A method for high-throughput streaming delivery order information queue assignment and scheduling

Abstract

This invention relates to the field of delivery order processing technology, and discloses a high-throughput streaming delivery order information queue assignment and scheduling method. The method includes: constructing a state storage layer, an order queue management layer, and an engine processing layer; dynamically updating the spatiotemporal characteristics of the delivery party and generating an order scheduling vector for the delivery party; detecting high-throughput new orders within the target area through multiple parallel processing units, obtaining the spatiotemporal characteristics of the new orders, comparing the spatiotemporal characteristics of new orders that match the delivery party's order scheduling vector with the delivery party's spatiotemporal characteristics, and determining the new orders as queued orders or streaming orders based on the optimal spatiotemporal matching result; and using the engine processing layer to perform streaming assignment of streaming orders and unified assignment and scheduling of multiple order queues. This invention helps solve the technical problem that existing order assignment and scheduling methods often lead to order delays.

Description

Technical Field

[0001] This invention relates to the field of delivery order processing technology, and in particular to a high-throughput streaming delivery order information queue assignment and scheduling method. Background Technology

[0002] A delivery order refers to a logistics and delivery task generated after a user places an order. Currently, with the rapid development of online sales and e-commerce models, the growth rate of logistics and delivery data generated through online platforms is significant.

[0003] Food delivery orders on food delivery platforms represent a massive volume of orders, encompassing the entire process from user order placement, pickup from the merchant or warehouse, to delivery to the designated location. It serves as a core data carrier connecting user demand, merchant supply, and delivery resources.

[0004] Meanwhile, food delivery orders on food delivery platforms have distinct characteristics, including: short delivery times and the need for real-time tracking of delivery personnel, with penalties for late orders; multiple variables, such as fluctuating food preparation times, user address changes, or unexpected situations with the delivery person; and the reliance on high-precision location technology. The efficient allocation and scheduling of food delivery orders directly impacts user experience, delivery personnel scheduling costs, and platform operational efficiency.

[0005] In recent years, the food delivery market has continued to grow at an extraordinary pace. For example, recently, the peak daily order volume of a certain food delivery platform reached tens of thousands. Currently, various food delivery platforms are engaged in comprehensive competition across all dimensions, including quality supply, delivery experience, and software experience. Among these, delivery timeliness and on-time rate, as important indicators in the fulfillment process, are one of the core competitive advantages of food delivery platforms.

[0006] Over the past decade, academic research on delivery algorithms has largely focused on "how to be faster," specifically how to compress delivery time through optimized route planning, aiming to deliver orders to users as quickly as possible within the expected delivery time. Based on this approach, current technologies typically allocate food delivery orders to riders within 3-5 seconds to ensure delivery efficiency. The allocation logic prioritizes riders closest to the merchant and monitors the number of orders a rider is currently carrying to prevent overloading (e.g., a maximum of 3-5 orders per rider), ensuring delivery efficiency. Furthermore, the system predicts the time it will take for a rider to complete their current order and prioritizes assigning it to the rider who can accept it fastest. Therefore, in existing order allocation methods, the most important allocation logic after an order is placed is: efficiently allocating orders to the nearest rider while minimizing the rider's workload. Thus, the existing food delivery order allocation logic aims to achieve efficient, short-distance, and low-volume deliveries, hoping to improve user satisfaction and order placement probability by delivering orders as quickly as possible.

[0007] The technical problem with this "how to speed up" order allocation and delivery is that while it improves delivery speed in busy areas with high rider density, it often leads to delays in less busy areas. Orders may be assigned to riders further away due to a lack of nearby riders, or riders may have to wait for others to finish their current orders before delivering new ones. This is especially true during peak ordering times and in inclement weather.

[0008] However, existing technologies rarely address a crucial issue: users don't continue ordering simply because delivery is a few minutes or even ten minutes faster; in fact, delayed delivery can lead to significant user churn. A recent study, based on actual transaction data from a leading food delivery platform (covering tens of millions of orders), empirically quantified for the first time the asymmetric impact of early and delayed delivery on user repurchase behavior. To accurately capture the long-term impact of delivery timeliness on user behavior, the research team constructed a two-stage analytical framework: first, a survival analysis model was built to empirically test the impact of early or delayed delivery on customers' future orders; second, a simulation-based evaluation framework was developed to compare the impact of different order allocation objective functions (such as minimizing average delivery time, minimizing average delay, and maximizing the number of future orders) on platform performance. The study found that delayed delivery significantly reduced customers' future orders, while the benefits of early delivery were limited. This research quantified the impact of early and delayed delivery on customers' repeat purchase behavior, revealing that the negative impact of delayed delivery on customers' future orders far outweighs the positive impact of early delivery.

[0009] Therefore, existing order dispatching and scheduling methods often lead to order delays, resulting in decreased user satisfaction, especially during peak order periods and inclement weather. Thus, there is an urgent need to propose a high-throughput streaming delivery order information queue dispatching and scheduling method. Summary of the Invention

[0010] The main objective of this invention is to provide a high-throughput streaming delivery order information queue assignment and scheduling method, which aims to solve the technical problem that existing order assignment and scheduling methods often lead to order delivery delays.

[0011] To achieve the above objectives, the present invention provides a high-throughput streaming delivery order information queue assignment and scheduling method, comprising the following steps:

[0012] Build a state storage layer, an order queue management layer, and an engine processing layer for streaming management of high-throughput orders;

[0013] The state storage layer detects the delivery location information, order delivery status, remaining delivery task information, and current load information of the delivery party within the target area to dynamically update the spatiotemporal characteristics of the delivery party and generate the order scheduling vector of the delivery party. The current load information of the delivery party is controlled by the delivery timeout time.

[0014] Multiple parallel processing units in the order queue management layer detect high-throughput new orders within the target area, obtain the spatiotemporal characteristics of the new orders, compare the spatiotemporal characteristics of new orders that match the delivery party's order scheduling vector with the delivery party's spatiotemporal characteristics, and determine the new orders as either queued orders to be queued and assigned to the delivery party or streaming orders to be streamed and assigned to the delivery party based on the optimal spatiotemporal matching result. New orders that belong to queued orders are added to the corresponding multi-order queue, and new orders that belong to streaming orders are sent to the engine processing layer as streaming information.

[0015] Through the engine processing layer, streaming orders are distributed to the corresponding delivery providers in the target area when the streaming distribution conditions are met, and multi-order queues are uniformly distributed and scheduled to delivery providers with spatiotemporal characteristics matching in the target area when the queue distribution conditions are met.

[0016] Optionally, the step of detecting the delivery party's location information, order delivery status, remaining delivery tasks, and current load information within the target area through the state storage layer to dynamically update the delivery party's spatiotemporal characteristics and generate the delivery party's order scheduling vector includes:

[0017] The state storage layer detects the delivery location information, order delivery status, remaining delivery tasks, and current load information of the delivery party within the target area.

[0018] Based on the first merchant address and first delivery address of each order to be delivered in the remaining delivery task information of the delivery party, generate a set of remaining task addresses of the delivery party;

[0019] Based on the first merchant pickup time, first delivery timeout time and set of remaining task addresses for each order to be delivered in the remaining delivery task information of the delivery party, the spatiotemporal vector of the delivery party's task path is determined to generate the spatiotemporal features of the delivery party;

[0020] Based on the delivery provider's location information and order delivery status, update the delivery provider's current load information and dynamically update the delivery provider's spatiotemporal characteristics;

[0021] Based on the start and end points of the remaining delivery tasks from the delivery provider, an order scheduling vector is fitted to form the order scheduling vector.

[0022] The order scheduling region is determined based on the order scheduling vector.

[0023] Optionally, the step of detecting high-throughput new orders within the target area through multiple parallel processing units of the order queue management layer includes:

[0024] The order queue management layer receives high-throughput new orders within the target region.

[0025] Add the new order to the order processing sequence of the order queue management layer, and determine whether the second merchant address and second delivery address of the new order belong to the order scheduling area;

[0026] If so, the new order will be distributed to the parallel processing unit corresponding to the order scheduling area;

[0027] If not, the new order will be distributed to a parallel processing unit that is not associated with the delivery party's order scheduling area.

[0028] Optionally, the step of distributing new orders to parallel processing units corresponding to order scheduling areas includes:

[0029] Each new order in the order processing sequence is mapped to a parallel processing unit corresponding to its order scheduling region, so that the new orders in the order processing sequence are the original orders, and the mapped orders in each parallel processing unit are the mapped images of the original orders.

[0030] Optionally, the step of obtaining the spatiotemporal characteristics of new orders, which involves comparing the spatiotemporal characteristics of new orders that match the delivery party's order scheduling vector with the delivery party's spatiotemporal characteristics, and determining the new order as a queued order to be queued and dispatched to the delivery party or a streaming order to be streamed and dispatched to the delivery party based on the optimal spatiotemporal matching result, includes:

[0031] Each parallel processing unit obtains the second merchant address, second delivery address, second merchant pickup time, and second delivery timeout time of the received new order to generate the spatiotemporal features of the new order;

[0032] Each parallel processing unit compares the spatiotemporal characteristics of new orders that conform to the delivery party's order scheduling vector with the delivery party's spatiotemporal characteristics to determine the spatiotemporal matching degree.

[0033] New orders whose spatiotemporal feature matching degree meets the preset conditions are identified as queue orders;

[0034] New orders whose spatiotemporal feature matching degree does not meet the preset conditions are identified as streaming orders.

[0035] Optionally, the step of each parallel processing unit comparing the spatiotemporal characteristics of new orders conforming to the delivery party's order scheduling vector with the delivery party's spatiotemporal characteristics to determine the spatiotemporal matching degree includes:

[0036] Each parallel processing unit will locate the new order that conforms to the delivery party's order scheduling vector according to the spatiotemporal characteristics of the order and the spatiotemporal characteristics of the delivery party to determine the time insertion position of the new order;

[0037] The new order is spatially located based on the spatiotemporal characteristics of the order and the spatiotemporal characteristics of the delivery party in order to determine the spatial insertion position of the new order.

[0038] Based on the time and space insertion positions, the spatiotemporal characteristics of the delivery party are corrected to determine the on-time delivery degree of each order under the corrected spatiotemporal characteristics, and the spatiotemporal matching degree is determined based on the on-time delivery degree of each order.

[0039] Optionally, the step of fitting an order scheduling vector based on the start and end points of the remaining delivery tasks from the delivery party includes:

[0040] Compare the delivery distance between the start and end points of the remaining delivery tasks of the delivery party with the set distance;

[0041] When the delivery distance reaches the set distance, the origin and destination will be fitted to form an order scheduling vector;

[0042] When the delivery distance is less than the set distance, the order scheduling vector is calibrated and extended from the starting point to the destination according to the order density direction.

[0043] Optionally, the distance can be set based on the current order density and delivery load information of the area where the delivery party has remaining delivery tasks.

[0044] Optionally, the steps of streaming orders to delivery providers within the corresponding target area when the streaming dispatch conditions are met through the engine processing layer, and uniformly dispatching and scheduling multiple order queues to delivery providers with matching spatiotemporal characteristics within the target area when the queue dispatch conditions are met, include:

[0045] The engine processing layer adds streaming orders to the streaming order set and matches each new order in the streaming order set with available delivery providers in the target area in a time and space manner.

[0046] Within the dispatch time limit, based on the optimal result of spatiotemporal matching, streaming orders are distributed to available delivery providers within the corresponding target area.

[0047] Optionally, the steps of streaming orders to delivery providers within the corresponding target area when the streaming dispatch conditions are met through the engine processing layer, and uniformly dispatching and scheduling multiple order queues to delivery providers with matching spatiotemporal characteristics within the target area when the queue dispatch conditions are met, include:

[0048] The capacity of multiple order queues can be dynamically adjusted through the engine processing layer;

[0049] When the capacity of the multi-order queue is full or the dispatch time limit is reached, the order list in the multi-order queue is uniformly dispatched and scheduled from the order processing sequence to the delivery party with spatiotemporal characteristics matching in the target area, and the mapped orders that have a mapping relationship with each original order in the order list in the parallel processing unit are deleted.

[0050] The technical solution of this invention helps to solve the technical problem that existing order dispatching and scheduling methods often lead to order delays. Specifically, the technical solution of this invention constructs a state storage layer, an order queue management layer, and an engine processing layer for streaming high-throughput orders. The state storage layer can detect delivery parties within the target area, dynamically update the spatiotemporal characteristics of delivery parties based on the detected information, and generate order scheduling vectors for delivery parties. Thus, the state storage layer can continuously track the rider's state, spatiotemporal information, and order scheduling vectors. The order queue management layer is used to receive new orders, obtain the spatiotemporal characteristics of new orders, and perform spatiotemporal feature matching between new orders that match the order scheduling vectors of delivery parties and delivery parties in a parallel processing unit. The best-matched new orders are used as queue orders for delivery parties, while new orders that cannot be spatiotemporally matched with delivery parties are used as streaming orders. The engine processing layer is used to manage the allocation of streaming orders and queue orders. Streaming orders are dispatched to delivery parties using a streaming dispatching method, while queue orders are extracted from multiple order queues and uniformly dispatched and scheduled to delivery parties with spatiotemporal matching. Therefore, the technical solution of this invention employs a dedicated state storage layer to track the spatiotemporal characteristics of each delivery party and a dedicated order queue management layer to detect the spatiotemporal characteristics of new orders. Based on the optimal spatiotemporal matching result, it determines whether to allocate orders to delivery parties in batches or in a streaming manner, and determines the spatiotemporal allocation position for batch allocation of queued orders to delivery parties. This facilitates the use of more advantageous spatiotemporal matching logic to insert new orders into delivery tasks, following the delivery dynamics of delivery parties. It realizes the order allocation method of high-throughput streaming delivery order information according to queue assignment, which is conducive to achieving the optimal spatiotemporal scheduling of delivery parties. The technical solution of this invention does not simply follow the principle of nearest and fast allocation, but continuously inserts new orders through spatiotemporal feature matching while other orders do not time out. Thus, it helps to solve the technical problem that existing order assignment and scheduling methods often lead to order timeouts. Attached Figure Description

[0051] Figure 1 This is a flowchart illustrating the high-throughput streaming delivery order information queue assignment and scheduling method in the first embodiment of the present invention.

[0052] Figure 2 This is a schematic diagram of the functional modules in this invention;

[0053] Figure 3 This is a schematic diagram of the order scheduling vector and the order scheduling region in this invention;

[0054] Figure 4 This is a schematic diagram illustrating the new order processing in this invention as either queued orders or streaming orders;

[0055] Figure 5This is a schematic diagram illustrating the insertion of new orders into the remaining delivery tasks of the delivery party in this invention.

[0056] The objectives, features, and advantages of this invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0057] It should be understood that the specific embodiments described herein are for illustrative purposes only and are not intended to limit the scope of the invention.

[0058] In the following description, the use of suffixes such as "unit," "component," or "element" to denote elements is solely for the purpose of illustrative purposes and has no specific meaning in itself. Therefore, "unit," "component," or "element" may be used interchangeably.

[0059] Please see Figures 1 to 5 The first embodiment of the present invention provides a high-throughput streaming delivery order information queue assignment and scheduling method, comprising the following steps:

[0060] Step S10: Construct a state storage layer, an order queue management layer, and an engine processing layer for streaming control of high-throughput orders;

[0061] Step S20: The delivery location information, order delivery status, remaining delivery task information and current load information of the delivery party in the target area are detected by the state storage layer to dynamically update the spatiotemporal characteristics of the delivery party and generate the order scheduling vector of the delivery party. The current load information of the delivery party is controlled by the delivery timeout time.

[0062] Step S30: Through multiple parallel processing units of the order queue management layer, high-throughput new orders in the target area are detected respectively, and the spatiotemporal characteristics of the new orders are obtained. The spatiotemporal characteristics of the new orders that match the order scheduling vector of the delivery party are compared with the spatiotemporal characteristics of the delivery party. Based on the optimal spatiotemporal matching result, the new orders are determined to be either queued orders to be queued and dispatched to the delivery party or streaming orders to be streamed and dispatched to the delivery party. New orders that belong to queued orders are added to the corresponding multi-order queue, and new orders that belong to streaming orders are sent to the engine processing layer as streaming information.

[0063] Step S40: Through the engine processing layer, streaming orders are distributed to the corresponding delivery providers in the target area when the streaming distribution conditions are met, and multi-order queues are uniformly distributed and scheduled to delivery providers with matching spatiotemporal characteristics in the target area when the queue distribution conditions are met.

[0064] The technical solution of this invention helps to solve the technical problem that existing order dispatching and scheduling methods often lead to order delays. Specifically, the technical solution of this invention constructs a state storage layer, an order queue management layer, and an engine processing layer for streaming high-throughput orders. The state storage layer can detect delivery parties within the target area, dynamically update the spatiotemporal characteristics of delivery parties based on the detected information, and generate order scheduling vectors for delivery parties. Thus, the state storage layer can continuously track the rider's state, spatiotemporal information, and order scheduling vectors. The order queue management layer is used to receive new orders, obtain the spatiotemporal characteristics of new orders, and perform spatiotemporal feature matching between new orders that match the order scheduling vectors of delivery parties and delivery parties in a parallel processing unit. The best-matched new orders are used as queue orders for delivery parties, while new orders that cannot be spatiotemporally matched with delivery parties are used as streaming orders. The engine processing layer is used to manage the allocation of streaming orders and queue orders. Streaming orders are dispatched to delivery parties using a streaming dispatching method, while queue orders are extracted from multiple order queues and uniformly dispatched and scheduled to delivery parties with spatiotemporal matching. Therefore, the technical solution of this invention employs a dedicated state storage layer to track the spatiotemporal characteristics of each delivery party and a dedicated order queue management layer to detect the spatiotemporal characteristics of new orders. Based on the optimal spatiotemporal matching result, it determines whether to allocate orders to delivery parties in batches or in a streaming manner, and determines the spatiotemporal allocation position for batch allocation of queued orders to delivery parties. This facilitates the use of more advantageous spatiotemporal matching logic to insert new orders into delivery tasks, following the delivery dynamics of delivery parties. It realizes the order allocation method of high-throughput streaming delivery order information according to queue assignment, which is conducive to achieving the optimal spatiotemporal scheduling of delivery parties. The technical solution of this invention does not simply follow the principle of nearest and fast allocation, but continuously inserts new orders through spatiotemporal feature matching while other orders do not time out. Thus, it helps to solve the technical problem that existing order assignment and scheduling methods often lead to order timeouts.

[0065] Specifically, the delivery provider can be a rider or other entity that can undertake order delivery tasks.

[0066] Furthermore, after a new order is generated, the order is assigned and scheduled within the set time limit for new order dispatch to avoid delays in new order delivery.

[0067] Each new order in the same multi-order queue is sorted according to its remaining delivery time to manage the timing of new orders. Whenever a new order is added to the multi-order queue, the orders are reordered based on their remaining delivery time to avoid delivery timeouts. Within each multi-order queue, information such as the second merchant address, second delivery address, second order creation time, and second overdue delivery time are stored for each new order. The overdue delivery time refers to the delivery deadline assigned to the order by the system; if the actual delivery time exceeds the overdue delivery time, the order is considered to have timed out.

[0068] Specifically, in the technical solution of this invention, new orders that conform to the order scheduling vector of the delivery party are added to different parallel processing units according to the spatiotemporal characteristics of the orders. The new orders are compared with the spatiotemporal characteristics of the delivery party based on the spatiotemporal characteristics of the orders. Based on the matching results, the new orders are determined as either queued orders to be queued and assigned to the delivery party or streaming orders to be streamed and assigned to the delivery party. Therefore, for new orders that match the spatiotemporal characteristics of the delivery party, without affecting the delivery party's on-time delivery, they can be batch-allocated as queued orders to the delivery party with the highest delivery efficiency. This avoids the unreasonable order matching caused by the allocation method of directly and quickly assigning orders after they are generated in the prior art, thereby facilitating the effective merging of new orders that can be spatiotemporally matched with the delivery party.

[0069] According to the first embodiment of the high-throughput streaming delivery order information queue assignment and scheduling method of the present invention, and the second embodiment of the high-throughput streaming delivery order information queue assignment and scheduling method of the present invention, step S20 includes:

[0070] Step S21: Detect the delivery location information, order delivery status, remaining delivery task information, and current load information of the delivery party within the target area through the state storage layer;

[0071] Step S22: Generate a set of remaining task addresses for the delivery party based on the first merchant address and first delivery address of each order to be delivered in the remaining delivery task information of the delivery party;

[0072] Step S23: Based on the first merchant pickup time, first delivery timeout time and remaining task address set of each order to be delivered in the remaining delivery task information of the delivery party, determine the spatiotemporal vector of the delivery party's task path to generate the spatiotemporal features of the delivery party.

[0073] Step S24: Update the current load information of the delivery party based on the delivery party's location information and order delivery status, and dynamically update the spatiotemporal characteristics of the delivery party.

[0074] Step S25: Based on the start and end points of the remaining delivery tasks of the delivery party, fit and form an order scheduling vector;

[0075] Step S26: Determine the order scheduling area based on the order scheduling vector.

[0076] Specifically, the state storage layer includes multiple parallel spatiotemporal feature calculation modules. Each spatiotemporal feature calculation module uses a distributed GPS positioning cluster to collect the coordinate data of the delivery party, and the update frequency can be twice per second.

[0077] Order delivery status includes: pending order, accepted order but not yet delivered to store, delivered to store but awaiting pickup, pickup completed and in transit, in transit but not yet delivered to customer, and completed.

[0078] The remaining tasks of the delivery party are displayed in real time using an LRU cache structure.

[0079] In step S22, the first merchant address and the first delivery address are converted into geographic coordinates to construct the set of remaining task addresses:

[0080] Let the set of remaining task addresses for delivery party k be . ,in, Let this be the address of the first merchant in the i-th task among the remaining tasks. Let be the first delivery address of the i-th task in the remaining tasks, 1≤i≤n, where n is the number of orders in the remaining tasks. It should be noted that when a task has been picked up, the corresponding merchant address and delivery address will not coexist in the set of addresses of the remaining tasks at the same time, but only the delivery address will exist.

[0081] Then the spatiotemporal characteristics of delivery party k are: ,in, For the current spatiotemporal feature points of the delivery party, For the current address, The current time; Let j be the j-th spatiotemporal feature point obtained by sorting the delivery process in the spatiotemporal features of the delivery party. For the address information corresponding to the j-th spatiotemporal feature, Let j be the time information corresponding to the j-th spatiotemporal feature, 1≤j≤m, where m is the number of addresses in the remaining tasks. The j-th address can be either the first merchant address or the first delivery address. When sorted by delivery process, the first merchant address of the i-th task is located before the first delivery address of the i-th task.

[0082] Based on the delivery provider's location information and order delivery status, completed orders can be removed from the delivery provider's spatiotemporal characteristics to achieve dynamic updates of the delivery provider's spatiotemporal characteristics.

[0083] In this embodiment, the remaining delivery task information of the delivery party expresses the current load information of the delivery party according to the time difference between the estimated delivery time and the order timeout time for each order. A negative time difference indicates expected early delivery, a positive time difference indicates expected late delivery, and a time difference of 0 indicates on-time delivery. The current load information can be marked with spatiotemporal feature points of the corresponding delivery address to form new order insertion constraints for that delivery address. Sorting by delivery process yields a set of new order insertion constraints for each delivery address. Therefore, the load information in this embodiment helps to insert new orders before orders expected to be delivered early in the delivery party's remaining delivery tasks, and conversely, disallows new order insertion before orders not expected to be delivered early, thereby achieving control over delivery timeouts.

[0084] In a second embodiment of the high-throughput streaming delivery order information queue assignment and scheduling method of the present invention, and in a third embodiment of the same method, step S30, which involves detecting high-throughput new orders within the target area through multiple parallel processing units of the order queue management layer, includes:

[0085] Step S31: Receive high-throughput new orders within the target area through the order queue management layer;

[0086] Step S32: Add the new order to the order processing sequence of the order queue management layer, and determine whether the second merchant address and the second delivery address of the new order belong to the order scheduling area;

[0087] If so, proceed to step S33 to distribute the new order to the parallel processing unit corresponding to the order scheduling area;

[0088] If not, proceed to step S34 to distribute the new order to a parallel processing unit that is not associated with the delivery party's order scheduling area.

[0089] Among them, the spatiotemporal characteristics of delivery party k In China, according to and Determine the distance to the order scheduling vector, and then determine the direction of the order scheduling vector based on the positional relationship between each order and the order scheduling vector. Please refer to... Figure 3 The order scheduling vector is used as the diameter to determine the order scheduling area. The order scheduling area is then moved towards the direction of the delivery order to obtain the order scheduling region.

[0090] Specifically, there are multiple delivery companies within the same target area. Therefore, each delivery company will generate an order scheduling vector. The order scheduling area is an address area determined by the direction and distance of the order scheduling vector. The order scheduling areas of different delivery companies can intersect, overlap, or not intersect.

[0091] Specifically, step S32 includes: displaying the new order in the target area using the second merchant address and the second delivery address, where the second merchant address and the second delivery address are the address attributes of the new order; allocating a parallel processing unit to each order scheduling area; and scanning the new order in the target area based on the address attributes using the parallel processing unit; multiple new orders may be generated in the same order scheduling area, so new orders corresponding to the same order scheduling area can be assigned to the same parallel processing unit; furthermore, the same new order may also correspond to multiple intersecting or overlapping order scheduling areas, so the same new order can be distributed to different parallel processing units.

[0092] Specifically, if a new order does not belong to an order scheduling area, it means that there is currently no delivery service available to deliver this new order. In this case, the new order is distributed to a parallel processing unit that does not correspond to an order scheduling area but corresponds to the address attribute of the new order. Therefore, this type of parallel processing unit will receive multiple new orders with similar addresses and form a multi-order queue based on the timeout delivery time to wait for a suitable delivery provider to be associated. Within the new order's dispatch time limit, if a new delivery provider's order scheduling area establishes an allocation relationship with the parallel processing unit, the new order will be matched with the new delivery provider in a spatiotemporal matching degree calculation. If the new order's dispatch time limit has expired and no new delivery provider has yet established an allocation relationship with the parallel processing unit, the new order will be distributed as a streaming order to a nearby delivery provider.

[0093] Based on the third embodiment of the high-throughput streaming delivery order information queue assignment and scheduling method of the present invention, and the fourth embodiment of the high-throughput streaming delivery order information queue assignment and scheduling method of the present invention,

[0094] Step S33 includes:

[0095] Step S331: Map each new order in the order processing sequence to the parallel processing unit corresponding to the order scheduling region according to the corresponding order scheduling region, so that the new orders in the order processing sequence are the original orders, and the mapped orders in each parallel processing unit are the mapped images of the original orders.

[0096] Specifically, in order to improve the efficiency of parallel order processing and prevent the same new order from being delivered repeatedly due to being distributed to different parallel processing units, in this embodiment, after each new order is generated, it is first added to the order processing sequence of the order queue management layer. After the new order is scanned by the parallel processing unit, each new order in the order processing sequence is mapped to the parallel processing unit corresponding to the order scheduling region according to the corresponding order scheduling region. Thus, the new orders in the order processing sequence are the original orders (real orders), while the mapped orders in each parallel processing unit are the mapped images of the original orders (virtual orders).

[0097] Since a single new order may correspond to multiple order scheduling regions, it may be mapped multiple times in different parallel processing units, forming multiple order queues in each unit. If a parallel processing unit needs to assign this new order to a delivery party, it assigns the original order from the order processing sequence. Because the original order is deleted from the order processing sequence, all parallel processing units that have mappings to the original order also delete their mappings, thus preventing multiple dispatches of a single order.

[0098] In the third embodiment of the high-throughput streaming delivery order information queue assignment and scheduling method of the present invention, and in the fifth embodiment of the high-throughput streaming delivery order information queue assignment and scheduling method of the present invention, the step S30 of obtaining the spatiotemporal characteristics of a new order, comparing the spatiotemporal characteristics of a new order that conforms to the order scheduling vector of the delivery party with the spatiotemporal characteristics of the delivery party, and determining the new order as a queue order to be queued and assigned to the delivery party or a streaming order to be streaming and assigned to the delivery party based on the optimal spatiotemporal matching result, includes:

[0099] Step S35: Each parallel processing unit obtains the second merchant address, second delivery address, second merchant pickup time, and second delivery timeout time of the received new order to generate the spatiotemporal features of the new order.

[0100] Step S36: Each parallel processing unit compares the spatiotemporal characteristics of new orders that match the delivery party's order scheduling vector with the delivery party's spatiotemporal characteristics to determine the spatiotemporal matching degree.

[0101] Step S37: New orders that meet the preset conditions for spatiotemporal feature matching are identified as queue orders;

[0102] Step S38: New orders whose spatiotemporal feature matching degree does not meet the preset conditions are identified as streaming orders.

[0103] Specifically, the spatiotemporal characteristics of the f-th new order in the parallel processing unit can be: ;in, The address of the second merchant for the f-th new order in the parallel processing unit. Let f be the pickup time of the second merchant for the f-th new order in the parallel processing unit. The second delivery address for the f-th new order in the parallel processing unit. The second delivery time for the f-th new order in the parallel processing unit;

[0104] In the third embodiment, only the address attribute of the new order is considered to allocate new orders falling within the delivery party's order scheduling area to the corresponding parallel processing unit of the delivery party. In this embodiment, the matching result between the delivery party's spatiotemporal characteristics and the new order's spatiotemporal characteristics is also considered.

[0105] In this embodiment, each parallel processing unit compares the spatiotemporal features of new orders that conform to the order scheduling vector of the delivery party with the spatiotemporal features of the delivery party. This means comparing the spatiotemporal features of new orders scanned in the parallel processing unit corresponding to the order scheduling area of ​​the delivery party with the spatiotemporal features of the delivery party of the corresponding delivery party.

[0106] After comparison, if the spatiotemporal matching degree can reach the preset condition, it indicates that the insertion of new orders in the parallel processing unit into the remaining delivery tasks of the courier will not cause significant interference to the timely delivery of other assigned orders. Conversely, if the spatiotemporal matching degree does not reach the preset condition, it indicates that the insertion of new orders in the parallel processing unit into the remaining delivery tasks of the courier will cause significant interference to the timely delivery of other assigned orders.

[0107] Please see Figure 5 In the fifth embodiment of the high-throughput streaming delivery order information queue assignment and scheduling method of the present invention, and in the sixth embodiment of the high-throughput streaming delivery order information queue assignment and scheduling method of the present invention, step S36 includes:

[0108] Step S361: The new order is spatially located based on the spatiotemporal characteristics of the order and the spatiotemporal characteristics of the delivery party to determine the spatial insertion position of the new order;

[0109] In step S362, each parallel processing unit locates the new order that conforms to the delivery party's order scheduling vector according to the spatiotemporal characteristics of the order, and then determines the time insertion position of the new order.

[0110] Step S363: Based on the time insertion position and spatial insertion position, the spatiotemporal characteristics of the delivery party are corrected to determine the on-time delivery degree of each order based on the corrected spatiotemporal characteristics of the delivery party, and the spatiotemporal matching degree is determined based on the on-time delivery degree of each order.

[0111] Specifically, in step S361, based on spatial attribute positioning, in the order scheduling vector determined by the spatiotemporal characteristics of the delivery party, a first insertion radius area corresponding to the second merchant's pickup address is delineated according to the preset control radius and the second merchant's pickup address; similarly, a second insertion radius area corresponding to the second delivery address is delineated according to the preset control radius and the second delivery address.

[0112] The first insertion radius defines a large, manageable area along the order scheduling vector, facilitating the insertion of pickup tasks while meeting spatial insertion requirements without affecting the on-time delivery of surrounding orders. For the delivery party, pickup tasks from a second merchant can be inserted within the area defined by the first insertion radius, but these pickup tasks cannot be inserted before nodes with high loads.

[0113] Similarly, the second insertion radius defines a larger, manageable area along the order scheduling vector, facilitating the insertion of delivery tasks while meeting spatial insertion requirements without affecting the on-time delivery of surrounding orders. For the delivery provider, a delivery task destined for the second delivery address can be inserted within the range defined by the second insertion radius, but this task cannot be inserted before a heavily loaded node.

[0114] In step S362, based on the load information of the delivery party in delivering each assigned order in the remaining delivery tasks, multiple optional time insertion positions are generated in areas where the load meets the insertion condition. Here, the time insertion position refers to which addresses are allowed to be inserted for pickup and delivery tasks, provided that delivery timeliness conditions permit. For example, some locations have a high risk of timeout, so inserting pickup and delivery tasks for new orders before these locations is not allowed. Conversely, some locations have a low risk of timeout, so inserting pickup and delivery tasks for new orders before these locations is permitted. Thus, multiple optional time insertion positions can be determined, and the time insertion positions are represented by address regions.

[0115] Furthermore, the spatial insertion position in step S361 and the temporal insertion position in step S362 are both represented by address regions. Therefore, by performing the intersection operation of the first insertion radius region and the temporal insertion position, the first allowed spatiotemporal insertion position, i.e., the insertion position corresponding to the second merchant address, can be obtained; by performing the intersection operation of the second insertion radius region and the temporal insertion position, the second allowed spatiotemporal insertion position, i.e., the insertion position corresponding to the second delivery address, can be obtained.

[0116] In step S363, the second merchant address is inserted according to the first allowed spatiotemporal insertion position, and the second delivery address is inserted according to the second allowed spatiotemporal insertion position, so as to correct the spatiotemporal characteristics of the delivery party and determine the on-time delivery degree of each order based on the corrected spatiotemporal characteristics of the delivery party.

[0117] When the same new order is added to multiple multi-order queues, the on-time delivery rate of each order differs depending on the modified spatiotemporal characteristics of the delivery party for each queue. Therefore, before the new order's dispatch deadline is reached, the optimal spatiotemporal matching result is the multi-order sequence with the optimal on-time delivery rate of each order based on the modified spatiotemporal characteristics of the delivery party. At this point, the spatiotemporal matching effect of adding the new order to other multi-order sequences cannot surpass this optimal spatiotemporal matching result. Based on the optimal spatiotemporal matching result, the order list assigned to a particular delivery party is determined.

[0118] In the second embodiment of the high-throughput streaming delivery order information queue assignment and scheduling method of the present invention, and in the seventh embodiment of the high-throughput streaming delivery order information queue assignment and scheduling method of the present invention, step S25 includes:

[0119] Step S251: Compare the delivery distance between the start and end points of the remaining delivery tasks of the delivery party with the set distance;

[0120] Step S252: When the delivery distance reaches the set distance, the starting point and the destination are fitted to form an order scheduling vector;

[0121] Step S253: When the delivery distance is less than the set distance, the order scheduling vector is calibrated and extended from the starting point to the end point according to the order density direction to fit and form an order scheduling vector.

[0122] Specifically, when the distance between the start and end points of the remaining delivery tasks reaches the set distance, it means that the current delivery distance of the delivery party meets the set distance.

[0123] If the distance between the start and end points of the remaining delivery tasks of the delivery party does not reach the set distance, it means that the current delivery distance of the delivery party is too short. In order to avoid the delivery party performing delivery with a light load during high-throughput periods, which would cause new orders to be unable to be allocated to the appropriate delivery party in a timely manner, the delivery distance between the current start and end points will be tilted and extended in the direction of higher order density to determine the order scheduling vector. Then, it will be determined whether new orders within the order scheduling area determined by the order scheduling vector can be merged into the remaining delivery tasks of the delivery party.

[0124] Specifically, the distance is set based on the current order density and delivery load information of the area where the delivery party has remaining delivery tasks.

[0125] When the current order density is low, a larger distance can be set so that each delivery party can accept orders over a wider area during the delivery process. This avoids delivery parties not receiving orders for a short period of time and ensures that orders in remote areas can also be delivered in a timely manner, thereby increasing the probability of new orders being merged into the remaining delivery tasks of the delivery party. Conversely, when the current order density is high, a smaller distance can be set so that delivery parties can quickly allocate orders to areas with high order density and deliver efficiently within a small area.

[0126] When the delivery provider is heavily loaded, it indicates that there are many orders in the planned delivery area. Orders in the current planned delivery area can be prioritized. Therefore, the set distance can be adjusted to a smaller value. Conversely, when the delivery provider is lightly loaded, it indicates that there are fewer orders in the planned delivery area. The delivery area can be expanded. Therefore, the set distance can be adjusted to a larger value.

[0127] In the first to seventh embodiments of the high-throughput streaming delivery order information queue assignment and scheduling method of the present invention, and in the eighth embodiment of the high-throughput streaming delivery order information queue assignment and scheduling method of the present invention, step S40 includes:

[0128] Step S41: The streaming orders are added to the streaming order set through the engine processing layer, and each new order in the streaming order set is matched with an available delivery party in the target area in time and space.

[0129] Step S42: Within the dispatch time limit, the streaming orders are dispatched to the idle delivery parties in the corresponding target areas according to the optimal result of spatiotemporal matching.

[0130] It is easy to understand that when an idle delivery party is assigned an order, it can determine the order scheduling vector based on the delivery start and end points, and associate the order scheduling vector with the parallel processing unit. Thus, new orders that conform to the order scheduling vector can be added to the parallel processing unit during the delivery process, and the method steps S20 to S40 in this invention can be executed.

[0131] In the fourth embodiment of the high-throughput streaming delivery order information queue assignment and scheduling method of the present invention, and in the ninth embodiment of the high-throughput streaming delivery order information queue assignment and scheduling method of the present invention, step S40 includes:

[0132] Step S43: Dynamically adjust the capacity of the multi-order queue through the engine processing layer;

[0133] Step S44: When the capacity of the multi-order queue is full or the dispatch time limit is reached, the order list in the multi-order queue is uniformly dispatched and scheduled from the order processing sequence to the delivery party with spatiotemporal characteristics matching in the target area, and the mapped orders that have a mapping relationship with each original order in the order list in the parallel processing unit are deleted.

[0134] Specifically, adjusting the capacity of the multi-order queue includes the following steps: When a new order is inserted into the remaining delivery tasks of the delivery party, if it does not affect the on-time delivery rate of each order in the corrected spatiotemporal characteristics of the delivery party, the first spatiotemporal insertion position and the second spatiotemporal insertion position corresponding to the new order can be stored in the order information of the new order, and the insertion calculation of the next new order can be performed until the insertion of a new order affects the on-time delivery rate of each order in the corrected spatiotemporal characteristics of the delivery party. At this point, the insertion calculation stops, and all new orders that have been successfully inserted are used as the order list to be allocated to the delivery party. At the same time, the capacity of the multi-order queue is marked as full, and the order list from the multi-order queue is packaged and allocated to the corresponding delivery party.

[0135] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods of the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms, or by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a readable storage medium (such as ROM / RAM, magnetic disk, optical disk) as described above, and includes several instructions to cause a terminal device to enter the methods described in the various embodiments of the present invention.

[0136] In the description of this specification, references to terms such as "one embodiment," "another embodiment," "other embodiments," or "first embodiment to Xth embodiment," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of the present invention. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, method steps, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0137] It should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or system that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or system. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or system that includes that element.

[0138] The sequence numbers of the above embodiments of the present invention are for descriptive purposes only and do not represent the superiority or inferiority of the embodiments.

[0139] The above are merely preferred embodiments of the present invention and do not limit the scope of the patent. Any equivalent structural or procedural transformations made based on the description and drawings of the present invention, or direct or indirect applications in other related technical fields, are similarly included within the scope of patent protection of the present invention.

Claims

1. A method for high-throughput streaming delivery order information queue assignment and scheduling, characterized in that, Includes the following steps: Build a state storage layer, an order queue management layer, and an engine processing layer for streaming management of high-throughput orders; The state storage layer detects the delivery location information, order delivery status, remaining delivery task information, and current load information of the delivery party within the target area to dynamically update the spatiotemporal characteristics of the delivery party and generate the order scheduling vector of the delivery party. The current load information of the delivery party is controlled by the delivery timeout time. Multiple parallel processing units in the order queue management layer detect high-throughput new orders within the target area, obtain the spatiotemporal characteristics of the new orders, compare the spatiotemporal characteristics of new orders that match the delivery party's order scheduling vector with the delivery party's spatiotemporal characteristics, and determine the new orders as either queued orders to be queued and assigned to the delivery party or streaming orders to be streamed and assigned to the delivery party based on the optimal spatiotemporal matching result. New orders that belong to queued orders are added to the corresponding multi-order queue, and new orders that belong to streaming orders are sent to the engine processing layer as streaming information. Through the engine processing layer, streaming orders are distributed to the corresponding delivery providers in the target area when the streaming distribution conditions are met, and multi-order queues are uniformly distributed and scheduled to delivery providers with spatiotemporal characteristics matching in the target area when the queue distribution conditions are met. The step of obtaining the spatiotemporal characteristics of new orders, which involves comparing the spatiotemporal characteristics of new orders that conform to the order scheduling vector of the delivery party with the spatiotemporal characteristics of the delivery party, and determining whether the new order is a queued order or a streaming order that is streamed to the delivery party based on the optimal spatiotemporal matching result, includes: each parallel processing unit obtaining the second merchant address, second delivery address, second merchant pickup time, and second delivery timeout time of the received new order to generate the spatiotemporal characteristics of the new order; each parallel processing unit comparing the spatiotemporal characteristics of new orders that conform to the order scheduling vector of the delivery party with the spatiotemporal characteristics of the delivery party to determine the spatiotemporal matching degree; new orders whose spatiotemporal characteristic matching degree reaches a preset condition are determined as queued orders; and new orders whose spatiotemporal characteristic matching degree does not reach the preset condition are determined as streaming orders. The step of each parallel processing unit comparing the spatiotemporal characteristics of new orders conforming to the delivery party's order scheduling vector with the delivery party's spatiotemporal characteristics to determine the spatiotemporal matching degree includes: each parallel processing unit performing time attribute positioning of new orders conforming to the delivery party's order scheduling vector based on the order spatiotemporal characteristics to determine the time insertion position of the new orders; performing spatial attribute positioning of new orders based on the order spatiotemporal characteristics to determine the spatial insertion position of the new orders; and correcting the delivery party's spatiotemporal characteristics based on the time insertion position and the spatial insertion position to determine the on-time delivery degree of each order under the corrected delivery party spatiotemporal characteristics, and determining the spatiotemporal matching degree based on the on-time delivery degree of each order.

2. The high-throughput streaming delivery order information queue assignment and scheduling method according to claim 1, characterized in that, The step of detecting the delivery party's location information, order delivery status, remaining delivery tasks, and current load information within the target area through the state storage layer to dynamically update the delivery party's spatiotemporal characteristics and generate the delivery party's order scheduling vector includes: The state storage layer detects the delivery location information, order delivery status, remaining delivery tasks, and current load information of the delivery party within the target area. Based on the first merchant address and first delivery address of each order to be delivered in the remaining delivery task information of the delivery party, generate a set of remaining task addresses of the delivery party; Based on the first merchant pickup time, first delivery timeout time and set of remaining task addresses for each order to be delivered in the remaining delivery task information of the delivery party, the spatiotemporal vector of the delivery party's task path is determined to generate the spatiotemporal features of the delivery party; Based on the delivery provider's location information and order delivery status, update the delivery provider's current load information and dynamically update the delivery provider's spatiotemporal characteristics; Based on the start and end points of the remaining delivery tasks from the delivery provider, an order scheduling vector is fitted to form the order scheduling vector. The order scheduling region is determined based on the order scheduling vector.

3. The high-throughput streaming delivery order information queue assignment and scheduling method according to claim 2, characterized in that, The step of detecting high-throughput new orders within the target area through multiple parallel processing units of the order queue management layer includes: The order queue management layer receives high-throughput new orders within the target region. Add the new order to the order processing sequence of the order queue management layer, and determine whether the second merchant address and second delivery address of the new order belong to the order scheduling area; If so, the new order will be distributed to the parallel processing unit corresponding to the order scheduling area; If not, the new order will be distributed to a parallel processing unit that is not associated with the delivery party's order scheduling area.

4. The high-throughput streaming delivery order information queue assignment and scheduling method according to claim 3, characterized in that, The step of distributing new orders to parallel processing units corresponding to the order scheduling regions includes: Each new order in the order processing sequence is mapped to a parallel processing unit corresponding to its order scheduling region, so that the new orders in the order processing sequence are the original orders, and the mapped orders in each parallel processing unit are the mapped images of the original orders.

5. The high-throughput streaming delivery order information queue assignment and scheduling method according to claim 2, characterized in that, The step of fitting an order scheduling vector based on the start and end points of the remaining delivery tasks from the delivery party includes: Compare the delivery distance between the start and end points of the remaining delivery tasks of the delivery party with the set distance; When the delivery distance reaches the set distance, the origin and destination will be fitted to form an order scheduling vector; When the delivery distance is less than the set distance, the order scheduling vector is calibrated and extended from the starting point to the destination according to the order density direction.

6. The high-throughput streaming delivery order information queue assignment and scheduling method according to claim 5, characterized in that, The distance is set based on the current order density and delivery load information of the area where the delivery party has remaining delivery tasks.

7. The high-throughput streaming delivery order information queue assignment and scheduling method according to claim 1, characterized in that, The steps of streaming orders to delivery providers within the corresponding target area when the streaming dispatch conditions are met, and uniformly dispatching and scheduling multiple order queues to delivery providers with matching spatiotemporal characteristics within the target area when the queue dispatch conditions are met, through the engine processing layer, include: The engine processing layer adds streaming orders to the streaming order set and matches each new order in the streaming order set with available delivery providers in the target area in a time and space manner. Within the dispatch time limit, based on the optimal result of spatiotemporal matching, streaming orders are distributed to available delivery providers within the corresponding target area.

8. The high-throughput streaming delivery order information queue assignment and scheduling method according to claim 4, characterized in that, The steps of streaming orders to delivery providers within the corresponding target area when the streaming dispatch conditions are met, and uniformly dispatching and scheduling multiple order queues to delivery providers with matching spatiotemporal characteristics within the target area when the queue dispatch conditions are met, through the engine processing layer, include: The capacity of multiple order queues can be dynamically adjusted through the engine processing layer; When the capacity of the multi-order queue is full or the dispatch time limit is reached, the order list in the multi-order queue is uniformly dispatched and scheduled from the order processing sequence to the delivery party with spatiotemporal characteristics matching in the target area, and the mapped orders that have a mapping relationship with each original order in the order list in the parallel processing unit are deleted.

Images