transport time calculated in step S2565 and the transport event warning time determined in step S2564 can be loaded. The two time values: the spent transport time and the warning time are compared. If the spent transport time exceeds the warning time, this indicates that the transport event has exceeded the expected maximum transport time. In this case, the status of the order will be updated to "transport interruption". This involves executing an SQL update statement to change the status field in the order database, for example: UPDATE orders SET status = 'Transportation interruption' WHERE order_id = [specific order ID]. After the update, the system will generate an order status classification identifier to reflect the new status of the order. This identifier can be a simple text label, such as "transportation interruption", or a more complex data structure containing the order ID and its new status information. For example, if the spent transport time of an order is 10 hours and the warning time is 8 hours, the order monitoring system will determine that the transportation of the order has exceeded the normal time range. Therefore, the order is marked as "transportation interruption" status and a status classification identifier is generated, such as: Order 100011 - Transportation Interruption.

[0079] The present invention can more accurately identify the initial state of an order by distinguishing between empty sets and non-empty sets of current transport events, thereby providing an accurate starting point for subsequent transport status classification. By matching the historical transport task schedule with the current transport event, the historical data can be used to predict and evaluate the transport time of the current transport task, thereby enhancing the understanding and control of the transport process. By determining the maximum duration of historical transport and adding it to the preset safety duration, a reasonable warning duration of a transport event can be set, which helps to identify and prevent potential transport delays or interruptions in advance. By obtaining the current system time in real time and comparing it with the start time of the transport event, the spent transport time is calculated, thereby achieving real-time monitoring and evaluation of the transport progress. By comparing the spent transport time with the warning duration, the status of the order can be dynamically adjusted, and the order can be marked as being in transport or in a transport interruption state, and this dynamic adaptability helps to quickly respond to changes in the transport process. By timely identifying transport interruptions and updating the status, transport delays and cargo losses caused by failure to discover problems in time can be reduced, thereby improving the efficiency and reliability of transportation. By analyzing the historical transport duration and setting the warning duration, future transport events can be predicted, providing support for advance planning and scheduling.

[0080] Preferably, step S3 comprises the following steps:

[0081] Step S31: Real-time vehicle information of the transport vehicle is collected through the freight platform to obtain a transport vehicle data set, wherein the transport vehicle data set includes a plurality of transport vehicle data, each transport vehicle data at least includes a vehicle ID, a maximum loading weight, a maximum loading space, a vehicle functional attribute, a vehicle working state, a current loading state and a list of orders received by the vehicle, and the vehicle working state is any one of a working state and an idle state;

[0082] Specifically, the information of transport vehicles can be collected in real time through an integrated vehicle management system that can communicate with the vehicle's GPS tracking device, sensors, and other data sources. The specific operation is as follows: the vehicle management system will automatically send requests to the GPS tracking devices and sensors of each vehicle to collect the vehicle's location, status, and other relevant information. The system will receive real-time data streams from the vehicle, including vehicle ID, maximum load weight, maximum load space, vehicle functional attributes (such as refrigeration, dangerous goods transportation, etc.), vehicle working status (working or idle), and current loading situation. For each vehicle, the system will also query its received order list to determine whether the vehicle has an assigned order. The collected data will be organized and stored in the transport vehicle profile data set, and the profile data of each vehicle is a record containing all necessary information. For example, assume that the vehicle management system collects the following information: the vehicle with vehicle ID VH123, the maximum load weight is 10 tons, the maximum load space is 50 cubic meters, has refrigeration function, the current working status is idle, the current loading situation is empty, and the received order list shows that all orders have been completed. This information will be recorded in the transport vehicle profile data set.

[0083] Step S32: traverse the transport vehicle data set, and for each transport vehicle data set, if the vehicle working state is idle and all the received orders in the received order list are in the completed transportation state, mark the corresponding transport vehicle as an available vehicle to obtain a first available vehicle set;

[0084] Specifically, the transport vehicle profile data set can be loaded and the records of each vehicle can be traversed. For each record, the software will check whether the working status of the vehicle is marked as "idle". If the vehicle status is "idle", the software will further check the list of received orders for the vehicle to confirm whether all orders have been marked as "completed transportation". If all conditions are met, that is, the vehicle status is idle and there are no unfinished orders, the vehicle will be marked as an available vehicle. All vehicles marked as available will be added to the first available vehicle set. For example, suppose that when the fleet management software processes the record with vehicle ID VH123, it finds that the vehicle status is idle and all orders in the list of received orders have been completed. Therefore, VH123 is marked as an available vehicle and added to the first available vehicle set. This process will be repeated for all vehicles in the transport vehicle profile data set to ensure that all available vehicles can be quickly identified and utilized.

[0085] Step S33: if the vehicle working state is working state, the completed orders are removed from the corresponding vehicle received order list to obtain an uncompleted order list;

[0086] Specifically, vehicles whose working status is marked as "working status" can be filtered out from the transport vehicle profile data set. For these vehicles, the software will check their received order lists and identify which orders have been transported. Using the order management system, the software will update the status of each completed order and mark it as "completed transportation". Once the order status is updated, the software will remove these completed orders from the vehicle's received order list to obtain an uncompleted order list. The updated uncompleted order list will be saved and associated with the corresponding vehicle profile. For example, suppose that a vehicle with vehicle ID VH456 currently has five received orders, three of which have been completed. The fleet management software will identify these five orders and update the status of the three completed orders to "completed transportation" through the order management system. Then, these three orders are removed from VH456's received order list to obtain a new list containing two uncompleted orders.

[0087] Step S34: extracting order cargo information from the uncompleted order list to obtain a transport cargo information set, wherein the transport cargo information set includes a plurality of transport cargo information, each transport cargo information includes at least a corresponding order ID, cargo weight and cargo volume, and the transport cargo information set corresponds to a transport vehicle;

[0088] Specifically, a list of uncompleted orders for each vehicle can be loaded, and data extraction operations can be performed for each order. For each order, the software will query the order database to obtain detailed information about the goods, including the order ID, the weight of the goods, and the volume of the goods. This information is collected and a transport cargo information record is created for each order, which includes the order ID, the weight of the goods, and the volume of the goods. All collected transport cargo information will be organized into a transport cargo information set, which corresponds to a specific transport vehicle. For example, suppose that the vehicle ID is VH456 and there are two uncompleted orders. The cargo weight of order 100A is 3 tons and the volume is 5 cubic meters; the cargo weight of order 200B is 2 tons and the volume is 4 cubic meters. The order processing software will extract this information from the order database and create the following transport cargo information set: [{'order_id': '100A', 'weight': 3000, 'volume': 5}, {'order_id': '200B', 'weight': 2000, 'volume': 4}]. This information set will be used to determine VH456's loading capacity and transportation requirements.

[0089] Step S35: Calculate the total weight of the transported goods and the total volume of the transported goods to obtain the total weight and total volume of the transported goods;

[0090] Specifically, a transport cargo information set can be loaded, which is a list of cargo information collected for each vehicle in step S34. The transport cargo information set of each vehicle is traversed, and the cargo weight and volume of each order are accumulated. For each vehicle, the cumulative weight and volume of all order cargo are calculated to obtain the total weight and total volume of the transport cargo of the vehicle. The calculation results will be recorded in the vehicle's profile data or stored in a new data structure. For example, assume that the transport cargo information set of vehicle ID VH789 contains cargo information for three orders: order 100A (weight 3 tons, volume 5 cubic meters), order 200B (weight 2 tons, volume 4 cubic meters), and order 300C (weight 1.5 tons, volume 3 cubic meters). The cargo loading optimization software will accumulate these values ​​and calculate that the total weight of the transport cargo of VH789 is 6.5 tons and the total volume is 12 cubic meters.

[0091] Step S36: comparing the total weight of the transported goods with the corresponding maximum loading weight, and comparing the total volume of the transported goods with the corresponding maximum loading space; if the total weight of the transported goods is greater than or equal to the corresponding maximum loading weight or the total volume of the transported goods is greater than or equal to the corresponding maximum loading space, marking the corresponding transport vehicle as a fully loaded vehicle to obtain a fully loaded vehicle set; otherwise, marking the corresponding transport vehicle as an available vehicle to obtain a second available vehicle set;

[0092] Specifically, the total weight and total volume data of the transported goods of each vehicle, as well as the maximum loading weight and maximum loading space data of the vehicle, can be loaded. Compare the total weight of the transported goods with the maximum loading weight of the vehicle, and the total volume of the transported goods with the maximum loading space of the vehicle. If the total weight of the transported goods is greater than or equal to the maximum loading weight of the vehicle, or the total volume of the transported goods is greater than or equal to the maximum loading space of the vehicle, the software will mark the vehicle as a fully loaded vehicle. All vehicles marked as fully loaded will be added to the fully loaded vehicle set. If the total weight and total volume of the transported goods do not exceed the maximum limit of the vehicle, the software will mark the vehicle as an available vehicle and add it to the second available vehicle set. For example, assuming that the maximum loading weight of the vehicle ID VH789 is 8 tons and the maximum loading space is 15 cubic meters. According to the calculation results of step S35, the total weight of the transported goods of VH789 is 6.5 tons and the total volume is 12 cubic meters, neither of which exceeds the maximum limit. Therefore, the cargo loading optimization software marks VH789 as an available vehicle and adds it to the second available vehicle set. If another vehicle with ID VH999 transports goods with a total weight of 10 tons, which exceeds its maximum loading weight of 8 tons, the software will mark VH999 as a fully loaded vehicle and add it to the fully loaded vehicle set.

[0093] Step S37: Record the first available vehicle set and the second available vehicle set as the vehicle data set to be allocated.

[0094] Specifically, the first available vehicle set can be accessed first, which contains all vehicles that are idle and have no uncompleted orders. Then, the second available vehicle set is accessed, which contains all vehicles that are not currently fully loaded, that is, those that have remaining loading capacity. Use data processing functions, such as database merge queries or set merge operations in programming languages, to integrate the two sets. During the integration process, the software will ensure that no vehicles are double-counted and that the information of each vehicle is up to date. The integrated data set, i.e., the vehicle data set to be allocated, will contain detailed information of all available vehicles, including vehicle ID, maximum load weight, maximum load space, current loading status, etc. For example, assume that the first available vehicle set contains vehicles with vehicle IDs VH123 and VH456, and the second available vehicle set contains vehicles with vehicle IDs VH789 and VH999. The fleet management software merges the information of these vehicles to form a vehicle data set to be allocated.

[0095] By collecting vehicle information in real time, the system can accurately grasp the current status and capacity of each vehicle, so as to more effectively allocate transportation tasks, reduce vehicle idle time, and improve vehicle utilization. By distinguishing between idle vehicles and vehicles in working state, and further analyzing the unfinished orders of vehicles in working state, this helps to dispatch vehicles more accurately and ensure that each vehicle can perform the right task at the right time. By analyzing the list of received orders of the vehicle, the completed orders can be eliminated and the unfinished orders can be focused on, which helps to more accurately evaluate the loading requirements of the vehicle and optimize the loading strategy. By calculating the total weight and volume of the goods and comparing them with the maximum loading limit of the vehicle, it can be ensured that the carrying capacity of the vehicle will not be exceeded, thereby improving the safety of cargo transportation. By accurately matching the loading capacity of the vehicle and the demand for goods, it can reduce the waste of resources caused by overloading or empty loading and improve transportation efficiency. By identifying fully loaded vehicles and available vehicles, this helps to quickly respond to transportation needs, ensure that the goods can be loaded and transported in time, and improve overall transportation efficiency.

[0096] Preferably, step S4 comprises the following steps:

[0097] Step S41: constructing a multi-objective optimization function for the to-be-allocated vehicle dataset, the transportation interruption order subset, and the unaccepted order subset to obtain a multi-dimensional optimization objective function set; wherein the multi-dimensional optimization objective function set includes a total transportation time minimization function, a vehicle utilization maximization function, and a transportation cost minimization function;

[0098] Specifically, a multi-objective optimization software called "FleetOptimizer" can be used to build and solve the optimization model. This software is specially designed to handle complex logistics and transportation optimization problems, and can consider multiple objectives and constraints at the same time. First, the dataset of vehicles to be assigned, the subset of transportation interruption orders, and the subset of unaccepted orders are imported into the FleetOptimizer software. These datasets contain detailed information about the vehicles, the requirements of the orders, and the characteristics of the transportation tasks. Next, three optimization objectives are defined: Minimize the total transportation time: Use the "travel time + loading and unloading time" model to calculate the total transportation time for each order. FleetOptimizer software has a built-in time calculation module that can estimate the travel time based on the average speed of the vehicle, road conditions, and traffic forecasts, while taking into account the time required for loading and unloading. Maximize vehicle utilization: Use FleetOptimizer's utilization analysis tool to calculate the ratio of the loading weight and volume of each vehicle to the maximum capacity. The software will analyze the frequency of vehicle use and the current loading situation to ensure that the vehicle utilization rate is maximized. Minimize transportation costs: Use FleetOptimizer's cost analysis module to estimate the cost of each transportation task. This module takes into account fuel consumption, driver wages, road usage charges and other related expenses to calculate the total cost of each transport.

[0099] After defining these objectives, FleetOptimizer will use a weighted summation approach to balance these objectives. This approach allows each objective to be assigned a weight that reflects its importance in the decision-making process. For example, a higher weight is assigned to the total transportation time minimization objective because it has the greatest impact on customer satisfaction. The software will then use a genetic algorithm or a particle swarm optimization algorithm to search for the optimal solution. These algorithms are several algorithms used in FleetOptimizer to solve complex optimization problems. They find the optimal or near-optimal solution to the problem by simulating natural selection or group behavior. Finally, FleetOptimizer will generate a set of optimal solutions that will constitute the set of multi-dimensional optimization objective functions. For example, suppose there is a vehicle data set to be assigned, including a vehicle with vehicle ID VH123, a maximum loading weight of 10 tons, a maximum loading space of 50 cubic meters, and currently empty. At the same time, there is a subset of transportation interruption orders, including order 100A, which needs to be transported from Shanghai to Beijing, with a cargo weight of 3 tons and a volume of 5 cubic meters. There is also a subset of unfilled orders, including order 200B, which needs to be transported from Guangzhou to Shanghai, with a cargo weight of 2 tons and a volume of 4 cubic meters. The optimization software will use this data to calculate the total transportation time, vehicle utilization, and transportation cost for each potential vehicle-order combination, and find the best combination to minimize the total transportation time, maximize vehicle utilization, and minimize transportation costs. Ultimately, the software will generate a set of multidimensional optimization objective functions containing these optimal combinations.

[0100] Step S42: defining hard constraints on the multidimensional optimization objective function set according to the to-be-allocated vehicle data set, the transportation interruption order subset, and the unaccepted order subset to obtain a hard constraint condition set, and defining soft constraints on the multidimensional optimization objective function set to obtain a soft constraint condition set;

[0101] Specifically, please refer to the sub-steps of step S42 for the detailed implementation process of this embodiment.

[0102] Step S43: integrating the hard constraint condition set, the soft constraint condition set and the multi-dimensional optimization objective function set by Lagrangian relaxation method to obtain a constraint relaxation scheduling function;

[0103] Specifically, a mathematical optimization technique called Lagrangian relaxation can be used to handle the hard constraint set, soft constraint set and multi-dimensional optimization objective function set. This can be achieved through professional mathematical modeling and optimization software, such as MATLAB or the SciPy library in Python. The specific operation is as follows: First, the hard constraint set (such as vehicle capacity limit, time window limit, etc.) and the soft constraint set (such as vehicle balance utilization, path continuity, etc.) are integrated into the multi-dimensional optimization objective function set. These constraints will be converted into mathematical expressions and become part of the optimization problem. Using MATLAB or Python, define the Lagrangian function, which combines the original optimization objective function with the Lagrangian multiplier to relax the hard constraints. In this way, the original constrained optimization problem is transformed into a more tractable unconstrained optimization problem. The value of the Lagrangian multiplier is adjusted by an iterative method to gradually approach the optimal solution of the original problem. This process involves solving a series of sub-problems, each of which is an unconstrained optimization problem under the current Lagrangian multiplier value. In this way, a constrained relaxation scheduling function is obtained, which can provide an optimized objective function value for scheduling decisions while considering the constraints.

[0104] Step S44: dynamic traffic data is collected on the real-time traffic network to obtain a real-time traffic data set, historical transportation records of the transport vehicles are collected to obtain a historical transportation data set, and order-vehicle-path relationships are captured on the real-time traffic data set and the historical transportation data set to obtain an order-vehicle-path association feature set;

[0105] Specifically, traffic information system integration tools, such as Google Maps API or similar real-time traffic data services, can be used to collect real-time traffic data. These data include road congestion, accident information, construction areas, etc., which will affect the vehicle's travel route and time. At the same time, historical transportation records are collected from the vehicle's GPS tracking system and transportation management system, which include the vehicle's past travel route, transportation time, cargo information, etc. Use data analysis software, such as R or Python's Pandas library, to merge the real-time traffic data set with the historical transportation data set and conduct in-depth analysis to identify the relationship between orders, vehicles, and paths. For example, analyze which routes are often congested during a specific time period and which vehicles are often used to transport specific types of goods. Through this analysis, an order-vehicle-path association feature set is obtained, which contains statistical information and patterns about the relationship between orders, vehicles, and paths. For example, suppose it is found that the route from Shanghai to Beijing is often congested during the morning rush hour, and historical data shows that vehicle VH123 often transports this route during this time period. Based on this information, VH123's dispatch plan can be adjusted to avoid scheduling transportation tasks during peak hours, or to choose alternative routes to reduce transportation time and costs.

[0106] Step S45: performing state space vectorization on the to-be-allocated vehicle data set, the transportation interruption order subset, and the unaccepted order subset to obtain an order-vehicle state vector set;

[0107] Specifically, data science tools such as Python and its machine learning library scikit-learn can be used to vectorize the state space of the assigned vehicle dataset, the transportation interruption order subset, and the unaccepted order subset. The specific operations are as follows: First, define the state features of each order and vehicle, such as the current location of the vehicle, the load status, the urgency of the order, the type of goods, etc. Then, use feature extraction techniques such as principal component analysis (PCA) or autoencoders to convert these state features into numerical vectors. These vectors will capture the current state of each order and vehicle and represent it as a point in the state space. For each order and vehicle, create a state vector that contains all the selected feature values. These state vectors will form the order-vehicle state vector set. For example, suppose there is a vehicle VH123, whose state features include current location (Shanghai), load status (empty), available capacity (10 tons), and special requirements (none). At the same time, there is an order 100A, whose state features include urgency (high), type of goods (fragile), and estimated delivery time (2024-10-15 08:00). Through feature extraction technology, these features are converted into numerical vectors, such as the state vector of VH123 is [1, 0, 10,0], and the state vector of order 100A is [1, 1, 1, 0].

[0108] Step S46: defining the action space for the to-be-allocated vehicle dataset, the transportation interruption order subset, and the unaccepted order subset to obtain an order-vehicle action space set;

[0109] Specifically, a decision tree or reinforcement learning algorithm can be used to define the set of order-vehicle action spaces. This can be implemented using Python with its machine learning libraries such as TensorFlow or PyTorch. The specific operation is as follows: First, define actions, such as assigning vehicles to orders, re-planning routes, postponing order processing, etc. Then, use a decision tree algorithm or reinforcement learning algorithm to define the potential action space for each order and vehicle. The decision tree algorithm will determine the best action based on the state vectors of the order and vehicle, while the reinforcement learning algorithm will learn the optimal action selection strategy through trial and error. For each order and vehicle, create an action space that contains all potential actions. These action spaces will form the set of order-vehicle action spaces. For example, suppose there is a vehicle VH123 and an order 100A. Potential actions include immediately assigning VH123 to order 100A, waiting for other orders, and assigning order 100A to other vehicles. Through decision trees or reinforcement learning algorithms, define the space of these actions and represent them as an element in the action space set, such as [assign VH123 to 100A, wait, assign other vehicles to 100A].

[0110] Step S47: Design a reward function for the constraint relaxation scheduling function to obtain a scheduling reward function, and perform Markov decision process modeling on the order-vehicle-path association feature set, the order-vehicle state vector set, the order-vehicle action space set and the scheduling reward function to obtain an order-vehicle scheduling decision model;

[0111] Specifically, please refer to the sub-steps of step S47 for the detailed implementation process of this embodiment.

[0112] Step S48: Perform a Monte Carlo tree search on the order-vehicle scheduling decision model to obtain an order-vehicle matching table, and update the order status of the freight platform based on the subset of orders being transported and the order-vehicle matching table.

[0113] Specifically, the Monte Carlo Tree Search (MCTS) algorithm can be used to optimize the order-vehicle scheduling decision model. MCTS is a heuristic search algorithm used in certain decision-making processes, especially suitable for complex decision-making problems with uncertainty. The specific operation is as follows: First, define the state space, action space, and transition probability of the order-vehicle scheduling decision model. The state space includes all potential order and vehicle combinations, the action space includes all potential scheduling decisions, and the transition probability describes the possibility of state transition under a specific decision. Next, use the MCTS algorithm to search for the optimal scheduling decision. The MCTS algorithm works through four steps: simulation, selection, expansion, and backpropagation. In the simulation step, we randomly generate a series of scheduling decisions and evaluate the results of each decision. In the selection step, we select those decisions with the highest expected benefits. In the expansion step, we expand the branches of the tree to consider more decision options. In the backpropagation step, we update the statistics of the nodes in the tree to reflect the new simulation results. The code of the MCTS algorithm can be written using the Python programming language and its machine learning libraries, such as numpy and scipy. Create a function to simulate the results of order-vehicle matching, and use this function to evaluate different scheduling decisions. After a series of MCTS iterations, an order-vehicle matching table is obtained, which contains the recommended order and vehicle matching solutions. Finally, the order status on the freight platform is updated according to this matching table and the subset of orders being transported. This can be achieved through the API of the freight platform or by directly updating the order status record in the database. For example, if order 200B is matched to vehicle VH123, the status of order 200B will be updated on the freight platform as "allocated to VH123", and the status of vehicle VH123 will be updated as "transporting order 200B". In this way, the MCTS algorithm helps find an optimal or near-optimal order-vehicle matching solution under uncertainty, and improves the efficiency and accuracy of scheduling by updating the order status on the freight platform in real time.

[0114] The present invention realizes a comprehensive consideration of total transportation time, vehicle utilization and transportation cost by constructing a multi-dimensional optimization objective function set, and significantly improves the efficiency and effect of scheduling. At the same time, the introduction of hard constraints and soft constraint condition sets increases the flexibility and adaptability of scheduling, ensuring that the system can adapt to changing transportation scenarios and customer needs while meeting the main optimization goals. The integrated application of Lagrangian relaxation method improves the scientificity and accuracy of decision-making, especially when dealing with scheduling problems under complex constraints. In addition, the system's prediction ability of transportation time and cost is enhanced by collecting and analyzing real-time road condition data and historical transportation data, optimizing path planning, and improving the timeliness and reliability of transportation. The application of state space vectorization and action space definition improves the intelligence level of scheduling, providing a solid foundation for the application of intelligent decision-making algorithms. The dynamic response capability of scheduling is further enhanced by the application of Markov decision process modeling, enabling the system to capture the relationship between orders, vehicles and paths in real time. The optimization accuracy of scheduling is improved by the design of reward functions and the application of Monte Carlo tree search, helping the system to comprehensively evaluate different scheduling schemes and find the optimal solution. The function of updating order status in real time improves the user experience of the system and increases customer trust and satisfaction. Finally, optimized scheduling plans help reduce empty runs, detours and delays, effectively reducing fuel consumption and labor costs, thereby reducing overall operating costs.

[0115] Preferably, step S42 includes the following steps:

[0116] Step S421: Capacity constraints are imposed on transport vehicles according to the vehicle data set to be allocated to obtain a vehicle load constraint matrix;

[0117] Specifically, professional logistics planning software can be used to perform capacity constraint analysis on the transport vehicles in the vehicle dataset to be allocated. The specific operation is as follows: The logistics planning software will load the vehicle dataset to be allocated, which contains detailed information about the vehicles, such as vehicle ID, maximum loading weight, and maximum loading space. The software will use a linear programming method to construct a vehicle load constraint matrix. This matrix will represent the relationship between the loading capacity of each vehicle and the order demand. For each vehicle, the software will define a matrix in which the rows represent vehicles, the columns represent orders, and the elements of the matrix represent the number of orders that the vehicle can load. The software will set constraints to ensure that the loading weight and volume of each vehicle do not exceed its maximum capacity. This will be achieved by adding linear inequalities to the planning model. After completing these steps, the software will generate a vehicle load constraint matrix that reflects the capacity restrictions of each vehicle. For example, suppose there are three vehicles, numbered VH1, VH2, and VH3, and their maximum loading weights are 5 tons, 8 tons, and 10 tons, respectively. There are four orders, orders A, B, C, and D, and their weights are 2 tons, 3 tons, 4 tons, and 1 ton, respectively. The logistics planning software will build a matrix that contains the capacity constraints of these vehicles and orders. The matrix is ​​shown in the attached figure. Figure 2 As shown in Figure 1, in this matrix, 1 means the vehicle can load an order, and 0 means the vehicle cannot load an order. For example, VH1 can load orders A and D, but cannot load orders B and C.

[0118] Step S422: performing time window constraints on the transportation interrupted order subset and the unaccepted order subset to obtain an order time window constraint set;

[0119] Specifically, the scheduling software can be used to perform time window constraint analysis on the subset of transportation interrupted orders and the subset of unaccepted orders. The specific operation is as follows: the scheduling software will load the subset of transportation interrupted orders and the subset of unaccepted orders, which contain detailed information of the orders, such as order ID, expected shipping time and expected arrival time. The software will define a time window for each order, which represents the time range in which the order can be accepted and processed. For each order, the software will set constraints to ensure that the shipping and arrival times of the order fall within the specified time window. The software will use time series analysis methods to determine the time window constraint set for the order. This involves calculating the latest shipping time and the earliest arrival time for each order. After completing these steps, the software will generate an order time window constraint set that reflects the time restrictions of each order. For example, suppose there are three orders, Order E, Order F, and Order G, whose expected delivery times are 2024-10-20, 2024-10-22, and 2024-10-25, and expected arrival times are 2024-11-05, 2024-11-10, and 2024-11-15, respectively. The scheduling software will define time windows for these orders and generate a time window constraint set, as shown in the attached figure. Figure 3 As shown, in this time window constraint set, each order has a clear shipping and arrival time range, and the scheduling software will ensure that the processing and transportation plan of the order meets these time constraints.

[0120] Step S423: performing cargo stacking attribute constraints on the cargo according to the transportation interruption order subset and the unaccepted order subset to obtain a cargo stacking rule matrix;

[0121] Specifically, logistics planning software can be used to perform constraint analysis on cargo stacking attributes. The specific operation is as follows: Using logistics planning software, first load the transportation interruption order subset and the unaccepted order subset, which contain detailed information of the order, including the size, weight and stacking attributes of the cargo. The logistics planning software will define the cargo stacking rule matrix based on the stacking attributes of the cargo. This matrix will indicate how different cargoes are stacked and how they are placed in the vehicle. The software will consider the physical properties of the cargo, such as whether it is fragile, whether it needs to avoid squeezing, stacking height restrictions, etc., to ensure that the cargo is safely stored in the vehicle. By analyzing the stacking attributes of the cargo and the loading space of the vehicle, the software will generate a cargo stacking rule matrix, which contains the constraints for cargo stacking. For example, suppose there are three types of cargo: A (fragile, cannot be stacked), B (can be stacked up to 3 meters high) and C (can be stacked up to 5 meters high). There are two vehicles, VH1 and VH2, whose loading spaces are 4 meters and 5 meters high respectively. The logistics planning software will analyze this information and generate a cargo stacking rule matrix, as shown in the attached figure. Figure 4As shown in the figure, in this matrix, the numbers represent the maximum stacking height of each cargo in the corresponding vehicle. For example, cargo B can be stacked up to 3 meters high in VH1 and 3 meters high in VH2.

[0122] Step S424: Record the vehicle load constraint matrix, the order time window constraint set and the cargo stacking rule matrix as a hard constraint condition set;

[0123] Specifically, the logistics planning software can be used to integrate the vehicle load constraint matrix, the order time window constraint set and the cargo stacking rule matrix into a hard constraint set. The specific operation is as follows: the logistics planning software will load the three matrices generated in steps S421, S422 and S423: the vehicle load constraint matrix, the order time window constraint set and the cargo stacking rule matrix. The software will integrate these matrices to form a comprehensive hard constraint set. This set will contain the loading restrictions of all vehicles, the time restrictions of orders and the stacking restrictions of goods. Use these hard constraints to guide the subsequent vehicle scheduling and order allocation process to ensure that all scheduling decisions meet these constraints. For example, suppose there is a hard constraint set that contains the loading restrictions of vehicles VH1 and VH2, the time restrictions of orders E and F, and the stacking restrictions of goods A, B and C. The logistics planning software will use this information to generate a scheduling plan that will ensure that the vehicle will not be overloaded, the order will be completed within the specified time, and the goods will be loaded according to the specified stacking rules. This hard constraint set will serve as the basis for scheduling decisions to ensure that all scheduling activities are carried out within the framework of these constraints.

[0124] Step S425: performing vehicle balance utilization constraints on the transport vehicles to obtain a vehicle load balance index matrix;

[0125] Specifically, the transportation management system can be used to perform vehicle balance utilization constraint analysis on transportation vehicles. The specific operation is as follows: the transportation management system will load the historical transportation data and current task allocation of all transportation vehicles, which include the loading records, transportation frequency and utilization efficiency of each vehicle. Statistical analysis methods, such as standard deviation or variance analysis, are used to evaluate the balance of vehicle utilization. This method can help identify which vehicles are often overused and which vehicles are underused. Based on these analyses, the system will construct a vehicle load balance indicator matrix. This matrix will contain the load balance index of each vehicle, reflecting the degree of balance between vehicle usage frequency and load volume. Set vehicle balance utilization constraints to ensure that the use of vehicles is not too concentrated on certain vehicles, thereby avoiding excessive wear and tear and improving the utilization efficiency of all vehicles. For example, suppose there are five vehicles, numbered VH101 to VH105, and their usage frequencies are 10 times, 5 times, 3 times, 8 times and 2 times respectively. The transportation management system will analyze these data and generate a vehicle load balance indicator matrix, as shown in the attached figure. Figure 5 As shown, in this matrix, the load balancing index is a value between 0 and 1. The closer the value is to 1, the closer the vehicle's usage frequency is to the average usage frequency of the fleet, thus reflecting that the vehicle's usage is more balanced.

[0126] Step S426: performing path continuity constraints on the transport vehicles according to the order-vehicle-path association feature set to obtain a path continuity index matrix;

[0127] Specifically, path analysis software can be used to perform path continuity constraint analysis on transport vehicles. The specific operations are as follows: the path analysis software will load the order-vehicle-path association feature set, which contains the order's origin, destination, and historical transportation path information. The software will use the shortest path algorithm in graph theory, such as the Dijkstra algorithm or the A* algorithm, to analyze the transportation path of each vehicle and evaluate the continuity of the path. Based on these analyses, the system will construct a path continuity indicator matrix. This matrix will contain the path continuity index of each vehicle, reflecting the path efficiency of the vehicle when performing continuous order tasks. Set path continuity constraints to ensure that the vehicle's transportation path is as continuous as possible, reduce empty driving and detours, and improve transportation efficiency. For example, suppose there are three vehicles, numbered VH201, VH202, and VH203, which are responsible for orders 201, 202, and 203, respectively. The path analysis software will analyze the transportation paths of these orders and generate a path continuity indicator matrix, as shown in the attached figure. Figure 6 As shown, in this matrix, 1 indicates that the vehicle can transport orders continuously, and 0 indicates that the vehicle needs to change the route to transport another order after transporting one order, thus reflecting the continuity of the route.

[0128] Step S427: performing customer priority constraints on customers according to the transportation interruption order subset and the unaccepted order subset to obtain a customer priority weight matrix;

[0129] Specifically, customer relationship management (CRM) can be used to load a subset of shipping interrupted orders and a subset of unaccepted orders, which contain detailed information about the orders, including customer ID, order details, and customer historical transaction data. The system will assign a priority score to each customer based on factors such as the customer's historical purchase volume, contract terms, credit rating, and historical service record. Using multi-criteria decision analysis methods such as AHP (Analytic Hierarchy Process) or TOPSIS (Technique for Order Preference by Similarity to Ideal Solution), the CRM system will determine the priority weight for each customer. Based on these analyses, the system will construct a customer priority weight matrix. This matrix will contain the priority weight of each customer, reflecting the customer's sensitivity to service response time and importance to the value of the enterprise. For example, suppose there are three customers, customers A, B, and C, whose historical purchase volumes are high, medium, and low, respectively, and whose credit ratings are excellent, good, and average, respectively. The CRM system will analyze this data and generate a customer priority weight matrix, as shown in the attached figure. Figure 7 As shown, in this matrix, the priority weight is a value between 0 and 1. The larger the value, the higher the priority of the customer, and the faster the service response is required.

[0130] Step S428: Record the vehicle load balancing index matrix, the path continuity index matrix and the customer priority weight matrix as a soft constraint condition set.

[0131] Specifically, the logistics planning software can be used to integrate the vehicle load balancing index matrix, the path continuity index matrix and the customer priority weight matrix into a soft constraint set. The specific operation is as follows: the logistics planning software will load the three matrices generated in steps S425, S426 and S427: the vehicle load balancing index matrix, the path continuity index matrix and the customer priority weight matrix. The software will integrate these matrices to form a comprehensive soft constraint set. This set will contain the load balancing index, path continuity index and customer priority weights of all vehicles. The logistics planning software will use these soft constraints to guide the subsequent vehicle scheduling and order allocation process, ensuring that all scheduling decisions try to meet these indicators and weights, but do not require full compliance. For example, suppose there is a soft constraint set that includes the load balancing index, path continuity index of vehicles VH101 to VH103, and the priority weights of customers A, B and C. The logistics planning software will use this information to generate a scheduling plan that will try to balance the use of vehicles, optimize path continuity, and give priority to the needs of high-priority customers.

[0132] The present invention ensures the compliance of vehicle loading, improves the timeliness of scheduling, and optimizes cargo management by introducing hard constraint condition sets such as vehicle load constraint matrix, order time window constraint set and cargo stacking rule matrix, thereby enhancing the safety and efficiency of transportation. At the same time, through soft constraint condition sets such as vehicle load balancing index matrix, path continuity index matrix and customer priority weight matrix, it is possible to flexibly balance vehicle utilization efficiency, path planning and customer needs while pursuing the optimal solution, further improving the adaptability and customer satisfaction of the scheduling solution. This method of comprehensively considering hard constraints and soft constraints not only improves the directness and timeliness of transportation, but also reduces operating costs, improves decision-making quality, and makes the entire freight scheduling process more efficient, flexible and customer-oriented.

[0133] Preferably, step S47 includes the following steps:

[0134] Step S471: Designing a reward function for the constraint relaxation scheduling function to obtain a scheduling reward function;

[0135] Specifically, professional operations research and optimization software, such as CPLEX or Gurobi, can be used to design a scheduling reward function to evaluate and optimize the performance of the constraint relaxation scheduling function. The specific operation is as follows: First, define the objectives of the scheduling reward function, which includes minimizing the total transportation time, maximizing vehicle utilization, and minimizing transportation costs. Then, use operations research principles to construct a reward function. This function will assign a numerical reward to different scheduling decisions, which reflects the contribution of the decision to the optimization goal. For example, if a scheduling decision significantly reduces the total transportation time, the scheduling reward function will give a higher positive reward. On the contrary, if a decision increases the transportation cost or reduces the vehicle utilization, the function will give a lower reward or penalty. The parameters of the reward function can be calibrated and adjusted through historical data and simulation results to ensure that the function can accurately reflect the actual effects of different scheduling decisions. Finally, a scheduling reward function is obtained, which can provide a quantified reward value for each scheduling decision. For example, suppose there are three scheduling decisions: decision 1 is to assign order O1 to vehicle V1, decision 2 is to assign order O2 to vehicle V2, and decision 3 is to assign order O3 to vehicle V3. The dispatch reward function gives these decisions rewards of +10, +5, and -3, respectively, based on their impact on total transportation time, vehicle utilization, and transportation cost.

[0136] Step S472: performing decision node attribute encapsulation on the order-vehicle-path association feature set, the order-vehicle state vector set, the order-vehicle action space set and the scheduling reward function to obtain an order-vehicle scheduling decision node set;

[0137] Specifically, for example, the decision support system will load the order-vehicle-route association feature set, which contains the association information between orders, vehicles and routes. The system will load the order-vehicle state vector set, which contains the current state of each order and vehicle, such as location, load and task progress. The system will load the order-vehicle action space set, which contains the potential actions of each order and vehicle, such as allocation, waiting and replanning. The system will load the scheduling reward function, which provides a reward value for each scheduling decision. Then, the system encapsulates this information into a decision node attribute, and each decision node represents a specific scheduling decision under a specific state. The system will generate an order-vehicle scheduling decision node set, which contains all potential decision nodes, each of which contains order, vehicle, route, state, action and reward information. For example, suppose there are three orders (O1, O2, O3) and three vehicles (V1, V2, V3), as well as the association features and state vectors between them. The decision support system will generate a decision node for each potential order-vehicle pair (such as O1-V1, O2-V2, O3-V3) and assign a reward value based on the scheduling reward function to each node. These decision nodes will form a set.

[0138] Step S473: constructing a multi-level decision tree for the order-vehicle scheduling decision node set to obtain an order-vehicle multi-level decision tree;

[0139] Specifically, a decision tree building algorithm can be used to create a multi-level decision tree for orders and vehicles. This can be achieved using a machine learning library such as Python's scikit-learn. The specific operation is as follows: First, define the construction rules of the decision tree, which is based on criteria such as information gain, Gini impurity, or error minimization. Use the decision tree algorithm in the scikit-learn library, such as DecisionTreeClassifier or DecisionTreeRegressor, to select features and labels from the set of order-vehicle scheduling decision nodes. Features include the urgency of the order, the current status of the vehicle, the type of goods, etc., while the label is the output of the scheduling reward function. The algorithm will recursively split the data, each split representing a decision node, until the stopping condition is met, such as reaching the maximum depth or the minimum number of samples of the node. Through this process, a multi-level decision tree is built, each node represents a scheduling decision under specific conditions, and the hierarchical structure of the tree reflects the priority and dependency of the decision. For example, assuming there is a set of scheduling decision nodes for orders and vehicles, the decision tree algorithm will build a decision tree based on the features and reward values ​​of these nodes. The root node of the tree represents the set of all orders and vehicles, while the child nodes represent decisions based on specific features (such as vehicle availability or order priority). Each decision node contains conditions that further split the data until the leaf nodes represent the final dispatch decision.

[0140] Step S474: performing UCB heuristic node selection on the order-vehicle multi-level decision tree to obtain an order-vehicle optimal extended node set, and performing dynamic width tree expansion on the order-vehicle optimal extended node set to obtain an extended order-vehicle multi-level decision tree;

[0141] Specifically, the upper confidence bound (UCB) heuristic algorithm can be used to select the optimal set of expansion nodes in the order-vehicle multi-level decision tree using reinforcement learning libraries such as OpenAI Gym or DeepMind Lab in Python, and perform dynamic width tree expansion. First, define the parameters of the UCB algorithm, such as the exploration-exploitation balance factor, which will determine the trade-off between the algorithm exploring new nodes and exploiting known information. Using the UCB algorithm in the OpenAI Gym or DeepMind Lab library, the UCB score is calculated for each node in the multi-level decision tree, which is based on the reward value and the number of visits of the node. The algorithm will select the nodes with the highest UCB score as the optimal expansion nodes, which represent the most potential scheduling decisions. Then, dynamic width tree expansion is performed on these optimal expansion nodes, and the expansion width of the tree is adjusted according to the current scheduling environment and the objective function value. Finally, an expanded order-vehicle multi-level decision tree containing the optimal expansion nodes is obtained. For example, assuming that there are several candidate expansion nodes in the multi-level decision tree, the UCB algorithm will calculate their UCB scores based on the historical rewards and the number of visits of these nodes. For example, node N1 has the highest UCB score because of its high expected reward and low number of visits, so it is selected as the optimal expansion node. Then, the child nodes of node N1 are taken into consideration, and the expansion width of the tree is determined according to the current scheduling requirements and reward function value, thereby generating an expanded decision tree.

[0142] Step S475: performing a Monte Carlo fast scenario simulation on the extended order-vehicle multi-level decision tree to obtain a simulated transport scheduling simulation result, and performing value back propagation on the simulated transport scheduling simulation result to obtain an updated order-vehicle multi-level decision tree;

[0143] Specifically, Python's Monte Carlo Simulation library can be used for fast scenario simulation. First, the simulation parameters are defined, including the number of simulations and the distribution of random variables in each scenario (such as traffic time, cargo loading time, etc.). Using the Monte Carlo Simulation library, multiple random simulations are performed on each decision node. Each simulation randomly generates scenario parameters and calculates the corresponding transportation scheduling results. After each simulation, the result data of transportation scheduling is collected, including total transportation time, vehicle utilization, and transportation cost. Through multiple simulations, a set of simulated transportation scheduling simulation results are obtained, which reflect the performance of different scheduling decisions in different scenarios. Next, the value backpropagation method is used to update the node value in the decision tree according to the simulation results. This involves calculating the average reward or expected utility of each node and updating the evaluation value of the node. Finally, an updated order-vehicle multi-level decision tree is obtained, which contains information about the simulation results. For example, suppose that a decision tree with three decision nodes is simulated 1,000 times. Each simulation randomly generates traffic time and loading time, and calculates the corresponding transportation cost and time. Based on the simulation results, the value of each node in the decision tree is updated to reflect the expected performance of each scheduling decision.

[0144] Step S476: screening the scheduling strategy for the update order-vehicle multi-level decision tree to obtain the optimal scheduling strategy set;

[0145] Specifically, Python's Pandas and SciPy libraries can be used to screen scheduling strategies. First, define the screening criteria, including the objective function (such as minimizing the total transportation time), constraints (such as vehicle capacity restrictions), and risk preferences (such as conservative or aggressive scheduling strategies). Using the Pandas library, process and analyze the data in the updated decision tree, including the value of the nodes and the simulation results. According to the screening criteria, use the optimization algorithms in the SciPy library, such as linear programming or dynamic programming, to search for the optimal scheduling strategy. The algorithm will evaluate each potential scheduling strategy and select the optimal strategy that meets the objective function and constraints. Finally, a set of optimal scheduling strategies is obtained, which contains a set of optimal scheduling decisions that provide the best expected performance while meeting all requirements. For example, suppose there is an updated decision tree with five decision nodes, each with different simulation results and values. Using decision analysis software, an optimal strategy set is screened out based on the goal of minimizing the total transportation time and vehicle capacity restrictions. This policy set includes decisions such as assigning order O1 to vehicle V1 and order O2 to vehicle V2, which perform best in the simulation and meet all constraints.

[0146] Step S477: According to the optimal scheduling strategy set, the strategy execution path of the updated order-vehicle multi-level decision tree is traced back to the order-vehicle scheduling decision model.

[0147] Specifically, the order-vehicle multi-level decision tree data and the optimal scheduling strategy set can be loaded. By analyzing the optimal scheduling strategy set, the key decision points corresponding to each strategy are identified. For example, the strategy set contains a strategy, namely "prioritize high-value orders to the nearest vehicle". This strategy is converted into a search condition in the decision tree. Using the path backtracking algorithm, starting from the leaf node of the decision tree (i.e., the specific strategy execution result), backtrack to the root node to determine the decision path leading to these optimal strategies. The depth-first search (DFS) algorithm is used to explore all potential paths by writing recursive functions until a path that meets the optimal strategy is found. In the backtracking process, the key decision nodes on each path are recorded, including the attributes of the nodes (such as order priority, vehicle location, etc.) and the decisions made (such as order allocation, route selection, etc.). Using data analysis tools, such as Python's Pandas library, the execution paths of all optimal strategies are analyzed and integrated to extract consistent decision patterns and key decision points. This information is integrated into an order-vehicle scheduling decision model, which reflects the execution path of the optimal strategy set in a data-driven manner. For example, suppose that DispatchOptimizer finds in its analysis that the match between order P1 (high value and urgent) and vehicle VW123 (closest to order P1 and with available capacity) is part of the optimal strategy. The decision path for this match is determined through the path backtracking algorithm, including selecting the branch of "high priority orders first" in the second layer of the decision tree and selecting "proximity principle for vehicle allocation" in the subsequent nodes. Record this path and mark it as a recommended strategy in the dispatch decision model. This model can be directly referenced in actual operations.

[0148] The present invention enhances the incentive of scheduling decisions through the designed reward function, effectively guiding decision makers towards the goal of improving efficiency and reducing costs. At the same time, by encapsulating orders, vehicles, paths and reward functions as decision node attributes, various factors in the scheduling process can be fully considered. The multi-level decision tree constructed not only improves the hierarchy and clarity of decisions, but also enables complex scheduling problems to be processed in a hierarchical and step-by-step manner. The efficiency of node selection is optimized by adopting the UCB heuristic algorithm, and exploration and utilization are effectively balanced under uncertainty and incomplete information, thereby improving the efficiency and accuracy of node selection. The dynamic expansion capability of the decision tree is enhanced by dynamic width tree expansion, so that the decision tree can be adjusted according to real-time data and simulation results, thereby improving flexibility and adaptability. The efficiency and practicality of the simulation are improved by Monte Carlo fast scenario simulation, and a large number of transportation scenarios can be quickly generated and evaluated. The value feedback speed of the result is further improved by value back propagation, so that the decision process can respond to environmental changes and new information in a timely manner. The optimal solution is ensured to be selected from a large number of strategies by scheduling strategy screening, thereby improving the quality and implementation effect of the scheduling strategy. By tracing back the strategy execution path, the execution steps of the optimal strategy can be clearly determined to ensure that the scheduling strategy is effectively implemented.

[0149] Preferably, the present invention further provides a freight dispatching information management system for executing the freight dispatching information management method as described above, the freight dispatching information management system comprising:

[0150] The order collection module is used to collect unfinished orders on the freight platform to obtain an unfinished order set, wherein the unfinished order set contains a number of unfinished orders, and each unfinished order at least includes the order ID, order release time, target goods attribute data, publisher identity data, origin, destination, order status identification, order transportation task schedule, estimated delivery time and expected arrival time of goods;

[0151] An order classification module is used to classify and mark the order status of the unfinished order set to obtain a classified order set, wherein the classified order set includes a subset of orders in transit, a subset of orders interrupted in transit, a subset of orders to be transported, and a subset of orders not received;

[0152] The vehicle screening module is used to collect real-time vehicle information of transport vehicles through the freight platform to obtain a transport vehicle data set; the loading capacity of the transport vehicles is screened according to the transport vehicle data set to obtain a vehicle data set to be allocated;

[0153] The scheduling module is used to perform intelligent multi-dimensional parameter scheduling optimization on the subset of transportation interrupted orders and the subset of unaccepted orders based on the data set of vehicles to be allocated, obtain the order-vehicle matching table, and update the order status of the freight platform based on the subset of orders being transported and the order-vehicle matching table.

[0154] Preferably, a computer-readable storage medium stores a freight scheduling information management method as described above that can be loaded and executed by a processor.

[0155] Therefore, the embodiments should be regarded as illustrative and non-restrictive from all points, and the scope of the present invention is limited by the appended claims rather than the above description, and it is therefore intended that all changes falling within the meaning and range of equivalent elements of the application documents are included in the present invention.

[0156] The above description is only a specific embodiment of the present invention, so that those skilled in the art can understand or implement the present invention. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but should conform to the widest scope consistent with the principles and novel features invented herein.

Claims

1. A freight dispatch information management method, characterized in that: The following steps are involved: Step S1: Collect unfinished orders on the freight platform to obtain an unfinished order set, wherein the unfinished order set includes a number of unfinished orders, and each unfinished order includes at least an order ID, order release time, target goods attribute data, publisher identity data, origin, destination, order status identifier, order transportation task schedule, goods stacking attributes, estimated delivery time and expected arrival time of goods; Step S2: classifying and marking the order status of the unfinished order set to obtain a classified order set, wherein the classified order set includes a subset of orders in transit, a subset of orders interrupted in transit, a subset of orders to be transported, and a subset of orders not received; Step S3: collecting real-time vehicle information of transport vehicles through the freight platform to obtain a transport vehicle data set; screening the loading capacity of transport vehicles according to the transport vehicle data set to obtain a vehicle data set to be allocated; Step S4: intelligent multi-dimensional parameter scheduling optimization is performed on the transportation interruption order subset and the unaccepted order subset according to the vehicle data set to be allocated, and an order-vehicle matching table is obtained, and the order status of the freight platform is updated according to the transportation order subset and the order-vehicle matching table; wherein step S4 includes the following steps: Step S41: constructing a multi-objective optimization function for the to-be-allocated vehicle dataset, the transportation interruption order subset, and the unaccepted order subset to obtain a multi-dimensional optimization objective function set; wherein the multi-dimensional optimization objective function set includes a total transportation time minimization function, a vehicle utilization maximization function, and a transportation cost minimization function; Step S42: hard constraints are defined for the multidimensional optimization objective function set according to the to-be-allocated vehicle data set, the transportation interruption order subset, and the unaccepted order subset to obtain a hard constraint condition set, and soft constraints are defined for the multidimensional optimization objective function set to obtain a soft constraint condition set; wherein step S42 includes the following steps: Step S421: Capacity constraints are imposed on transport vehicles according to the vehicle data set to be allocated to obtain a vehicle load constraint matrix; Step S422: performing time window constraints on the transportation interrupted order subset and the unaccepted order subset to obtain an order time window constraint set; Step S423: performing cargo stacking attribute constraints on the cargo according to the transportation interruption order subset and the unaccepted order subset to obtain a cargo stacking rule matrix; Step S424: Record the vehicle load constraint matrix, the order time window constraint set and the cargo stacking rule matrix as a hard constraint condition set; Step S425: performing vehicle balance utilization constraints on the transport vehicles to obtain a vehicle load balance index matrix; Step S426: performing path continuity constraints on the transport vehicles according to the order-vehicle-path association feature set to obtain a path continuity index matrix; Step S427: performing customer priority constraints on customers according to the transportation interruption order subset and the unaccepted order subset to obtain a customer priority weight matrix; wherein step S47 includes the following steps: Step S471: Designing a reward function for the constraint relaxation scheduling function to obtain a scheduling reward function; Step S472: performing decision node attribute encapsulation on the order-vehicle-path association feature set, the order-vehicle state vector set, the order-vehicle action space set and the scheduling reward function to obtain an order-vehicle scheduling decision node set; Step S473: constructing a multi-level decision tree for the order-vehicle scheduling decision node set to obtain an order-vehicle multi-level decision tree; Step S474: performing UCB heuristic node selection on the order-vehicle multi-level decision tree to obtain an order-vehicle optimal extended node set, and performing dynamic width tree expansion on the order-vehicle optimal extended node set to obtain an extended order-vehicle multi-level decision tree; Step S475: performing a Monte Carlo fast scenario simulation on the extended order-vehicle multi-level decision tree to obtain a simulated transport scheduling simulation result, and performing value back propagation on the simulated transport scheduling simulation result to obtain an updated order-vehicle multi-level decision tree; Step S476: screening the scheduling strategy for the update order-vehicle multi-level decision tree to obtain the optimal scheduling strategy set; Step S477: backtracking the strategy execution path of the updated order-vehicle multi-level decision tree according to the optimal scheduling strategy set to the order-vehicle scheduling decision model; Step S428: Recording the vehicle load balancing index matrix, the path continuity index matrix and the customer priority weight matrix as a soft constraint condition set; Step S43: integrating the hard constraint condition set, the soft constraint condition set and the multi-dimensional optimization objective function set by Lagrangian relaxation method to obtain a constraint relaxation scheduling function; Step S44: dynamic traffic data is collected on the real-time traffic network to obtain a real-time traffic data set, historical transportation records of the transport vehicles are collected to obtain a historical transportation data set, and order-vehicle-path relationships are captured on the real-time traffic data set and the historical transportation data set to obtain an order-vehicle-path association feature set; Step S45: performing state space vectorization on the to-be-allocated vehicle data set, the transportation interruption order subset, and the unaccepted order subset to obtain an order-vehicle state vector set; Step S46: defining the action space for the to-be-allocated vehicle dataset, the transportation interruption order subset, and the unaccepted order subset to obtain an order-vehicle action space set; Step S47: Design a reward function for the constraint relaxation scheduling function to obtain a scheduling reward function, and perform Markov decision process modeling on the order-vehicle-path association feature set, the order-vehicle state vector set, the order-vehicle action space set and the scheduling reward function to obtain an order-vehicle scheduling decision model; Step S48: Perform a Monte Carlo tree search on the order-vehicle scheduling decision model to obtain an order-vehicle matching table, and update the order status of the freight platform based on the subset of orders being transported and the order-vehicle matching table.

2. The freight dispatch information management method according to claim 1, characterized in that: Step S2 includes the following steps: Step S21: extracting order status identifiers from the unfinished order set to obtain an order status identifier set, wherein the order status identifier set includes a plurality of order status identifiers, and the order status identifier is any one of a pending order and an accepted order; Step S22: if the order status is marked as pending order, the corresponding order is recorded as an unaccepted order, thereby obtaining an unaccepted order subset; Step S23: if the order status is marked as received, the corresponding order is recorded as an received order, thereby obtaining a subset of received orders; Step S24: extracting the transport task schedule for the received order subset to obtain a plurality of order transport task schedules, wherein each order transport task schedule includes at least one transport stage event; Step S25: classifying the transportation progress status of the received order subset according to the order transportation task schedule to obtain a subset of orders being transported, a subset of orders interrupted in transportation, and a subset of orders to be transported; Step S26: Integrate the subset of unaccepted orders, the subset of orders in transit, the subset of orders with interrupted transit, and the subset of orders to be transported to obtain a classified order set.

3. The freight dispatch information management method according to claim 2, characterized in that: Step S25 includes the following steps: Step S251: extracting the current transportation event from the order transportation task schedule to obtain the current transportation event, wherein the current transportation event includes the transportation event ID, transportation event description, transportation event start time and identity information of the transportation event person in charge; Step S252: acquiring the corresponding origin and the corresponding destination of the order transportation task schedule based on the received order subset, and taking the origin as the first origin and the destination as the first destination; Step S253: collecting completed orders from the freight platform to obtain a completed order set, wherein the completed order set includes a number of completed orders; Step S254: performing historical path matching on the completed order set according to the first origin and the first destination to obtain a matched completed order set, wherein the matched completed order set includes a plurality of matched completed orders; Step S255: extracting historical transport task schedules from the matched completed order set to obtain a number of historical transport task schedules; Step S256: classifying the transport progress status of the received orders according to the historical transport task schedule and the current transport event, and obtaining an order status classification identifier; Step S257: Traverse the subset of received orders, and for each received order, execute steps S251-S256, that is, obtain several order status classification identifiers, and classify and integrate the progress status of the subset of received orders according to the order status classification identifiers to obtain the subset of orders in transit, the subset of orders with interrupted transportation, and the subset of orders to be transported.

4. The freight dispatch information management method according to claim 3, characterized in that: Step S256 includes the following steps: Step S2561: if the current transport event is an empty set, the corresponding received order is marked as a waiting-for-transportation state, thereby obtaining an order state classification identifier, otherwise, proceed to the next step; Step S2562: for each historical transportation task schedule, extract the transportation events in the historical transportation task schedule that are consistent with the transportation event description in the current transportation event, and obtain a plurality of matching transportation events; Step S2563: Determine the transportation time length for each matching transportation event based on the historical transportation task schedule to obtain a number of historical transportation durations; Step S2564: Screen the historical transportation time for the maximum duration to obtain the historical maximum transportation time, add the preset safety time to the historical maximum transportation time, and obtain the transportation event warning time; Step S2565: obtaining the current system time, and calculating the time spent on the current transportation event according to the current system time and the transportation event start time in the current transportation event, to obtain the current transportation time spent; Step S2566: If the current transport time is less than or equal to the transport event warning time, the corresponding received order is marked as being in transport status, and the order status classification identifier is obtained; Step S2567: If the current transport time is longer than the transport event warning time, the corresponding received order is marked as a transport interruption status, thereby obtaining an order status classification identifier.

5. The freight dispatch information management method according to claim 1, characterized in that: Step S3 includes the following steps: Step S31: Real-time vehicle information of the transport vehicle is collected through the freight platform to obtain a transport vehicle data set, wherein the transport vehicle data set includes a plurality of transport vehicle data, each transport vehicle data at least includes a vehicle ID, a maximum loading weight, a maximum loading space, a vehicle functional attribute, a vehicle working state, a current loading state and a list of orders received by the vehicle, and the vehicle working state is any one of a working state and an idle state; Step S32: traverse the transport vehicle data set, and for each transport vehicle data set, if the vehicle working state is idle and all the received orders in the received order list are in the completed transportation state, mark the corresponding transport vehicle as an available vehicle to obtain a first available vehicle set; Step S33: if the vehicle working state is working state, the completed orders are removed from the corresponding vehicle received order list to obtain an uncompleted order list; Step S34: extracting order cargo information from the uncompleted order list to obtain a transport cargo information set, wherein the transport cargo information set includes a plurality of transport cargo information, each transport cargo information includes at least a corresponding order ID, cargo weight and cargo volume, and the transport cargo information set corresponds to a transport vehicle; Step S35: Calculate the total weight of the transported goods and the total volume of the transported goods to obtain the total weight and total volume of the transported goods; Step S36: comparing the total weight of the transported goods with the corresponding maximum loading weight, and comparing the total volume of the transported goods with the corresponding maximum loading space; if the total weight of the transported goods is greater than or equal to the corresponding maximum loading weight or the total volume of the transported goods is greater than or equal to the corresponding maximum loading space, marking the corresponding transport vehicle as a fully loaded vehicle to obtain a fully loaded vehicle set; otherwise, marking the corresponding transport vehicle as an available vehicle to obtain a second available vehicle set; Step S37: Record the first available vehicle set and the second available vehicle set as the vehicle data set to be allocated.

6. A freight dispatch information management system, characterized in that: Used to execute the freight dispatch information management method according to claim 1, the freight dispatch information management system comprises: The order collection module is used to collect unfinished orders on the freight platform to obtain an unfinished order set, wherein the unfinished order set contains a number of unfinished orders, and each unfinished order at least includes the order ID, order release time, target goods attribute data, publisher identity data, origin, destination, order status identification, order transportation task schedule, estimated delivery time and expected arrival time of goods; An order classification module is used to classify and mark the order status of the unfinished order set to obtain a classified order set, wherein the classified order set includes a subset of orders in transit, a subset of orders interrupted in transit, a subset of orders to be transported, and a subset of orders not received; The vehicle screening module is used to collect real-time vehicle information of transport vehicles through the freight platform to obtain a transport vehicle data set; the loading capacity of the transport vehicles is screened according to the transport vehicle data set to obtain a vehicle data set to be allocated; The scheduling module is used to perform intelligent multi-dimensional parameter scheduling optimization on the subset of transportation interrupted orders and the subset of unaccepted orders based on the data set of vehicles to be allocated, obtain the order-vehicle matching table, and update the order status of the freight platform based on the subset of orders being transported and the order-vehicle matching table.

7. A computer-readable storage medium storing a freight dispatch information management method as claimed in any one of claims 1 to 5 that can be loaded and executed by a processor.

Images