Automatic order scheduling method based on delivery address and emergency degree

Abstract

The invention relates to the technical field of data processing and logistics management, and discloses an automatic order scheduling method based on a delivery address and an emergency degree, and the method comprises the steps: carrying out the preprocessing of order data, achieving the address structuring, and quantifying the order emergency degree; a K-means algorithm is adopted to carry out geographical clustering on the orders to form delivery districts; through a two-dimensional sorting strategy, sorting is carried out according to the total emergency degree of the districts, and then secondary sorting is carried out on orders in the districts, so that global and local delivery priorities are determined; the optimal transport capacity vehicle model is intelligently matched for the ranked districts, and a dynamic adjustment mechanism for handling emergency situations such as emergency order insertion and vehicle faults is established; and finally, outputting structured data for system docking and a visual scheme for manual monitoring. According to the invention, automation and intellectualization of the whole scheduling process are realized, and the logistics scheduling efficiency and accuracy are significantly improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of data processing and logistics management, in particular to an order automatic scheduling method based on delivery address and emergency degree. BACKGROUND

[0002] In modern logistics and supply chain management, order scheduling and vehicle scheduling are the core links to determine the operation efficiency, cost control and customer satisfaction. With the rapid development of e-commerce and instant delivery business, the number of orders that logistics enterprises need to handle every day is growing explosively, and order information is becoming increasingly complex, including different delivery time limits, cargo specifications and geographical locations.

[0003] At present, the order scheduling work of many enterprises still largely depends on manual scheduling. The dispatcher manually or with the help of simple spreadsheet tools to group orders, plan routes and assign vehicles according to personal experience. The efficiency bottleneck of this traditional mode is obvious. When faced with a large number of orders, manual processing is not only slow and prone to errors, but the scheduling quality is highly dependent on the business proficiency and subjective judgment of the dispatcher, making it difficult to form a standardized and replicable efficient operation process. The scheduling quality is highly dependent on personal experience, and when making manual decisions, it is difficult to fully weigh multiple dynamic variables such as order urgency, geographical location, vehicle load, etc., resulting in non-optimal scheduling solutions and high empty running rates, uneven loading, etc., directly increasing operating costs.

[0004] At the same time, this mode is rigid and slow to respond to sudden situations such as urgent orders and vehicle breakdowns, seriously affecting service reliability. Therefore, the existing technology has been difficult to meet the urgent needs of modern logistics for efficient, intelligent and flexible scheduling. SUMMARY

[0005] In view of the shortcomings of the prior art, the present application provides an order automatic scheduling method based on delivery address and emergency degree, which solves the problems of low order scheduling efficiency and high dependence on manual experience in the prior art.

[0006] To achieve the above purpose, the present application is realized by the following technical scheme: an order automatic scheduling method based on delivery address and emergency degree, comprising the following steps:

[0007] Step 1: data preprocessing

[0008] The received order data containing delivery address, delivery time, cargo weight and volume is standardized.

[0009] In this step, first, the GeoHash encoding technology is used to convert the delivery address of the order. Specifically, the delivery address of each order is converted into a 32-bit binary string, which is more suitable for subsequent data processing and analysis. The delivery address of the order, including the address with ambiguous description (such as near the industrial park, logistics park), is parsed and converted into an 8-precision GeoHash code . Based on the coding rule, the orders with the same first 6 bits are determined to be located in the same geographic area.

[0010] Secondly, the urgency of the order is quantified, and according to the time difference between the current time and the delivery deadline of the order, the order Set a calculable urgency score, the calculation rule is as follows:

[0011] ;

[0012] Wherein:

[0013] Indicates an independent order;

[0014] Indicates the urgency score of the order ;

[0015] Indicates the remaining delivery time window of the order , which is calculated as ;

[0016] Indicates the delivery deadline of the order ;

[0017] Indicates the current time for scheduling calculation.

[0018] The calculation rule allows the operator to manually fine-tune the calculated urgency score according to the specific business needs.

[0019] Step two: address grouping

[0020] Based on the GeoHash code of the preprocessed order, the clustering algorithm is used to automatically cluster the orders with similar geographical positions into at least one area .

[0021] Specifically, first determine the number of clusters , which can be calculated according to the total weight of the orders to be scheduled and the average carrying capacity of the vehicles in the vehicle fleet:

[0022] ;

[0023] Wherein:

[0024] Indicates the number of areas finally generated;

[0025] Total weight of all orders in the current queue.

[0026] Average upper limit of load capacity of available vehicles in the fleet.

[0027] Ceiling function.

[0028] Subsequently, the top geographical locations with the highest order density are taken as initial cluster centers. Orders are assigned to the nearest center by iteratively calculating the geographical distance between orders and cluster centers, and constantly updating the center positions until convergence, eventually forming pieces of zones.

[0029] In this step, an overloaded zone splitting mechanism is also included. After clustering is completed, the total weight and total volume of each zone are calculated. If it meets the condition or , the overloaded zone is split into at least two sub-zones according to the geographical distance of orders in the zone, where:

[0030] Independent zone.

[0031] Total weight of all orders in the zone .

[0032] Sum of the volume of all orders in the zone .

[0033] Standard vehicle model rated load used to determine whether it is overloaded.

[0034] Standard vehicle model rated volume used to determine whether it is overloaded.

[0035] Step three: scheduling

[0036] The delivery sequence of the zones and orders in the zones generated in step two is determined.

[0037] First, the delivery priority between zones is determined. The total zone urgency score of each zone is obtained by calculating the sum of the urgency scores of all orders in the zone, and the calculation formula is:

[0038] ;

[0039] Where:

[0040] the slice emergency score of the slice ;

[0041] the slice emergency score of the slice ;

[0042] the slice emergency score of the slice ;

[0043] all slices are ranked in descending order according to their scores, and the slice with the highest score has the highest priority in delivery.

[0044] Secondly, the delivery order of orders within each slice is determined. For orders within a single slice, they are first sorted in descending order according to their scores; if there are orders with the same score, they are sorted in ascending order according to the geographical distance from their addresses to the delivery center.

[0045] Step four: vehicle matching

[0046] According to the total weight and total volume of goods in each slice, a transport vehicle is matched for the slice from a preset vehicle type library containing parameters such as vehicle number, upper limit of load, upper limit of volume, and state .

[0047] The matching process includes:

[0048] vehicles that are idle and meet the demand for transport capacity are screened out; further, from the vehicles that meet the conditions, vehicles that make the load utilization rate and the volume utilization rate both in the preset interval of 80%-90% are selected, and the calculation formulas of the load utilization rate and the volume utilization rate are as follows:

[0049] ;

[0050] ;

[0051] wherein:

[0052] represents an independent transport vehicle in the vehicle type library;

[0053] represents the load utilization rate of the vehicle when delivering the slice ;

[0054] represents the volume utilization rate of the vehicle when delivering the slice ;​

[0055] and total weight and total volume of goods in the slice area respectively;

[0056] represent the maximum load capacity of the vehicle ;

[0057] represent the maximum volume capacity of the vehicle .

[0058] Step five: schedule output

[0059] Based on the results of the above steps, a structured scheduling scheme is generated and output, which contains vehicle number, slice area name, detailed delivery list and order of slice area, expected departure time, etc.

[0060] Further, the method provided by the application further comprises a dynamic adjustment mechanism for handling emergency situations, which performs non-global rearrangement adjustment.

[0061] When receiving a new emergency order with an emergency score greater than or equal to a preset threshold, the dynamic adjustment mechanism is triggered, which first parses the GeoHash code of the new order to determine its belonging slice area, then inserts the order directly at the top of the order sorting list of the belonging slice area, and then recalculates the total weight and total volume of the slice area, and updates the vehicle matching and departure time of the slice area. This process only affects the single slice area, and the generated scheduling scheme of other slice areas remains unchanged.

[0062] When receiving a fault signal of a matched vehicle, the dynamic adjustment mechanism is triggered, which selects and assigns a new standby vehicle from the standby vehicle pool according to the preset scheduling priority rules to replace the slice delivery task originally undertaken by the fault vehicle, and adjusts the expected departure time of the slice area according to the current position of the standby vehicle. This adjustment is also limited to the slice area originally responsible by the fault vehicle, and does not affect the scheduling of other slice areas.

[0063] The application provides an order automatic scheduling method based on delivery address and emergency degree. It has the following beneficial effects:

[0064] 1、The application uses GeoHash encoding technology to accurately map the delivery address to geographic coordinates, providing an accurate basis for order clustering. On this basis, a two-dimensional sorting logic of slice area first and order second is created: macroscopically, the slice areas are sorted according to the total emergency degree, and microscopically, the orders are sorted according to the emergency degree and distance. This method effectively coordinates the two demands of emergency priority and smooth path, solves the conflict between the two in the prior art, and significantly reduces the vehicle empty running rate.

[0065] 2、The application sets up a screening rule based on the load utilization rate and the volume utilization rate in the 80%-90% interval in the vehicle model matching link, which further limits the utilization efficiency of the transport capacity resources under the basic premise of meeting the upper limit of the vehicle load and volume, ensuring the accurate adaptation of the selected vehicle to the total transport capacity of the area, which avoids the waste of large vehicles carrying small goods due to human judgment negligence, or the overloading and secondary shunting problems caused by insufficient transport capacity estimation.

[0066] 3、The application establishes a dynamic adjustment mechanism without global rearrangement, which only performs insertion order or vehicle replacement adjustment operation on the affected local area when facing emergency insertion order or vehicle failure, without the need to recalculate the overall scheduling scheme, which changes the response mode of interrupting and overturning the overall plan in the prior art, significantly improving the response speed and scheme stability of the scheduling system in a dynamic environment. BRIEF DESCRIPTION OF DRAWINGS

[0067] Figure 1 is the flow chart of the method of the application;

[0068] Figure 2 is a schematic diagram of the data preprocessing step of the application;

[0069] Figure 3 is a two-dimensional scheduling sorting logic flow chart of the application;

[0070] Figure 4 is a vehicle model matching screening logic flow chart of the application;

[0071] Figure 5 is an emergency insertion order dynamic adjustment flow chart of the application;

[0072] Figure 6 is a vehicle failure dynamic adjustment flow chart of the application. DETAILED DESCRIPTION

[0073] The technical solutions in the embodiments of the application will be described clearly and completely below with reference to the drawings in the embodiments of the application. Obviously, the described embodiments are only part of the embodiments of the application, not all. Based on the embodiments in the application, all other embodiments obtained by those skilled in the art without creative labor are within the scope of protection of the application.

[0074] Embodiment:

[0075] Please refer to the drawings of the application Figure 1 - the drawings of the application Figure 6The embodiment of the application provides an order automatic scheduling method based on a delivery address and an emergency degree, and the method comprises the following steps:

[0076] S100, data preprocessing, the purpose of the step is to convert received original order data with non-uniform formats into standardized and structured data which can be directly calculated and analyzed in subsequent steps, in the embodiment, the step can be specifically divided into three sub-steps of address resolution and coding, emergency quantification, and order validity verification.

[0077] Firstly, address resolution and coding are performed. The system receives a delivery address field in order data, and the field is a text string. In order to realize quantification of a geographical position, a GeoHash coding technology is adopted. For the address string, the system first calls a geographical information resolution service to attempt to directly convert the address string into a longitude and latitude coordinate. If resolution is successful and confidence is higher than a preset threshold, an 8-bit length GeoHash coding string is generated based on the longitude and latitude coordinate. An 8-bit length GeoHash code corresponds to a geographical grid with a physical size of about 19m*19m. The accuracy is sufficient to distinguish different building positions in the same street or large park.

[0078] For ambiguous addresses in the address string, such as nearby, surrounding and the like, or addresses that cannot be directly resolved due to incomplete expression, the system starts a secondary matching mechanism, and the mechanism performs keyword matching in a preset core geographical landmark database. The database stores standard names of key places such as local logistics parks, large factories and industrial parks and corresponding longitude and latitude coordinates, if a keyword in the order address is matched with a certain landmark point in the database, the system generates a GeoHash code for the order by using the longitude and latitude coordinates of the landmark point, and after processing, each order obtains a unique GeoHash code as a quantitative representation of the geographical position.

[0079] Secondly, order emergency quantification is performed. In order to objectively measure the delivery timeliness requirement of the order, the system converts the qualitative delivery time into a quantitative emergency score. The system obtains the current system time and reads the delivery deadline of the order, and obtains the remaining delivery time window of the order by calculating the time difference between the two. Then, the system maps the time difference into an emergency score according to a preset segmented function, and the specific mapping rule is as follows:

[0080] ;

[0081] Wherein:

[0082] represents an independent order;

[0083] representing the urgency score of the order ;

[0084] representing the remaining delivery time window of the order , which is calculated as ;

[0085] representing the delivery deadline of the order ;

[0086] representing the current time for schedule calculation.

[0087] In addition, the system provides an interface allowing dispatchers to manually modify the automatically calculated urgency score in specific cases (e.g. VIP customer orders), and subsequent steps will use the modified urgency score for calculation.

[0088] Finally, order validity checks are performed to ensure that the order data entering the schedule calculation is valid. The system automatically checks the weight and volume of each order, and has two preset thresholds: the maximum load upper limit and the maximum volume upper limit of the vehicle, which usually correspond to the carrying capacity of the largest vehicle in the fleet. The verification logic is as follows: if or , the order is marked as over-limit and removed from the current automatic scheduling pool for manual processing, if or , the order is marked as invalid data and also removed from the scheduling pool. Only orders that pass the validity check can enter the subsequent address grouping step. S200, address grouping, this step receives the preprocessed order data with GeoHash encoding in S100, its purpose is to aggregate discrete orders into several geographically concentrated areas based on the geographical location of the order using K-means clustering algorithm, to provide a basis for subsequent centralized distribution and path planning.

[0089] Before executing the clustering algorithm, the number of target clusters, i.e. the number of areas

[0090] , is first determined, which is calculated based on the total volume of all orders to be scheduled and the average carrying capacity of the fleet vehicles to ensure that the number of areas generated matches the available carrying capacity. The calculation formula is as follows:

[0091] ; ​

[0092] in:

[0093] Indicates the final number of regions generated;

[0094] This indicates the total weight of all goods in the current pending orders;

[0095] This indicates the average maximum load capacity of available vehicles in the fleet;

[0096] This represents the function for rounding up.

[0097] In determining After setting the value, it is necessary to select a value for the K-means algorithm. To ensure more reasonable clustering results, initial cluster centers are selected based on the geographical density of orders. The system divides the geographical area covered by the scheduling into equally sized grid cells and calculates the number of orders within each grid cell. The initial cluster centers are selected based on the geographical density of orders. The geographic center point of each grid cell was selected for the K-means algorithm. Initial cluster centers .

[0098] Subsequently, the algorithm enters an iterative process, which includes two steps: allocation and update. In the allocation step, for each order... The system calculates the geographic coordinates corresponding to its GeoHash encoding and compares them with all current... Cluster centers The spherical distance between them (calculated using the Haversine formula) is used to assign the order to the area represented by the nearest cluster center. During the update process, for each area The system recalculates its centroid by taking the average of the geographic coordinates of all orders within the region and updating the centroid location as the new cluster center for that region. .

[0099] The above allocation and update steps are repeated until a preset convergence condition is met. The convergence condition can be set to one of the following two:

[0100] In two consecutive iterations, the positional changes of all cluster centers were less than the preset distance threshold. ;

[0101] The number of iterations has reached the preset maximum number of iterations. .

[0102] After the clustering process converges, the system will analyze each generated region. Capacity check is performed to handle overloading situation. The system calculates the total weight and total volume of each zone . If the condition or is met, the zone is determined as an overloading zone, where and are the standard vehicle's rated load and volume respectively for the judgment. For the overloading zone, the system will start the splitting mechanism: find the two orders with the furthest distance in the zone as the initial center to split the zone into two sub-zones, and re-allocate the rest of the orders according to the nearest distance principle. The splitting process can be repeated until the capacity of all sub-zones is within the single vehicle's carrying range.

[0103] S300, scheduling sequence, this step receives the zone set and the order data contained in it generated in S200, and its purpose is to determine the global and unique delivery execution sequence. The sequencing process is divided into two levels: first determine the macro delivery priority between zones, and then determine the micro delivery sequence of orders within each zone.

[0104] The first level is the zone delivery priority sequence. The system traverses all the zones generated in step S200 . For each zone , the system calculates its zone urgency total score, which is the arithmetic sum of the urgency scores of all orders in the zone. The calculation formula is as follows:

[0105] ;

[0106] Where:

[0107] represents an independent zone;

[0108] represents the zone urgency total score of zone ;

[0109] represents the summation of each order in zone ;

[0110] is the urgency score of order calculated in step S100.

[0111] After calculating the zone urgency total score of all zones, the system sorts all zones in descending order according to their values. The result of the sorting is an ordered zone list , where The list defines the order of execution of the delivery tasks corresponding to different geographical zones.

[0112] The second level is to sort the orders within a zone. After the priority of the zones is determined, the system sorts the list of orders within each zone to determine the specific delivery route order of the vehicle within the zone. For any order set within a zone, the sorting follows a double standard.

[0113] First, the orders are sorted in descending order by the urgency score of the order, which ensures that orders with higher urgency scores are delivered first.

[0114] Second, when there are two or more orders with the same urgency score, a secondary sorting key is used, which is the straight-line distance from the delivery address of the order to the geographical coordinates of the distribution center. The system obtains the pre-set geographical coordinates of the distribution center and the geographical coordinates of the order, and calculates the distance between the two. For the subset of orders with the same urgency score, the system sorts them in ascending order by the value of the distance. This ensures that when the urgency is the same, the vehicle delivers the orders closer to the distribution center first.

[0115] After the above two levels of sorting, the final output of step S300 is a complete and hierarchical delivery sequence that not only specifies which zone the vehicle should go to first, but also specifies the order in which the orders should be delivered after arriving at the zone. This complete sorting result is passed as input to the subsequent S400 vehicle matching step.

[0116] S400, vehicle matching, the purpose of this step is to match a vehicle with the most suitable transport capacity for each zone that has been sorted in S300 from the dynamically updated vehicle pool. This process is performed for each zone in the order determined by the zone priority list in S300.

[0117] First, the system reads the information of all available vehicles from the pre-set vehicle pool, which stores multiple attributes of each vehicle, including vehicle number, maximum load limit, maximum volume limit, and current state.

[0118] The system first matches the vehicle for the highest priority zone in the order of the zone priority list The matching process includes multiple layers of filtering logic.

[0119] ​​​​​​​​First, the system selects all vehicles with idle status from the vehicle pool, and then further selects vehicles that can meet the basic demand of the area from the idle vehicle set, i.e. must meet the following two conditions:

[0120] ;

[0121] ;

[0122] After this step, the system gets a candidate vehicle set, if the set is empty, the system marks the area as insufficient capacity and continues to match the next area.

[0123] Second, the system calculates the load utilization rate and volume utilization rate of each vehicle in the candidate vehicle set when it is used to deliver the area The calculation formula is as follows:

[0124] ;

[0125] ;

[0126] Wherein:

[0127] represents a transport vehicle in the candidate vehicle set;

[0128] represents the current area to be matched;

[0129] represents the load utilization rate of the vehicle when it is used to deliver the area ;

[0130] represents the volume utilization rate of the vehicle when it is used to deliver the area ;

[0131] and are the total weight and total volume of the goods in the area ;

[0132] represents the maximum load upper limit of the vehicle ;

[0133] represents the maximum volume upper limit of the vehicle .

[0134] The system preferentially selects vehicles that meet the following conditions simultaneously:​

[0135] If multiple vehicles meet these two conditions, the vehicle with the largest maximum load capacity will be selected. The smallest vehicle is reserved for larger vehicles to carry heavier cargo.

[0136] If there are no vehicles in the candidate vehicle set that simultaneously meet both of the above utilization ranges, the system will execute a rollback strategy:

[0137] Select the maximum load capacity directly from the candidate vehicle set. The minimum vehicle allocation strategy ensures that even without an optimal match, a vehicle that meets basic capacity requirements and minimizes capacity waste can be assigned to a region.

[0138] After a successful match, the system will assign the area... With the selected vehicle Bind and connect the vehicle The vehicle's status in the vehicle database is updated to "assigned," preventing it from participating in matching for other areas in this scheduling calculation. Subsequently, the system continues with the next area in the list, following the same logic. Perform vehicle matching until all areas have been processed.

[0139] S500, Dynamic Adjustment Mechanism: This mechanism is an event-driven process used to modify the generated partial schedule without interrupting or recalculating the overall scheduling plan, in order to deal with unexpected situations that occur during transportation execution. This embodiment mainly describes two typical triggering events: the insertion of emergency orders and the failure of assigned vehicles.

[0140] The first scenario involves handling the insertion of urgent orders. Once a complete scheduling plan containing multiple regions and their corresponding vehicles has been generated, the system receives a new order. The system first performs data preprocessing in S100 on the new order and calculates its urgency score. If the score is greater than or equal to the preset urgency threshold For example, if the threshold is set to 9, this dynamic adjustment process will be triggered.

[0141] After the process is initiated, the system first obtains the GeoHash code of the new order. Then, the geographical coordinates of the order are calculated and compared with all existing areas. Geographical center The distance between them is used to assign the new order to the area closest to it, denoted as . .

[0142] Next, the system will process the new order. Insert directly into the target area The order will be placed at the top of the internal order sorting list. After this operation is completed, the system will recalculate. Total weight of goods and total volume .

[0143] The system verification was originally assigned to vehicles Whether it can still carry the updated total amount of goods, that is, whether it still meets the requirements. and If the conditions are met, only the order list and estimated total transit time for that area will be updated, and the adjustment will be complete. If the conditions are not met, the system will relocate the vehicles. The state is released back to idle, and only applies to the updated state. From all currently available vehicles (including those that have just been released) In the process, the vehicle matching process for S400 will be re-executed to match a new vehicle with sufficient capacity for the area. Throughout the adjustment process, all other unaffected areas and their scheduling plans will remain unchanged.

[0144] The second scenario involves handling faults in assigned vehicles. This occurs when the system receives information about a specific assigned vehicle from the vehicle monitoring terminal or through manual input. When a fault signal is received, the dynamic adjustment mechanism is triggered.

[0145] The system first identified the faulty vehicle. The area they are responsible for delivering to is denoted as Then, the system will The vehicle's status in the depot is updated to "under maintenance," thus removing it from any scheduling process.

[0146] Next, the system filters out all vehicles that meet the requirements from the set of vehicles in the vehicle database that are either idle or ready for use. The selection criteria for alternative vehicles to meet capacity requirements are as follows: and .

[0147] For all candidate vehicles, the system selects the optimal replacement vehicle according to preset scheduling rules. These rules can be set as follows: prioritize the candidate vehicle that is closest to the distribution center from the current geographical location; if multiple vehicles are at the same distance, select the one with the largest load capacity. The smallest one.

[0148] Selecting an alternative vehicle Then, the system compares it with the area. Bind it and update its status to assigned. Simultaneously, the system... current location and preparation time, recalculate and update the estimated departure time of the task in the zone, and the process is limited to modifying the schedule of the zone and does not affect the established plan of any other zone.

[0149] S600, schedule output, which is the terminal link of the entire automated scheduling process, and the purpose is to present and distribute the final results calculated and decided by the preceding steps S100 to S500 in one or more preset formats, forming a specific execution plan for dispatchers, drivers and other related information systems.

[0150] In an embodiment, the output of this step contains two main parts: structured data output and visual interface output.

[0151] First, structured data output is performed, and the system encapsulates each final delivery task into an independent data object in JSON format, which contains the following hierarchical fields to form a complete and machine-readable scheduling instruction:

[0152] The top-level object represents an independent delivery task.

[0153] The fields contained under this object are: vehicle number, license plate number, assigned driver information, and estimated departure time.

[0154] The object also contains a sub-object representing the zone information it is responsible for, which contains the following fields: zone ID, total number of orders in the zone, total weight of goods in the zone, and total volume of goods in the zone.

[0155] The object contains an array representing the delivery sequence list of orders in the zone, and each element in the array is an order object containing the following fields: delivery sequence number, order number, customer name, detailed delivery address, GeoHash code, goods weight, goods volume, and delivery deadline.

[0156] The system collects data objects of all delivery tasks and provides them to external systems (such as transportation management system TMS or warehouse management system WMS) through a preset API interface. External systems can call the interface to obtain complete scheduling data for subsequent warehouse picking, waybill printing, etc.

[0157] Second, visual interface output is performed, and the system renders structured data on the human-machine interface, which provides at least two view modes.

[0158] The first one is a list view, which displays all generated delivery tasks in a table form, each row represents a delivery task of a vehicle, and the column contents include vehicle number, order quantity, total weight, total square, estimated departure time and other summary information. Users can click on any row to expand and view the detailed list of all orders under the task and their scheduled delivery sequence.

[0159] The second one is a map view, which visualizes all delivery tasks on an electronic map. Each task corresponds to a path distinguished by different colors. The path is a polyline connecting the distribution center and the geographic location points of all orders under the task. The connection order of the polyline is the delivery sequence determined in S300. The order location points are displayed as markers on the map. Users can view the specific order information of the points through interactive operations.

[0160] In addition, the visualization interface provides manual intervention functions. Before confirming the final plan, dispatchers can fine-tune the system-generated scheduling. The executable operations include: in the list view, adjusting the order delivery sequence within a vehicle task through drag-and-drop operations; or in the map view, dragging an order marker from one path to another to realize cross-vehicle allocation. When users perform manual allocation operations, the system will immediately recalculate the total weight and total volume of the affected two areas and check whether the vehicle after allocation has overloading. If there is overloading, the system will prohibit the operation and issue a prompt to the user. All manual adjustments are recorded and updated to the structured data output after final confirmation.

Claims

1. A method for automatically scheduling orders based on delivery address and urgency, characterized in that, Includes the following steps: S1. Data preprocessing: Preprocessing the received order data, including using encoding technology to convert the order's delivery address into an encoding, and setting an urgency score based on the order's delivery time. S2. Address grouping: Based on the order code, total order quantity and cargo load limit, all orders are clustered into initial cluster centers. Each order is assigned to an initial cluster center for iteration until the position change of the cluster center after iteration is less than a preset value. The cluster centers that have been iterated are taken as cluster regions. S3. Scheduling and sorting: Based on the urgency score of the orders, the orders between the clustered areas are scheduled and sorted to determine the delivery priority between each clustered area and the order delivery order within each clustered area. S4. Vehicle type matching: Obtain and match transport vehicles for each region from the vehicle type database based on the total weight and volume of goods in each region. S5. Schedule Output: Generates and outputs the schedule plan.

2. The automatic order scheduling method based on delivery address and urgency according to claim 1, characterized in that, The encoding technology employed includes converting the order's delivery address into GeoHash encoding. The specific steps for converting the order's delivery address into GeoHash encoding are as follows: Delivery addresses, including fuzzy addresses, are converted into 8-bit GeoHash codes. Orders with the same first 6 bits of the code are considered to belong to the same geographic area.

3. The automatic order scheduling method based on delivery address and urgency according to claim 1, characterized in that, The specific steps for setting the urgency score based on the order's delivery time are as follows: Based on the time difference between the current time and the delivery deadline, orders are assigned different preset urgency scores, and manual adjustment of the urgency scores is supported.

4. The automatic order scheduling method based on delivery address and urgency according to claim 1, characterized in that, The address grouping step further includes: When the total weight of goods in a region exceeds the preset load ratio of the matching vehicle model or the total volume exceeds the preset load ratio of the matching vehicle model's volume, the region is divided into at least two sub-regions based on their distance.

5. The automatic order scheduling method based on delivery address and urgency according to claim 1, characterized in that, The specific steps for determining the delivery priority between areas are as follows: Calculate the sum of the urgency scores of all orders in each region to obtain the total urgency score for that region; then sort the regions in descending order of their total urgency scores, with regions having higher total scores having higher delivery priority.

6. The automatic order scheduling method based on delivery address and urgency according to claim 1, characterized in that, The specific steps for determining the order delivery order within each region are as follows: For orders within a single region, sort them in descending order by urgency score; when urgency scores are the same, sort them in ascending order by distance from the order address to the distribution center.

7. The automatic order scheduling method based on delivery address and urgency according to claim 1, characterized in that, The specific steps for vehicle matching are as follows: From the available vehicles that meet the upper limits of load capacity and volume capacity, select vehicles with a load utilization rate and volume utilization rate in the range of 80%-90% for matching.

8. The automatic order scheduling method based on delivery address and urgency according to claim 1, characterized in that, The method also includes a dynamic adjustment step: When an emergency order insertion or vehicle malfunction signal is received, a dynamic adjustment that is not a global reordering is performed. When an urgent order with an urgency score greater than or equal to a preset threshold is received, the dynamic adjustment step is as follows: Parse the GeoHash code of the new order to determine its region; insert the order at the top of the order sorting of the region; and only update the schedule of the region, while the schedules of other regions remain unchanged.

9. The automatic order scheduling method based on delivery address and urgency according to claim 8, characterized in that, When a fault signal is received from a matched vehicle, the dynamic adjustment steps are as follows: According to the preset scheduling priority rules, a backup vehicle is selected from the backup vehicles and assigned to take over the task of the faulty vehicle; and the departure time of the original area of ​​the faulty vehicle is slightly adjusted, while the schedule of other areas remains unchanged.

10. The automatic order scheduling method based on delivery address and urgency according to claim 1, characterized in that, The scheduling order includes: Calculate the total urgency score for each region and determine the delivery priority between regions accordingly. Also, determine the order delivery order within each region based on the urgency score and the distance from the order address to the delivery center. The scheduling plan includes vehicle number, area name, and order list.

Images