Delivery real-time monitoring system applied to take-out platform

Abstract

The invention discloses a distribution real-time monitoring system applied to a take-out platform, relates to the technical field of distribution real-time monitoring, and solves the technical problem that in the prior art, differential management and control are not implemented for different imbalance scenes. A multi-dimensional detection unit, a sectional distribution unit, a regional order distribution unit and a track accurate updating unit are planned through a distribution monitoring platform, and a full-closed-loop management and control system of monitoring, research and judgment, scheduling and optimization is constructed. The multi-dimensional detection unit collects multi-dimensional data, weights the multi-dimensional data to obtain an influence coefficient, and combines order demand fluctuation to realize early warning scene classification; the sectional distribution unit implements refined batch division and order transfer scheduling for a low-supply and high-backlog scene, and the regional order distribution unit realizes order differentiation distribution based on a rider track feature for a supply-demand imbalance scene; according to the invention, through multi-dimensional collaborative management and control, accurate identification and differentiated scheduling of the distribution scene are realized, and the distribution efficiency and the supply and demand matching stability are effectively improved.

Description

Technical Field

[0001] This invention relates to the field of real-time delivery monitoring technology, specifically a real-time delivery monitoring system applied to food delivery platforms. Background Technology

[0002] With the rapid development of the food delivery industry, food delivery services have become an important part of urban life services, and users' demands for delivery timeliness and stability continue to increase. Currently, the delivery scheduling of food delivery platforms is mostly based on the basic matching of the number of orders and the number of riders, lacking a comprehensive consideration of the multi-dimensional influencing factors of the entire delivery chain.

[0003] The existing technology has the following technical defects:

[0004] First, the delivery efficiency prediction dimension is too single. The existing system only relies on order volume prediction and does not take into account multiple dynamic factors such as the delay at the end of the collection end (e.g., elevator waiting), the matching degree of the order preparation and waiting order volume at the food preparation end, and weather changes. This results in large deviations in efficiency prediction and makes it impossible to avoid the risk of delivery congestion in advance.

[0005] Secondly, in handling the imbalance between order supply and demand, a uniform rider dispatch strategy is mostly adopted, without implementing differentiated management for different imbalance scenarios (such as low supply and high backlog, supply and demand imbalance); and the rationality of order batch division and rider route matching is lacking, which easily leads to problems such as long rider waiting time to pick up food, high overlap rate of delivery routes, and chaotic order transfer.

[0006] Third, in scenarios of supply and demand imbalance, order allocation is unreasonable. The familiarity and suitability of riders to the delivery route are not taken into account. Complex orders are randomly assigned to simple orders, resulting in large differences in rider delivery efficiency. Riders waiting at food preparation points become congested, further exacerbating delivery delays.

[0007] To address the aforementioned technical shortcomings, a solution is proposed that achieves refined and precise control over the entire food delivery chain by constructing a multi-dimensional collaborative real-time monitoring and scheduling system. Summary of the Invention

[0008] The purpose of this invention is to solve the problems mentioned above by proposing a real-time delivery monitoring system for food delivery platforms.

[0009] The objective of this invention can be achieved through the following technical solutions:

[0010] A real-time delivery monitoring system for food delivery platforms includes a delivery monitoring platform, wherein the delivery monitoring platform has communication connections to:

[0011] The multi-dimensional detection unit monitors the coverage area of ​​the food delivery platform in real time and predicts the delivery efficiency of the food delivery platform through multi-dimensional data collection.

[0012] Segmented delivery units allow for segmented delivery control within the delivery area;

[0013] The regional order allocation unit performs order allocation analysis for delivery areas.

[0014] Furthermore, the process of the multi-dimensional detection unit is as follows:

[0015] Identify the coverage area of ​​the food delivery platform and mark it as the delivery area. Identify business districts within the delivery area using a map and mark them as food preparation points. At the same time, identify office buildings or residential buildings within the delivery area and mark them as food collection points.

[0016] Using a 15-minute time window, multi-dimensional data is collected from the delivery area to obtain the average delay time of elevator waiting time corresponding to the collection point; the overlap time between the period when the food preparation speed increases and the period when the total number of food items to be prepared increases at the food preparation point; and the duration of weather characteristic parameter fluctuations within the road network corresponding to the food preparation and collection points. The average delay time, overlap time, and duration of parameter fluctuations are then uniformly labeled as a multi-dimensional data set.

[0017] Furthermore, the system acquires real-time order demand fluctuations within the delivery area and determines the order demand at each time point. Simultaneously, it weights the multi-dimensional data set to obtain multi-dimensional influence coefficients, and uses the order demand at each time point and the multi-dimensional influence coefficients to determine early warning scenarios.

[0018] If the multi-dimensional influence coefficients show an increasing trend during the floating phase, the order supply in the delivery area will decrease, and the current period will be marked as a low supply period; if the order demand does not increase during the low supply period, a low supply with no backlog signal will be generated and sent to the delivery monitoring platform; if the order demand increases during the low supply period, a low supply with high backlog signal will be generated and sent to the delivery monitoring platform.

[0019] If order demand shows an increasing trend, the current period is marked as a high-supply period; if the multi-dimensional influence coefficients show an increasing trend during the high-supply period, a supply-demand imbalance signal is generated and sent to the delivery monitoring platform; if the multi-dimensional influence coefficients show a decreasing trend during the high-supply period, a supply-demand stability signal is generated and sent to the delivery monitoring platform.

[0020] Furthermore, if the delivery monitoring platform receives a low supply with no backlog signal or a stable supply and demand signal, it infers that there is a fluctuation in the supply and demand balance in the delivery area, but the delivery efficiency is within the acceptable range. If the delivery monitoring platform receives a low supply with high backlog signal, it generates a segmented delivery signal and sends it to the segmented delivery unit. If the delivery monitoring platform receives a supply and demand imbalance signal, it generates a regional order allocation signal and sends it to the regional order allocation unit.

[0021] Furthermore, the process for segmented delivery units is as follows:

[0022] The order supply and demand gap of each business district in the delivery area is collected in 15-minute increments; when the order supply and demand gap of a business district is higher than 20% of the set value, riders within a 5-kilometer radius of the business district are dispatched to fill the gap; the location of the food collection point is identified based on the real-time orders in the business district, and the real-time food order batches are divided with a 2-minute interval between food preparation times as the threshold, that is, orders within a 2-minute interval are uniformly planned.

[0023] The riders are divided into the same batch of food orders based on the path overlap algorithm. That is, the overlapping sections of the delivery routes of the same batch of orders received by the rider are less than a set threshold. After the order is divided into the same batch of food orders, the rider uses the first order received as the delivery location and records it in the background.

[0024] Based on the order information received by riders in the backend, the continuous pickup area is determined. When a rider moves to the pickup area, the backend identifies the pickup point location of the order to be processed in the pickup area and the rider's delivery direction. If the directions match, the current order is matched with the rider. If the directions do not match, the backend counts the orders of the riders currently in the pickup area. The riders with completed order matching are used as the dispatcher, and the pickup point locations of the orders to be delivered on the dispatcher are collected and compared with the delivery directions of the riders who have not completed matching.

[0025] Furthermore, during delivery location comparison, if the locations match, the pending orders from the dispatch terminal are transferred to the remaining pending orders of the riders who currently have no matching orders. The backend directly switches the orders and replans the delivery routes of each rider. If the planned route increases the delivery time or delivery distance, the order transfer is canceled. If the locations do not match, rider replacement continues until the rider with the no-match order enters the store, at which point the current food order is temporarily matched with the current rider.

[0026] Furthermore, the system generates segmented delivery instructions, marks the temporarily matched rider trajectory with the corresponding food delivery order's pickup point, and simultaneously performs real-time statistics on the remaining trajectories of all riders within the delivery area. If a rider's real-time remaining trajectory overlaps with the temporarily matched rider's trajectory, and the distance between the food delivery order's pickup point and any pickup point in the rider's real-time remaining trajectory is less than a set threshold, then the overlapping point is identified and a waiting interval is set. Riders then alternate orders at the overlapping point; and the system updates the rider information in the background.

[0027] Furthermore, the process for regional order allocation units is as follows:

[0028] Orders are allocated based on the riders waiting to pick up their meals at the designated locations. Once a rider enters the store, the data on their pending delivery orders is analyzed. These pending delivery orders include orders currently waiting in the store.

[0029] The locations of the pickup points for the orders to be delivered are extracted, and the features of the delivery trajectory for each order to be delivered are extracted based on the extracted locations. The trajectory familiarity features and trajectory matching features of the riders waiting to pick up the food are also extracted.

[0030] If any quantitative parameter corresponding to the extracted features of the delivery trajectory of an order exceeds the threshold set by the corresponding quantitative parameter, the order to be delivered by the corresponding rider will be set as an easy order; if any quantitative parameter corresponding to the extracted features of the delivery trajectory of an order does not exceed the threshold set by the corresponding quantitative parameter, the order to be delivered by the corresponding rider will be set as a difficult order.

[0031] Furthermore, the orders awaiting delivery by riders waiting to pick up food in the store are compared with each other. If the order types of adjacent riders are correspondingly alternating, the corresponding adjacent riders are set as an allocable combination, and the orders awaiting delivery are allocable orders. Once the allocable combination is determined, the allocable orders are allocated. If the total number of orders for riders does not decrease, the allocation combination is adjusted, and the order delivery information is modified in the background. At this time, the orders at the food preparation point are among the allocable orders. Riders are selected based on the optimal delivery route analysis, and the orders at the current food preparation point are assigned to the same rider. Riders who have executed the order allocation combination but have not been assigned to the current food preparation point do not need to wait and can directly deliver the orders awaiting delivery.

[0032] Furthermore, the quantitative parameters of the trajectory familiarity feature are the total historical delivery overlap and the frequency of no wrong routes in the historical delivery for the rider to be picked up; the quantitative parameters of the trajectory fitting feature are the delivery speed and the frequency of no pauses for the rider to be picked up corresponding to the delivery type of the delivery trajectory. The delivery type is represented by the walking type or the stair climbing type in the delivery trajectory.

[0033] Compared with the prior art, the beneficial effects of the present invention are:

[0034] 1. The multi-dimensional detection unit monitors the coverage area of ​​the food delivery platform in real time and predicts the delivery efficiency, which breaks through the limitations of traditional single-dimensional prediction. It achieves accurate prediction of delivery efficiency through full-link data collection, can identify potential delivery risks in advance, provides data support for subsequent scheduling decisions, and improves the system's foresight.

[0035] By combining the fluctuation stages of order demand with multi-dimensional influence coefficients to determine early warning scenarios, the system achieves accurate classification of delivery supply and demand status. By distinguishing different interactive relationships (influence coefficients and order volume increase / decrease trends) under low / high supply scenarios, it accurately identifies four core scenarios: low supply with no backlog, low supply with high backlog, supply and demand imbalance, and stable supply and demand. This provides accurate scenario basis for subsequent differentiated scheduling and solves the problem of insufficient targeting of traditional scheduling strategies.

[0036] The delivery monitoring platform issues corresponding dispatch instructions based on different early warning signals, realizing precise linkage between "scenario analysis and instruction issuance". For the two core problem scenarios of low supply and high backlog and supply and demand imbalance, it triggers segmented delivery and regional order allocation respectively, ensuring accurate matching between dispatch strategies and scenario needs and improving the effectiveness of dispatch.

[0037] 2. By precisely scheduling riders to fill order gaps in business districts, dividing orders into batches based on meal preparation intervals, and matching rider orders based on path overlap algorithms, the utilization rate of rider resources and order delivery efficiency have been effectively improved. The design of trajectory and time verification during order transfer avoids the accumulation of delivery delays caused by order transfer, ensuring the rationality of scheduling. The segmented delivery and order alternation mechanism further optimizes the delivery link, reduces rider waiting time, and improves the overall delivery smoothness. At the same time, the clear commission sharing rules protect rider rights and improve rider cooperation.

[0038] 3. The regional order allocation unit implements differentiated order allocation steps for supply and demand imbalance scenarios. By extracting familiar and relevant features of rider trajectories, it achieves accurate matching between orders and riders. Based on the classification of order difficulty and the allocation design of adjacent rider combinations, it optimizes rider task load and reduces rider waiting congestion at food preparation points. The consideration of the total number of orders and optimal routes during the order allocation process ensures a balance between scheduling efficiency and delivery stability, which alleviates the pressure on the food preparation end and guarantees delivery timeliness. Attached Figure Description

[0039] To facilitate understanding by those skilled in the art, the present invention will be further described below with reference to the accompanying drawings.

[0040] Figure 1 This is a system principle block diagram of the present invention;

[0041] Figure 2 This is a flowchart of the method for the multi-dimensional detection unit in this invention. Detailed Implementation

[0042] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0043] In this document, the term "embodiment" means that a particular feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of the invention. The appearance of this phrase in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a separate or alternative embodiment mutually exclusive with other embodiments. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments.

[0044] Example 1

[0045] Please see Figure 1 As shown, a real-time delivery monitoring system applied to a food delivery platform includes a delivery monitoring platform, wherein the delivery monitoring platform is connected to a multi-dimensional detection unit, a segmented delivery unit, and a regional order allocation unit.

[0046] The delivery monitoring platform generates multi-dimensional detection signals and sends them to the multi-dimensional detection unit;

[0047] Please see Figure 2 As shown, the multi-dimensional detection unit receives multi-dimensional detection signals and monitors the coverage area of ​​the food delivery platform in real time. It also predicts the delivery efficiency of the food delivery platform through multi-dimensional data collection.

[0048] Identify the coverage area of ​​the food delivery platform and mark it as the delivery area. Identify business districts within the delivery area using a map and mark them as food preparation points. At the same time, identify office buildings or residential buildings within the delivery area and mark them as food collection points.

[0049] Using a 15-minute time window, multi-dimensional data is collected from the delivery area to obtain the average delay time of elevator waiting time at each collection point. It should be noted that, to cope with the diversity of office buildings or residential buildings, the delay time is obtained by comparing adjacent time windows as the delay time of the corresponding collection point, and the average value is calculated for all collection points. The overlap time between the period when the food preparation speed increases and the period when the total number of food items waiting to be prepared increases is obtained at the food preparation point.

[0050] The duration of the fluctuation is based on the weather characteristic parameters within the road network corresponding to the meal preparation and collection points. The weather characteristic parameters are represented by the weather characteristics such as rainfall, visibility, and wind force in the area. The duration of the parameter fluctuation is represented by the duration of rainfall, visibility, and wind force.

[0051] The average delay duration, overlap duration, and parameter fluctuation duration are uniformly marked as a multi-dimensional data set. In actual scenarios, when there is no corresponding data, the corresponding time set will not have data of the corresponding type.

[0052] The system acquires real-time order demand fluctuations within the delivery area and determines the order demand at each time point. It also weights the multi-dimensional data sets to obtain multi-dimensional influence coefficients, and uses the order demand at each time point and the multi-dimensional influence coefficients to determine early warning scenarios.

[0053] If the multi-dimensional influence coefficients show an increasing trend during the floating phase, the order supply in the delivery area will decrease, and the current period will be marked as a low supply period; if the order demand does not increase during the low supply period, a low supply with no backlog signal will be generated and sent to the delivery monitoring platform; if the order demand increases during the low supply period, a low supply with high backlog signal will be generated and sent to the delivery monitoring platform.

[0054] If order demand shows a prioritizing upward trend, then the current period will be marked as a high-supply period;

[0055] If the multi-dimensional influence coefficients show a synchronous increasing trend during periods of high supply, a supply-demand imbalance signal is generated and sent to the delivery monitoring platform; if the multi-dimensional influence coefficients show a decreasing trend during periods of high supply, a supply-demand stability signal is generated and sent to the delivery monitoring platform.

[0056] If the delivery monitoring platform receives a low supply with no backlog or a stable supply and demand signal, it infers that there is a fluctuation in the supply and demand balance in the delivery area, but the delivery efficiency is within the acceptable range. If the delivery monitoring platform receives a low supply with high backlog signal, it generates a segmented delivery signal and sends it to the segmented delivery unit. If the delivery monitoring platform receives a supply and demand imbalance signal, it generates a regional order allocation signal and sends it to the regional order allocation unit.

[0057] It should be explained that if the multi-dimensional impact coefficient shows an increasing trend, it indicates that delivery interference is the main influencing factor. If the order demand shows an increasing trend, it indicates that the order supply efficiency is not timely. Although there are slight differences between the two scenarios, by processing the two scenarios differently and summarizing them into the same system, delivery monitoring can be completed efficiently.

[0058] After receiving the segmented delivery signal, the segmented delivery unit performs segmented delivery control over the delivery area;

[0059] The order supply and demand gap of each business district in the delivery area is collected in 15-minute increments; when the order supply and demand gap of a business district is higher than 20% of the set value, riders within a 5-kilometer radius of the business district are dispatched to fill the gap; the location of the food collection point is identified based on the real-time orders in the business district, and the real-time food order batches are divided with a 2-minute interval between food preparation times as the threshold, that is, orders within a 2-minute interval are uniformly planned.

[0060] Riders are grouped into batches of food delivery orders based on a path overlap algorithm, meaning the overlapping sections of delivery routes for orders received by a rider in the same batch are below a set threshold. After grouping by batch, the rider's first received order is used as the delivery location, which is recorded in the backend. The system determines the continuous pickup areas based on the rider's order history. As the rider moves towards a pickup area, the backend identifies the pickup point location of the order within that area against the rider's delivery location. If the locations match, the current order is matched with the rider; otherwise, a new order is placed with the rider currently heading to the pickup area. In single-logic analysis, riders with completed order matching are used as the dispatch end. The dispatch end collects the pickup location of the orders to be delivered and compares it with the delivery location of riders who have not completed matching. If the location matches, the orders to be delivered by the dispatch end are transferred to the remaining orders of the riders who have not completed matching. The backend directly switches the orders and replans the delivery routes of each rider. If the planned route increases the delivery time or delivery distance, the order transfer is canceled. If the location does not match, riders are continuously changed until the rider with the unmatched order enters the store, at which point the current food order is temporarily matched with the current rider.

[0061] Simultaneously, the system generates segmented delivery instructions, marks the temporarily matched rider trajectory with the corresponding food order's pickup point, and performs real-time statistics on the remaining trajectories of all riders within the delivery area. If a rider's real-time remaining trajectory overlaps with the temporarily matched rider's trajectory, and the distance between the food order's pickup point and any pickup point in the rider's real-time remaining trajectory is less than a set threshold, then the overlapping point is identified and a waiting interval is set. Riders then alternate orders at the overlapping point; and the system updates the rider information in the background.

[0062] It should be explained that in real-world scenarios, when riders plan their orders along the same route and need to deliver them in segments, the rider's actual location commission is divided based on the actual delivery distance of the current order and the total commission.

[0063] After receiving the regional order allocation signal, the regional order allocation unit analyzes the meal preparation points in the delivery area;

[0064] Orders are allocated based on the riders waiting to pick up their meals at the designated locations. Once a rider enters the store, the data on their pending delivery orders is analyzed. These pending delivery orders include orders currently waiting in the store.

[0065] The locations of the pickup points for each order to be delivered are extracted, and features of the delivery trajectory for each order are extracted based on the extracted locations. The trajectory familiarity features and trajectory matching features of the riders to be picked up are extracted. The quantitative parameters of the trajectory familiarity features are the total historical delivery overlap and the frequency of no wrong routes in the historical delivery trajectory of the rider to be picked up. The quantitative parameters of the trajectory matching features are the delivery speed and the frequency of no pauses of the rider to be picked up corresponding to the delivery type of the delivery trajectory. The delivery type is represented by walking or climbing type in the delivery trajectory.

[0066] If any quantitative parameter corresponding to the extracted features of the delivery trajectory of an order exceeds the threshold set by the corresponding quantitative parameter, the order to be delivered by the corresponding rider will be set as an easy order; if any quantitative parameter corresponding to the extracted features of the delivery trajectory of an order does not exceed the threshold set by the corresponding quantitative parameter, the order to be delivered by the corresponding rider will be set as a difficult order.

[0067] The system compares the orders awaiting pickup by riders in the store. If adjacent riders have corresponding alternating order types, the adjacent riders are set as an allocable combination, and the orders awaiting pickup become allocable orders. Alternating order types mean: a difficult order for the current rider is an easy order for the corresponding adjacent rider; or a difficult order for the corresponding adjacent rider is an easy order for the current rider. Once an allocable combination is determined, allocable orders are assigned. If the total number of orders for a rider does not decrease, the allocation combination is adjusted, and the order delivery information is modified in the backend. At this point, orders at the food preparation point are among the allocable orders. Rider selection is based on optimal delivery route analysis, and orders at the current food preparation point are assigned to the same rider. Riders who have not yet been assigned orders at their current food preparation point can directly deliver orders without waiting. This alleviates the waiting intensity at food preparation points and improves the delivery efficiency of orders awaiting pickup. It should be noted that if the total number of orders for either rider decreases when an allocable combination is determined, rider authorization is required before the combination can be determined.

[0068] The delivery monitoring platform collects and processes data from each unit to conduct overall planning and scheduling of the delivery area;

[0069] Example 2

[0070] Building upon Example 1, the final stage of the delivery trajectory is prone to positioning deviations, leading to decreased delivery efficiency. This example addresses this by incorporating a precise trajectory update unit into the delivery monitoring platform's communication connection. To resolve the pain point of positioning deviations in last-mile delivery in Example 1, the system's functional architecture has been improved, achieving precise control over all stages of the delivery trajectory and filling a gap in existing systems for last-mile delivery optimization. This ensures the timeliness and accuracy of delivery trajectory data, providing real-time data support for last-mile delivery optimization, enabling timely detection of positioning deviations and initiation of optimization processes. It also solves the problem of navigation failure in areas with weak signals, ensuring delivery continuity for riders in deviation areas through offline data support, reducing detours and positioning waiting time, and effectively improving last-mile delivery efficiency. Furthermore, the delivery method, combining regional structure and offline content display, enhances rider flexibility and autonomy in delivery.

[0071] When Example 1 is executed, the trajectory accuracy update unit in this system monitors the delivery trajectory within the delivery area in real time and updates it continuously based on the monitoring results to improve delivery efficiency within the delivery area.

[0072] Perform distribution trajectory analysis on the delivery area by analyzing real-time delivery and historical delivery logs.

[0073] The system acquires the area of ​​the indoor passageway of the building corresponding to the real-time distribution trajectory, and collects the update delay frequency between the rider's actual position and the navigation position in real time as the communication area increases. If the update delay frequency appears and shows an increasing trend, or if the rider's actual position and navigation position deviate after the update delay frequency is generated, the passageway of the corresponding distribution trajectory is marked as a deviation area; if the update delay frequency does not appear, and the rider's actual position and navigation position do not deviate after the update delay frequency is generated, the passageway of the corresponding distribution trajectory is marked as a precise area.

[0074] The system uploads data to the backend for deviation areas. Based on the rider's historical delivery trajectory records, it statistically analyzes and spatially distributes the pickup points in the corresponding deviation areas. Simultaneously, it analyzes the movement trajectories of the pickup points in the spatial distribution to improve the road distribution map within the current deviation area. By comparing and contrasting the overlapping movement trajectories, it infers the delivery trajectory of each pickup point. If there is a delay in the rider's location update in the deviation area, the system displays the backend-stored delivery trajectory and corresponding spatial distribution offline. The rider then performs deliveries based on the structure of the deviation area and the offline display content.

[0075] In summary, this invention discloses a real-time delivery monitoring system for food delivery platforms. Its core objective is to achieve dynamic monitoring and differentiated scheduling of the entire delivery chain through multi-dimensional data collection and precise scenario analysis, thereby improving delivery efficiency and the accuracy of supply and demand matching.

[0076] Thresholds, preset values, preset ranges, etc. are set for result comparison and analysis to determine whether they are good or bad. The value of these thresholds is determined by a combination of large-scale model analysis of sample data and human experience. They can also be adjusted appropriately based on seasonal or common-sense influences.

[0077] Furthermore, the settings for weighting ratios, influence factors, etc., are based on the magnitude of each parameter's influence on the results. The specific values ​​are allocated to ultimately reflect the impact on the results. The settings for input and storage are also determined by a combination of large-scale model analysis of sample data and human experience. Appropriate adjustments can also be made based on seasonal or rational influence conditions.

[0078] The preferred embodiments of the present invention disclosed above are merely illustrative of the invention. These preferred embodiments do not exhaustively describe all details, nor do they limit the invention to any specific implementation. Clearly, many modifications and variations can be made based on the content of this specification. This specification selects and specifically describes these embodiments to better explain the principles and practical applications of the invention, thereby enabling those skilled in the art to better understand and utilize the invention. The invention is limited only by the claims and their full scope and equivalents.

Claims

1. A real-time delivery monitoring system applied to a food delivery platform, characterized in that, This includes a delivery monitoring platform, whose communication connections include: The multi-dimensional detection unit monitors the coverage area of ​​the food delivery platform in real time and predicts the delivery efficiency of the food delivery platform through multi-dimensional data collection. Segmented delivery units allow for segmented delivery control within the delivery area; The regional order allocation unit performs order allocation analysis for delivery areas.

2. The real-time delivery monitoring system for a food delivery platform according to claim 1, characterized in that, The process of multi-dimensional detection unit is as follows: Identify the coverage area of ​​the food delivery platform and mark it as the delivery area. Identify business districts within the delivery area using a map and mark them as food preparation points. At the same time, identify office buildings or residential buildings within the delivery area and mark them as food collection points. Using a 15-minute time window, multi-dimensional data is collected from the delivery area to obtain the average delay time of elevator waiting time corresponding to the collection point; the overlap time between the period when the food preparation speed increases and the period when the total number of food items to be prepared increases at the food preparation point; and the duration of weather characteristic parameter fluctuations within the road network corresponding to the food preparation and collection points. The average delay time, overlap time, and duration of parameter fluctuations are then uniformly labeled as a multi-dimensional data set.

3. The real-time delivery monitoring system applied to a food delivery platform according to claim 2, characterized in that, The system acquires real-time order demand fluctuations within the delivery area and determines the order demand at each time point. It also weights the multi-dimensional data sets to obtain multi-dimensional influence coefficients, and uses the order demand at each time point and the multi-dimensional influence coefficients to determine early warning scenarios. If the multi-dimensional influence coefficients show an increasing trend during the floating phase, the order supply in the delivery area will decrease, and the current period will be marked as a low supply period; if the order demand does not increase during the low supply period, a low supply no backlog signal will be generated and sent to the delivery monitoring platform. If order demand increases during periods of low supply, a low supply and high backlog signal will be generated and sent to the delivery monitoring platform. If order demand shows an increasing trend, the current period is marked as a high-supply period; if the multi-dimensional influence coefficients show an increasing trend simultaneously during the high-supply period, a supply-demand imbalance signal is generated and sent to the delivery monitoring platform. If the multi-dimensional influence coefficients show a downward trend during periods of high supply, a supply and demand stabilization signal is generated and sent to the delivery monitoring platform.

4. The real-time delivery monitoring system applied to a food delivery platform according to claim 3, characterized in that, If the delivery monitoring platform receives a low supply with no backlog or a stable supply and demand signal, it infers that there is a fluctuation in the supply and demand balance in the delivery area, but the delivery efficiency is within the acceptable range. If the delivery monitoring platform receives a low supply with high backlog signal, it generates a segmented delivery signal and sends it to the segmented delivery unit. If the delivery monitoring platform receives a supply and demand imbalance signal, it generates a regional order allocation signal and sends it to the regional order allocation unit.

5. A real-time delivery monitoring system for a food delivery platform according to claim 4, characterized in that, The process of segmented delivery units is as follows: The order supply and demand gap of each business district in the delivery area is collected in 15-minute increments; when the order supply and demand gap of a business district is higher than 20% of the set value, riders within a 5-kilometer radius of the business district are dispatched to fill the gap; the location of the food collection point is identified based on the real-time orders in the business district, and the real-time food order batches are divided with a 2-minute interval between food preparation times as the threshold, that is, orders within a 2-minute interval are uniformly planned. The riders are divided into the same batch of food orders based on the path overlap algorithm. That is, the overlapping sections of the delivery routes of the same batch of orders received by the rider are less than a set threshold. After the order is divided into the same batch of food orders, the rider uses the first order received as the delivery location and records it in the background. The system determines the continuous pickup area based on the rider's order status in the system. When the rider moves to the pickup area, the system identifies the pickup point location of the order in the pickup area and the rider's delivery location. If the locations match, the current order is matched with the rider. If the locations are inconsistent, the system will perform backend order statistics for riders heading to the current pick-up area, use riders with completed order matching as the dispatch end, and collect the pickup location of the orders to be delivered on the dispatch end and compare it with the delivery location of riders who have not completed matching.

6. A real-time delivery monitoring system for a food delivery platform according to claim 5, characterized in that, When comparing delivery locations, if the locations match, the pending orders from the dispatch terminal will be transferred to the remaining pending orders of the riders who currently have no matching orders. The backend will directly switch orders and replan the delivery routes of each rider. If the planned route increases the delivery time or delivery distance, the order transfer will be canceled. If the locations do not match, riders will continue to be changed until the rider with no matching orders enters the store. Then, the current food order will be temporarily matched with the current rider.

7. A real-time delivery monitoring system for a food delivery platform according to claim 6, characterized in that, The system generates segmented delivery instructions, marks the temporarily matched rider trajectory with the corresponding food order's pickup point, and simultaneously performs real-time statistics on the remaining trajectories of all riders within the delivery area. If a rider's real-time remaining trajectory overlaps with the temporarily matched rider's trajectory, and the distance between the food order's pickup point and any pickup point in the rider's real-time remaining trajectory is less than a set threshold, then the overlapping point is identified, and a waiting interval is set. Riders then alternate orders at the overlapping point; and the system updates the rider information in the background.

8. A real-time delivery monitoring system for a food delivery platform according to claim 1, characterized in that, The process of allocating regional orders to units is as follows: Orders are allocated based on the riders waiting to pick up their meals at the designated locations. Once a rider enters the store, the data on their pending delivery orders is analyzed. These pending delivery orders include orders currently waiting in the store. The locations of the pickup points for the orders to be delivered are extracted, and the features of the delivery trajectory for each order to be delivered are extracted based on the extracted locations. The trajectory familiarity features and trajectory matching features of the riders waiting to pick up the food are also extracted. If any quantitative parameter corresponding to the delivery trajectory extraction feature of the order to be delivered exceeds the corresponding quantitative parameter setting threshold, the order to be delivered for the corresponding rider will be set as an easy order; if any quantitative parameter corresponding to the delivery trajectory extraction feature of the order to be delivered does not exceed the corresponding quantitative parameter setting threshold, the order to be delivered for the corresponding rider will be set as a difficult order.

9. A real-time delivery monitoring system for a food delivery platform according to claim 8, characterized in that, The system compares the pending orders of riders waiting to pick up food in the store. If the order types of adjacent riders are correspondingly alternating, the corresponding adjacent riders are set as an allocable combination, and the pending orders become allocable orders. Once the allocable combination is determined, the allocable orders are assigned. If the total number of orders for a rider does not decrease, the allocation combination is adjusted, and the order delivery information is modified in the background. At this time, if the order at the food preparation point is among the allocable orders, the rider is selected based on the optimal delivery route analysis, and the current food preparation point order is assigned to the same rider. Riders who have executed the order allocation combination but have not been assigned to the current food preparation point order will be delivered directly without waiting.

10. A real-time delivery monitoring system for a food delivery platform according to claim 9, characterized in that, The quantitative parameters for trajectory familiarity features are the total historical delivery overlap and the frequency of no wrong routes in the historical delivery for the rider to be picked up; the quantitative parameters for trajectory fit features are the delivery speed and the frequency of no pauses for the rider to be picked up corresponding to the delivery type of the delivery trajectory. The delivery type is represented by the walking type or the stair climbing type in the delivery trajectory.

Images