A dynamic supply chain warehouse intelligent order splitting method

Abstract

This invention discloses a dynamic, intelligent order splitting method for supply chain warehouses. The method comprises the following steps: acquiring historical order data; establishing a mathematical model with the goal of minimizing costs and constraints including carrier uniqueness, timeliness, and order splitting ratios; and determining the carriers assigned to historical orders of the same dimension, which serve as the basis for assigning carriers to new orders. This method is computationally efficient and easy to use, and can be easily expanded and adjusted by subsequently considering more information, such as network capacity and line capacity.

Description

Technical Field

[0001] The present invention relates to supply chain warehouse logistics management, and more specifically, to a dynamic supply chain warehouse intelligent order splitting method. Background Art

[0002] The supply chain warehouse receives multiple orders from customers and sends them to various parts of the country. Without specifying a carrier, choosing the best carrier for each order is an important decision.

[0003] Each carrier has a different quotation system, and each has its own price advantages in different weight ranges and on different routes. The service quality of different carriers, such as the timeliness of order receipt, also has regional differences.

[0004] For orders received by the supply chain warehouse in real time, it is necessary to promptly assign appropriate carriers so that indicators such as the timeliness of final order receipt meet customer requirements.

[0005] The current method of order splitting is manual, based on multiple dimensions such as timeliness, cost, and other rules. Due to objective factors, order splitting based on timeliness is primarily used to improve operational quality. Timeliness splitting involves manually pulling a week or two-week historical warehouse receipt data for the corresponding project from a set of available carriers, aggregating and calculating timeliness. This timeliness data is then analyzed and compared on a city-by-city basis, and carriers are ranked by timeliness. The final order split is then determined based on a month's worth of historical orders, carrier costs, capacity, maximum share limits, and the principle of balancing remote areas.

[0006] The current manual order splitting method is labor-intensive, tedious, and repetitive, with limited data and insufficient accuracy. It also wastes significant manpower and time, and fails to meet customer needs in a timely manner. For example, there are approximately 330 prefecture-level cities, and each carrier must be compared. With order splitting now being refined down to the district and county level, manual labor is no longer sufficient to meet all of these needs.

[0007] In terms of cost dimension, personal subjectivity plays a big role in order splitting or it is not accurate enough. Orders are basically split in a one-size-fits-all manner without much consideration of cost.

[0008] As the customer base grows, the demands on delivery carriers are becoming more differentiated and demanding. Manual order splitting is a sluggish and limited method that can hardly fully meet customer needs.

[0009] Therefore, it is necessary to provide a dynamic supply chain warehouse intelligent order splitting method that selects a carrier from the optional carrier combination for orders under each route (warehouse-destination city) and each weight range while ensuring timeliness requirements, so as to minimize the total cost. Summary of the Invention

[0010] The purpose of the present invention is to provide a dynamic supply chain warehouse intelligent order splitting method to solve the above problems.

[0011] In order to achieve the above object, the present invention adopts the following technical solutions:

[0012] A dynamic supply chain warehouse intelligent order splitting method, comprising the following steps:

[0013] Get historical order data information;

[0014] Establish a mathematical model with the goal of minimizing costs, and constraints including carrier uniqueness, on-time rate constraints, and order splitting ratio constraints;

[0015] The carriers assigned to historical orders with the same dimensions are determined as the basis for assigning carriers to new orders.

[0016] Preferably, the historical order data information includes route, weight range, carrier, actual receipt time, expected receipt time and quotation list.

[0017] Preferably, the order classification dimensions include route and weight range.

[0018] Preferably, the carrier unique constraint is implemented using a 0-1 planning method.

[0019] Preferably, the mathematical model is

[0020]

[0021] in,

[0022] d is the order classification dimension;

[0023] p is the carrier;

[0024] c is the customer;

[0025] is the quote from carrier p for the order with dimension d;

[0026] The total number of orders for customer c with dimension d in the past month;

[0027] It is a 0-1 binary variable, where y=1 means that the order with dimension d is picked up by carrier p, otherwise it is not picked up;

[0028] The dimension is d, the on-time signing rate of carrier p;

[0029] The average signing-on-time rate required by customers;

[0030] is the total order quantity of the warehouse;

[0031] is the total order quantity of customer c;

[0032] The upper limit of the proportion of orders that the warehouse can allocate to carrier p;

[0033] The lower limit of the proportion of orders allocated by the warehouse to carrier p.

[0034] Preferably, the method further comprises adjusting the timely receipt rate in real time within the cycle, and calculating the timely receipt rate for the remaining days within the cycle based on the timely rates of the first t days within the cycle and the timely rate constraint.

[0035] Preferably, the calculation of the on-time receipt rate for the remaining days in the period based on the on-time rate of the first t days in the period and the on-time rate constraint includes:

[0036] make

[0037] Right now

[0038] Deformation

[0039]

[0040] Get a new timely receipt rate

[0041] Where T is the period;

[0042] The number of orders from customer c that were fulfilled t days ago;

[0043] The number of orders that can be fulfilled within Tt days after the customer's order is c;

[0044] is the number of customer orders from the previous t days;

[0045] The future Tt-day order volume for customer c with dimension d;

[0046] is the actual on-time rate of orders placed by customer c at the current time t.

[0047] Preferably, the method also includes real-time adjustment of the order splitting ratio parameters. If the proportion of orders of a certain carrier on that day is greater than the upper limit of the order splitting ratio limit, the corresponding upper limit of the order splitting ratio in the mathematical model will be reduced accordingly; if the proportion of orders of a certain carrier on that day is less than the lower limit of the order splitting ratio limit, the corresponding lower limit of the order splitting ratio in the mathematical model will be increased accordingly.

[0048] Preferably, the real-time adjustment of the order splitting ratio parameters includes the calculation of the order volume proportion of each carrier on that day. , compared with the upper and lower limits of the order quantity ratio allocated by the initialized warehouse to the carrier p,

[0049] like ,but = , ,

[0050] like ,but = , ,

[0051] in, 、 is the set experience value parameter, The upper limit of the proportion of orders allocated to carrier p by the initialized warehouse, The lower limit of the proportion of orders allocated to carrier p by the initialized warehouse.

[0052] Preferably, if the split ratio parameter appears in the historical calculation process of the day, the calculation result of the same dimension is directly used without recalculating the plan.

[0053] The beneficial effects of the present invention are as follows:

[0054] This invention provides timely order splitting for customers. The system can adjust order splitting based on the quality of receipt data, avoiding timeliness issues and improving overall delivery operational quality. Furthermore, it implements intelligent and automated order splitting, improving work efficiency, ensuring operational quality, saving labor costs, reducing freight costs, and increasing profits. Furthermore, this method boasts high computational efficiency and ease of use, and can be easily expanded and adjusted by considering additional information, such as network capacity and line capacity. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] The specific embodiments of the present invention will be further described in detail below with reference to the accompanying drawings.

[0056] Figure 1 A flow chart of a dynamic supply chain warehouse intelligent order splitting method is shown. DETAILED DESCRIPTION

[0057] In order to more clearly illustrate the present invention, the present invention is further described below in conjunction with preferred embodiments and accompanying drawings. Similar components in the accompanying drawings are represented by the same reference numerals. It should be understood by those skilled in the art that the following detailed description is illustrative rather than restrictive and should not be used to limit the scope of protection of the present invention.

[0058] like Figure 1As shown, the present invention provides a dynamic supply chain warehouse intelligent order splitting method, the steps comprising:

[0059] Get historical order data information;

[0060] Establish a mathematical model with the goal of minimizing costs, and constraints including carrier uniqueness, on-time rate constraints, and order splitting ratio constraints;

[0061] The carriers assigned to historical orders with the same dimensions are determined as the basis for assigning carriers to new orders.

[0062] Specifically, selecting the carrier with the best cost for orders of different dimensions can reduce costs and increase revenue as much as possible, and can also meet the kilogram-segment transportation needs of carriers of specific commodities.

[0063] In an optional implementation, the historical order data information includes routes, weight ranges, carriers, actual receipt time, expected receipt time, and quotation list.

[0064] In an optional implementation, the order classification dimensions include route and weight range.

[0065] In an optional implementation, a 0-1 planning method is used to implement the carrier unique constraint.

[0066] In an optional implementation, the mathematical model is

[0067]

[0068] in,

[0069] d is the order classification dimension;

[0070] p is the carrier;

[0071] c is the customer;

[0072] is the quote from carrier p for the order with dimension d;

[0073] The total number of orders for customer c with dimension d in the past month;

[0074] It is a 0-1 binary variable, where y=1 means that the order with dimension d is picked up by carrier p, otherwise it is not picked up;

[0075] The dimension is d, the on-time signing rate of carrier p;

[0076] The average signing-on-time rate required by customers;

[0077] is the total order quantity of the warehouse;

[0078] is the total order quantity of customer c;

[0079] The upper limit of the proportion of orders that the warehouse can allocate to carrier p;

[0080] The lower limit of the proportion of orders allocated by the warehouse to carrier p.

[0081] In an optional implementation, the method further includes adjusting the timely receipt rate in real time within the cycle, and calculating the timely receipt rate for the remaining days within the cycle based on the timely rates of the first t days within the cycle and the timely rate constraint.

[0082] Specifically, the service quality of different carriers, such as their order receipt timeliness, also varies regionally. To provide customers with optimal operational quality, each carrier's expected on-time delivery rate is a key consideration when selecting a carrier. In practice, we set timeliness standards for each customer. If order splitting results fail to meet these standards, there's a risk of losing customers.

[0083] In an optional implementation, the calculation of the on-time receipt rate for the remaining days in the period based on the on-time rate of the first t days in the period and the on-time rate constraint includes:

[0084] make

[0085] Right now

[0086] Deformation

[0087]

[0088] Get a new timely receipt rate

[0089] Where T is the period;

[0090] The number of orders from customer c that were fulfilled t days ago;

[0091] The number of orders that can be fulfilled within Tt days after the customer's order is c;

[0092] is the number of customer orders from the previous t days;

[0093] The future Tt-day order volume for customer c with dimension d;

[0094] is the actual on-time rate of orders placed by customer c at the current time t.

[0095] In an optional implementation, the method also includes real-time adjustment of the order splitting ratio parameters. If the proportion of orders of a certain carrier on that day is greater than the upper limit of the order splitting ratio limit, the corresponding upper limit of the order splitting ratio in the mathematical model is reduced accordingly; if the proportion of orders of a certain carrier on that day is less than the lower limit of the order splitting ratio limit, the corresponding lower limit of the order splitting ratio in the mathematical model is increased accordingly.

[0096] Specifically, order splitting needs to ensure that each carrier receives a certain volume to avoid dissatisfaction with partner carriers due to too few orders, which could negatively impact the partnership. Furthermore, each carrier has a limited daily capacity, so a daily cap is required to prevent a single carrier from overloading its capacity and potentially causing a warehouse overflow. While maintaining these upper and lower limits, we aim to maximize allocation to carriers with lower costs and higher expected timeliness.

[0097] In an optional implementation, the real-time adjustment of the order splitting ratio parameter includes calculating the order volume ratio of each carrier on that day. , compared with the upper and lower limits of the order quantity ratio allocated by the initialized warehouse to the carrier p,

[0098] like ,but = , ,

[0099] like ,but = , ,

[0100] in, 、 is the set experience value parameter, The upper limit of the proportion of orders allocated to carrier p by the initialized warehouse, The lower limit of the proportion of orders allocated to carrier p by the initialized warehouse.

[0101] In an optional implementation, if the split ratio parameter appears in the historical calculation process of the day, the calculation result of the same dimension is directly used without recalculating the solution.

[0102] In one specific embodiment of the present invention, to ensure timely carrier assignment for orders, historical data for similar orders, such as those on the same route or within the same weight range, is used for comparison when order information cannot be predicted. Assuming stable order volume along these dimensions, and the on-time delivery rates and quotes of each carrier remaining largely unchanged, recent historical data can be obtained to determine the carrier assigned to each order type, with optimal cost as the goal, while ensuring timeliness and order splitting ratios. This data then serves as the basis for assigning carriers to new orders.

[0103] In this embodiment, under the premise of ensuring timeliness requirements, a carrier is selected from the optional carrier combination for orders under each route (warehouse-destination city) and each weight range to minimize the total cost.

[0104] Specifically, historical order data is pulled from Best Inc.'s order management system (OMS) to obtain information such as the order's actual carrier, optional carrier combinations and quotes, actual receipt time, and expected receipt time, as shown in Table 1.

[0105] Collect statistics on the historical on-time receipt rate of each carrier for the same route and order weight range, as well as the sum of order quotations.

[0106] Timely receipt rate The number of orders promptly signed for by carrier p in dimension d (order classification dimension, including route, weight range, etc.) / the total number of historical orders by carrier p in dimension d.

[0107] Table 1 Historical order receipt information table

[0108]

[0109] Use recent historical order data with the goal of minimizing costs. Constraints include using a single carrier, on-time rate constraints, and order splitting ratio constraints.

[0110] To decide which carrier to assign an order to, a 0-1 binary variable is set to indicate whether the order under the same dimension d is assigned to carrier p, and an integer programming model is established. The specific model is as follows.

[0111] Mathematical model

[0112] min:

[0113] st

[0114] (1)

[0115] (2)

[0116] (3)

[0117] , (4)

[0118] in

[0119] d: Order classification dimension, including route, weight range, etc.

[0120] p: carrier.

[0121] c: customer.

[0122] : Represents the quote from carrier p for an order with dimension d.

[0123] : Indicates the total number of orders for customer c with dimension d in the past month.

[0124] : 0-1 binary variable, y=1 means that the order with dimension d is picked up by carrier p, otherwise it is not picked up.

[0125] : Dimension d, carrier p’s on-time receipt rate.

[0126] : The average timely receipt rate required by customers.

[0127] : The total number of orders in the warehouse.

[0128] : Total order quantity of customer c.

[0129] : The upper limit of the proportion of orders that the warehouse allocates to carrier p.

[0130] : The lower limit of the proportion of orders allocated by the warehouse to carrier p.

[0131] In this embodiment, since the model is based on historical data, changes in route order volume and timeliness may result in failure to achieve the expected timeliness and order splitting ratio in future operations. Therefore, a dynamic adjustment strategy is considered:

[0132] Real-time adjustment of the split order ratio parameter: If the order volume of a carrier on the day does not meet the ratio limit (upper / lower limit), reduce / increase the upper / lower limit of the corresponding split order ratio in the model ( ). Before calculation, initialize and , and then calculate the proportion of orders of each carrier on that day , with the initialized and In contrast, if ,but = , ,if ,but = , , 、 It is an empirical value parameter and can be set based on historical data analysis.

[0133] Adjust parameter signing timeliness rate : Assuming that the timeliness required by customers (timely receipt rate) is calculated based on the dimension of a natural month, the number of orders that need to meet the customer's timeliness requirements this month should be adjusted according to the number of orders that have already occurred in that month. That is, if the actual timely receipt rate from the beginning of the month to the current day is relatively low, assuming that the order volume remains unchanged, the number of orders that need to meet customer timeliness requirements in the future will have to increase.

[0134] Timely receipt rate Adjustment strategy: Based on the assumption that historical data can be used to represent the data for this month, the real-time receipt timeliness rate from the beginning of the month to the current day (day t) is brought into the first t days of the historical month to obtain the number of orders that met the timeliness requirement in the first t days. Then, we only need to calculate the solution for the next Tt days under the current situation to make the total number of orders that meet the timeliness requirement greater than the target value. :

[0135] (5)

[0136] : The number of orders from customer c that reached a deadline t days ago

[0137] : The number of orders that can be fulfilled within Tt days after customer c

[0138] Assuming that the daily order volume is balanced, the customer order volume in the previous t days can be expressed as , the number of orders in each dimension in the next Tt days will be .

[0139] Get the actual on-time rate of customer c's orders that have occurred at the current time (day) t this month , It can be expressed as , the number of orders that will be completed in Tt days in the future can be expressed by scaling the order quantity using formula (2), which is , so formula (5) can be expressed as follows.

[0140] (6)

[0141] T: total number of days in this month

[0142] C: The collection of customers

[0143] Through deformation we can get:

[0144]

[0145] Get a new timely receipt rate

[0146] Plan Calculation Frequency: Based on actual conditions, you can recalculate the plan using the latest parameters at regular intervals (e.g., hourly) and use this plan until the next calculation. The order splitting ratio parameter can be consistent with the plan calculation frequency (e.g., hourly), and the on-time receipt rate parameter can be updated daily. During each plan calculation, if the order splitting ratio parameter has appeared in previous calculations for the day, the plan can be recalculated without recalculation; the results for the same dimensions can be used directly.

[0147] This method is computationally efficient and easy to use, and can be easily expanded and adjusted if more information is subsequently considered, such as network capacity, line capacity, etc.

[0148] Obviously, the above embodiments of the present invention are merely examples for clearly illustrating the present invention, and are not limitations on the implementation methods of the present invention. For ordinary technicians in the relevant field, other different forms of changes or modifications can be made based on the above description. It is impossible to list all the implementation methods here. All obvious changes or modifications derived from the technical solution of the present invention are still within the scope of protection of the present invention.

Claims

1. A dynamic supply chain warehouse intelligent order splitting method, characterized by the following steps: include: Get historical order data information; Establish a mathematical model with the goal of minimizing costs, and constraints including carrier uniqueness, on-time rate constraints, and order splitting ratio constraints; Determine the carriers assigned to historical orders with the same dimensions as the basis for assigning carriers to new orders; The mathematical model is in, d is the order classification dimension; p is the carrier; c is the customer; is the quote from carrier p for the order with dimension d; The total number of orders for customer c with dimension d in the past month; It is a 0-1 binary variable, where y=1 means that the order with dimension d is picked up by carrier p, otherwise it is not picked up; The dimension is d, the on-time signing rate of carrier p; The average signing-on-time rate required by customers; is the total order quantity of the warehouse; is the total order quantity of customer c; The upper limit of the proportion of orders that the warehouse can allocate to carrier p; The lower limit of the proportion of orders allocated by the warehouse to carrier p.

2. The method according to claim 1, characterized in that The historical order data information includes routes, weight ranges, carriers, actual receipt time, expected receipt time and quotation sheet.

3. The method according to claim 1, characterized in that Order classification dimensions include route and weight range.

4. The method according to claim 1, characterized in that The carrier unique constraint is implemented using a 0-1 planning method.

5. The method according to claim 1, characterized in that: The method further includes adjusting the timely receipt rate in real time within the cycle, and calculating the timely receipt rate for the remaining days within the cycle based on the timely rates of the first t days within the cycle and the timely rate constraint.

6. The method according to claim 5, characterized in that The calculation of the on-time receipt rate for the remaining days in the period based on the on-time rate of the first t days in the period and the on-time rate constraint includes: make Right now Deformation Get a new timely receipt rate Where T is the period; The number of orders from customer c that were fulfilled t days ago; The number of orders that can be fulfilled within Tt days after the customer's order is c; is the number of customer orders from the previous t days; The future Tt-day order volume for customer c with dimension d; is the actual on-time rate of orders placed by customer c at the current time t.

7. The method according to claim 1, characterized in that: The method also includes adjusting the order splitting ratio parameter in real time. If the order volume of a carrier on that day exceeds the upper limit of the order splitting ratio, the corresponding order splitting ratio upper limit in the mathematical model is correspondingly reduced. If the proportion of orders of a certain carrier on that day is less than the lower limit of the order splitting ratio, the corresponding lower limit of the order splitting ratio in the mathematical model will be increased accordingly.

8. The method according to claim 7, characterized in that: The real-time adjustment of the order splitting ratio parameters includes statistics on the order volume share of each carrier on that day. , compared with the upper and lower limits of the order quantity ratio allocated by the initialized warehouse to the carrier p, like ,but = , , like ,but = , , in, 、 is the set experience value parameter, The upper limit of the proportion of orders allocated to carrier p by the initialized warehouse, The lower limit of the proportion of orders allocated to carrier p by the initialized warehouse.

9. The method according to claim 1, characterized in that: If the split ratio parameter appears in the historical calculation process of the day, the calculation result of the same dimension will be directly used without recalculating the plan.

Images