Logistics order distribution method, device and electronic equipment

Abstract

The present application provides a logistics order allocation method, device and electronic device, the method comprising: using historical logistics orders to solve a pre-built optimal matching algorithm model to obtain the order splitting results of the historical logistics orders; using the order splitting results to run with the real-time logistics orders in the online environment to obtain the running results; judging whether the running results meet the constraints of the optimal matching algorithm model; if so, pushing the optimal matching algorithm model to the online environment, so that the optimal matching algorithm model can optimize and allocate future real-time logistics orders in the online environment. In the implementation process of the above scheme, the logistics order allocation method based on real data, real-time logistics orders and optimal matching algorithm model is more scientific and objective than manually maintaining weights, and effectively avoids the deviation caused by personal experience or subjective judgment. Therefore, it can achieve the maximization effect of logistics resource utilization under constraints.

Description

Technical Field

[0001] The present application relates to the technical field of operations research, supply chain management, and logistics order optimization, and in particular, to a logistics order allocation method, device, and electronic equipment. Background Art

[0002] In the modern logistics industry, the optimal allocation of orders is a key link to improve operational efficiency and service quality, while the traditional logistics order allocation method usually relies on manually maintained weights to determine the priority of carriers. Specifically, business personnel will configure corresponding allocation strategies and weights for different carriers based on multiple factors (such as timeliness, cost, consignment attributes, cooperation agreements, etc.). When an order is confirmed, a set of candidate transportation service products is screened out based on the specific characteristics of the order (such as price range, weight range, whether it is charged, whether it is air embargoed, etc.). In this result set, each transportation service product can theoretically carry the order, and the final selection is usually based on the highest weight among multiple pre-set weights. The setting of such weights often relies on the experience and judgment of business personnel, lacks scientificity and objectivity, and is easily affected by personal bias. In addition, the market environment and customer needs often change, but the manually maintained weights are updated less frequently and are difficult to adapt to these changes in a timely manner. Therefore, it is difficult to maximize the utilization of logistics resources under constraints by currently relying on manually maintained weights to allocate logistics orders. Summary of the invention

[0003] The purpose of the embodiments of the present application is to provide a logistics order allocation method, device and electronic device, which are used to improve the problem that it is difficult to maximize the utilization of logistics resources under constraints when allocating logistics orders.

[0004] The embodiment of the present application provides a logistics order allocation method, including: using historical logistics orders to solve a pre-built optimal matching algorithm model to obtain the order splitting results of the historical logistics orders; using the order splitting results to run with real-time logistics orders in an online environment to obtain the running results; judging whether the running results meet the constraints of the optimal matching algorithm model; if so, pushing the optimal matching algorithm model to the online environment, so that the optimal matching algorithm model can optimize and allocate future real-time logistics orders in the online environment. In the implementation process of the above scheme, by using historical logistics orders to solve the pre-built optimal matching algorithm model, it can be ensured that the performance of the model in actual scenarios is more reliable, so that the model can learn and adapt to different actual situations, and using real-time logistics orders for running verification can further verify the performance of the model in the current market environment, which ensures that the model not only performs well on historical data, but also can cope with the latest market changes. This logistics order allocation method based on real data, real-time logistics orders and optimal matching algorithm model is more scientific and objective than manually maintained weights, and effectively avoids deviations caused by personal experience or subjective judgment. Therefore, when the accompanying results meet the constraints and are pushed to the actual business production environment, it can achieve the effect of maximizing the utilization of logistics resources under the constraints.

[0005] Optionally, in an embodiment of the present application, the total transportation cost of the transportation service product is settled in a freight ladder manner; before using historical logistics orders to solve the pre-built optimal matching algorithm model, it also includes: using a piecewise linear function to model the freight ladder of the transportation service product to obtain a first linear expression, and the first linear expression is used to construct the optimal matching algorithm model. In the implementation process of the above scheme, the nonlinear problem is converted into a linear problem through a piecewise linear function. Therefore, the piecewise linear function decomposes the complex nonlinear relationship into multiple linear segments, which reduces the computational complexity and makes the model more efficient in solving large-scale data sets. The model can be solved using an efficient linear programming solver. The linear programming solver usually has a faster solution speed and better solution stability. Compared with the nonlinear programming solver, the solution efficiency can be significantly improved.

[0006] Optionally, in an embodiment of the present application, the total transportation cost of the transportation service product is settled in a step-by-step manner based on the order quantity; before using historical logistics orders to solve the pre-built optimal matching algorithm model, it also includes: using a nonlinear function to perform step-by-step modeling of the transportation service product to obtain a nonlinear expression, and the nonlinear expression is used to construct the optimal matching algorithm model. In the implementation process of the above scheme, the complex relationship between order quantity and freight can be more accurately captured through nonlinear functions, especially in the presence of nonlinear discounts, step-by-step pricing, etc. This precise modeling helps to better reflect the actual business situation, and the nonlinear function allows the model designer to more flexibly define the relationship between freight and order quantity, and can more finely adjust the model parameters to adapt to different business needs, thereby effectively improving the accuracy of modeling.

[0007] Optionally, in an embodiment of the present application, the nonlinear expression includes nonlinear terms; after obtaining the nonlinear expression, it also includes: after introducing auxiliary variables and inequality constraints in the nonlinear expression, eliminating the nonlinear terms to obtain a second linear expression, and the second linear expression and the inequality constraints are used to construct an optimal matching algorithm model. In the implementation process of the above scheme, complex nonlinear problems are simplified to linear problems through linearization processing, which reduces the computational complexity and makes the model more efficient in solving large-scale data sets. In addition, auxiliary variables and inequality constraints are introduced to convert nonlinear expressions into linear expressions, so that the model can be solved using an efficient linear programming solver. Linear programming solvers usually have faster solution speeds and better solution stability, and can significantly improve the efficiency of solving compared to nonlinear programming solvers.

[0008] Optionally, in an embodiment of the present application, the optimal matching algorithm model includes: multiple decision variables; before using historical logistics orders to solve the pre-built optimal matching algorithm model, it also includes: for each decision variable in the multiple decision variables, judging whether the decision variable is a known variable at the current moment; if so, presetting the decision variable to a fixed value. In the implementation process of the above scheme, after fixing some variables, the structure of the model becomes simpler, the complexity of the solution is reduced, and the solution process is more stable and faster, and by pre-setting the known decision variables to fixed values, the number of variables in the model can be significantly reduced. This makes the solver more efficient in processing the model, thereby reducing calculation time and improving the speed and efficiency of the solution.

[0009] Optionally, in an embodiment of the present application, historical logistics orders are used to solve a pre-built optimal matching algorithm model, including: obtaining all interval combinations of discount variables, and re-arranging all interval combinations in reverse order to obtain the interval combinations after reversing; in the process of traversing each interval combination in the interval combinations after reversing, judging whether the optimal matching algorithm model is successfully solved to obtain a feasible solution using historical logistics orders, and the feasible solution is an interval combination that satisfies the constraints of the optimal matching algorithm model; if the optimal matching algorithm model is successfully solved to obtain a feasible solution using historical logistics orders, then the order splitting ratio or order splitting quantity corresponding to the feasible solution is determined as the order splitting result. In the implementation process of the above scheme, by re-arranging all interval combinations in reverse order, interval combinations with higher discounts can be given priority, the search range can be gradually narrowed, and the solution efficiency can be improved. Furthermore, by traversing all possible interval combinations (such as freight intervals and unit quantity intervals), the model can be ensured to comprehensively consider various situations, ensure that the global optimal solution is found, and avoid the limitations of local optimal solutions. In addition, the method of reverse rearrangement and step-by-step screening makes the model more stable when processing different data and scenarios, reducing model instability caused by data fluctuations.

[0010] The embodiment of the present application also provides a logistics order allocation device, including: an order splitting result acquisition module, which is used to use historical logistics orders to solve a pre-built optimal matching algorithm model to obtain the order splitting results of the historical logistics orders; a running-accompanying result acquisition module, which is used to use the order splitting results to run-accompanying real-time logistics orders in an online environment to obtain running-accompanying results; a running-accompanying result judgment module, which is used to judge whether the running-accompanying results meet the constraints of the optimal matching algorithm model; an algorithm model pushing module, which is used to push the optimal matching algorithm model to the online environment if the constraints of the optimal matching algorithm model are met, so that the optimal matching algorithm model can optimize and allocate future real-time logistics orders in the online environment.

[0011] Optionally, in an embodiment of the present application, the logistics order allocation device also includes: a cost ladder modeling module, which is used to use a piecewise linear function to perform freight ladder modeling on the transportation service product to obtain a first linear expression, and the first linear expression is used to construct an optimal matching algorithm model.

[0012] Optionally, in an embodiment of the present application, the logistics order distribution device also includes: a single quantity ladder modeling module, which is used to use a nonlinear function to perform single quantity ladder modeling on the transportation service product to obtain a nonlinear expression, and the nonlinear expression is used to construct an optimal matching algorithm model.

[0013] Optionally, in an embodiment of the present application, the nonlinear expression includes nonlinear terms; the logistics order distribution device also includes: a nonlinear term elimination module, which is used to eliminate the nonlinear terms after introducing auxiliary variables and inequality constraints into the nonlinear expression to obtain a second linear expression. The second linear expression and the inequality constraints are used to construct an optimal matching algorithm model.

[0014] Optionally, in an embodiment of the present application, the optimal matching algorithm model includes: multiple decision variables; the logistics order distribution device also includes: a decision variable judgment module, used to judge whether each decision variable among the multiple decision variables is a known variable at the current moment; a variable fixed setting module, used to pre-set the decision variable to a fixed value if the decision variable is a known variable at the current moment.

[0015] Optionally, in an embodiment of the present application, an order splitting result obtaining module includes: a combination reverse order rearrangement submodule, used to obtain all interval combinations of discount variables, and reverse order rearrange all interval combinations to obtain interval combinations after reverse order; an interval combination judgment submodule, used to judge whether a feasible solution is successfully solved for the optimal matching algorithm model using historical logistics orders during the traversal of each interval combination in the interval combinations after reverse order, and a feasible solution is an interval combination that satisfies the constraints of the optimal matching algorithm model; an order splitting result determination submodule, used to determine the order splitting ratio or order splitting quantity corresponding to the interval combination as the order splitting result if a feasible solution is successfully solved for the optimal matching algorithm model using historical logistics orders.

[0016] An embodiment of the present application further provides an electronic device, including: a processor and a memory, wherein the memory stores machine-readable instructions executable by the processor, and the machine-readable instructions execute the method described above when executed by the processor.

[0017] An embodiment of the present application further provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the method described above is executed.

[0018] The embodiment of the present application also provides a computer program product, including: a computer program or a computer instruction, and the computer program or the computer instruction executes the method described above when executed by a processor. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings required for use in the embodiments of the present application will be briefly introduced below. It should be understood that the following drawings only show certain embodiments of the embodiments of the present application and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without creative work.

[0020] Figure 1 A flow chart of a logistics order allocation method provided in an embodiment of the present application is shown;

[0021] Figure 2 A schematic diagram of the structure of a logistics order distribution device provided in an embodiment of the present application is shown;

[0022] Figure 3 A schematic diagram of the structure of an electronic device provided in an embodiment of the present application is shown. DETAILED DESCRIPTION

[0023] To make the purpose, technical scheme and advantages of the embodiment of the present application clearer, the technical scheme in the embodiment of the present application will be clearly and completely described below in conjunction with the drawings in the embodiment of the present application. It should be understood that the drawings in the embodiment of the present application only serve the purpose of explanation and description, and are not used to limit the protection scope of the embodiment of the present application. The flowchart used in the embodiment of the present application shows the operations implemented according to some embodiments of the embodiment of the present application. It should be understood that the operations of the flowchart can be implemented out of sequence, and the steps without logical context can be reversed in order or implemented simultaneously. In addition, those skilled in the art can add one or more other operations to the flowchart under the guidance of the content of the embodiment of the present application, or remove one or more operations from the flowchart.

[0024] In addition, the described embodiments are only a part of the embodiments of the present application, rather than all embodiments. The components of the embodiments of the present application usually described and shown in the accompanying drawings can be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present application provided in the accompanying drawings is not intended to limit the scope of the embodiments of the present application claimed, but merely represents the selected embodiments of the embodiments of the present application.

[0025] It is understandable that the "first" and "second" in the embodiments of the present application are used to distinguish similar objects. Those skilled in the art will understand that the words "first" and "second" do not limit the quantity and execution order, and the words "first" and "second" do not necessarily limit the difference. In the description of the embodiments of the present application, the term "and / or" is merely a description of the association relationship of associated objects, indicating that there may be three relationships, such as A and / or B, which can represent: A exists alone, A and B exist at the same time, and B exists alone. In addition, the character " / " in this article generally indicates that the related objects before and after are in an "or" relationship. The term "multiple" refers to more than two (including two), and similarly, "multiple groups" refers to more than two groups (including two groups).

[0026] Before introducing the logistics order allocation method provided in the embodiment of the present application, some concepts involved in the embodiment of the present application are first introduced:

[0027] Optimal Assignment refers to the allocation of orders to the most suitable logistics carrier (and its transportation service products) through intelligent algorithms to achieve the business goal of minimizing total freight costs while ensuring the constraints of service quality grading.

[0028] Companionship refers to generating instant companionship results in a simulation environment after accessing real-time logistics orders in an online environment (also known as an online business production environment), but without intervening in the actual online results, that is, without affecting the actual online results. Its main purpose is to verify whether the companionship results meet the constraints of the optimal matching algorithm model by comparing the actual online results with the instant companionship results. If the conditions are met, it means that the optimal matching algorithm model meets the usage conditions of the online environment, and the optimal matching algorithm model can be pushed to the online environment.

[0029] It should be noted that the logistics order allocation method provided in the embodiment of the present application can be executed by an electronic device, where the electronic device refers to a device terminal or server with the function of executing a computer program, such as a smart phone, a personal computer, a tablet computer, a personal digital assistant or a mobile Internet device. A server refers to a device that provides computing services through a network, such as an x86 server and a non-x86 server, and a non-x86 server includes a mainframe, a minicomputer and a UNIX server.

[0030] The following is an example of an application scenario in which the logistics order allocation method is applicable: Assume that an e-commerce platform needs to send a large number of goods to the corresponding buyer users after receiving a reminder message that a large number of buyer users have paid for the shopping orders. In this case, the transportation routes of these large quantities of goods (one route is from the origin city to the destination city) are different, and the speeds and prices of different carriers (such as standard express, special express, land transportation, etc.) are also different. In the case where a large number of goods need to be transported by multiple carriers, the logistics order allocation method can be used to solve and optimize the pre-built optimal allocation algorithm model based on historical logistics orders. The optimal allocation algorithm model can make the allocation of future real-time logistics orders closer to the maximization of logistics resource utilization under constraints.

[0031] See also Figure 1 The flowchart of the logistics order allocation method provided by the embodiment of the present application is shown; the embodiment of the present application provides a logistics order allocation method, including:

[0032] Step S110: Use historical logistics orders to solve the pre-built optimal matching algorithm model to obtain the order splitting results of the historical logistics orders.

[0033] The order splitting result refers to the proportion of logistics orders (also referred to as the order splitting ratio) or the number of logistics orders allocated to each carrier for each transportation service product on each transportation route during the logistics order allocation process. The above order splitting result can be in the form of a recommended data table, which is a strategic plan for splitting historical logistics orders from multiple dimensions using an optimal matching algorithm model. The multiple dimensions here can include: transportation route, package weight, business type, whether it is an air embargo, candidate carriers and transportation service products and / or weight priority, etc.

[0034] It is understandable that the transportation service product in the embodiment of the present application refers to the product of the transportation service provided by the carrier in different modes of transportation, which is also called the carrier product, for example: a company's standard express product or special express product, etc.; among them, the special express product is to transport goods by air, which has the fastest speed, but the highest price. For example, standard express is to transport goods by land, which has a lower price than air transportation, but a slower speed than air transportation. The above-mentioned historical logistics orders are generally distributed on multiple transportation routes. On one of the multiple transportation routes, there may be a large number of goods that need to be allocated to one or more transportation service products. At this time, it is necessary to determine the order distribution ratio of each transportation service product on the transportation route.

[0035] It is understandable that the historical logistics orders are used to solve the optimal matching algorithm model. This solution process can be carried out in an offline simulation environment. The offline simulation environment is a simulation environment that is separated from the online environment for simulation. Therefore, this process is also called offline simulation, and this process does not affect the real online environment.

[0036] There are many types of constraints in the above-mentioned optimal matching algorithm model, including: order-based constraints and / or service quality constraints, etc. For ease of understanding, some constraints are illustrated below.

[0037] The above-mentioned order-based constraints may be order quantity ratio constraints, such as: the order quantity control range of the carrier in the overall dimension, the fluctuation range of the carrier's order quantity in the city-origination dimension compared with the previous cycle, and the constraints on the proportion of the carrier's order quantity in the route dimension. Specifically, the above-mentioned order quantity ratio constraints may be, for example: a contract has been signed with carrier B, which stipulates that the number of logistics orders of carrier B needs to be greater than 40% of all logistics orders in the overall market, the number of logistics orders of carrier C needs to be greater than 50% of all logistics orders in the overall market, and the number of logistics orders of carrier D needs to be greater than 5% of all logistics orders in the overall market. The above-mentioned overall market refers to all logistics orders of an e-commerce platform within a preset time range. The above-mentioned order quantity ratio constraints may be, for example: the constraint on the total number of orders on each transport route is expressed as: The constraints on the upper and lower limits of the global split ratio for the carrier are expressed as follows: The upper and lower limits of the order allocation ratio for a transport order from a specific origin city to a carrier can be expressed as: For the meaning of letters not explained above, please refer to the following explanation of the service quality score constraints and the following explanation of the decision variables.

[0038] For example, the constraints of the service quality score are: the transport capacity of a carrier, the overtime rate on the transport route, and the customer complaint rate are weighted and calculated to convert them into a service quality score. Specifically, the constraints of the service quality score for the market can be expressed as:∑ i,m X im q im ≥NQ. Where i represents the route number, j represents the city number, m represents the carrier number, a represents the transportation service product number provided by the carrier, Q represents the overall service quality benchmark score, N represents the total number of orders per month, and N i represents the total monthly order volume of line i, N j represents the total number of monthly orders originating from city j, q im represents the service quality score of carrier m on transport route i, and They represent the upper and lower limits of the global order splitting ratio of carrier m, and They represent the upper and lower limits of the global order splitting ratio of carrier m in origin city j respectively. For the meaning of the letters not explained above, please refer to the explanation of decision variables below.

[0039] Step S120: Use the order splitting result to run the real-time logistics order in the online environment to obtain the running result.

[0040] Step S130: Determine whether the accompanying running result meets the constraint conditions of the optimal matching algorithm model.

[0041] Step S140: If the accompanying result meets the constraint conditions of the optimal matching algorithm model, the optimal matching algorithm model is pushed to the online environment so that the optimal matching algorithm model can optimize and allocate future real-time logistics orders in the online environment.

[0042] The decision variables of the above-mentioned optimal matching algorithm model may include: ia represents the number of orders assigned to transport service product a on route i, x ia is the smallest granularity decision variable. im represents the total number of orders of carrier m for route i. A carrier may have multiple transportation service products, so Among them, A m is the set of transportation service products provided by carrier m and is a known quantity. jmIt represents the total number of orders of carrier m in origin city j. A city has multiple origin routes, so Among them, I j is the set of routes from city j, which is a known quantity. It can be understood that the above decision variables (x ia ,X im ,X jm ) can be continuous variables.

[0043] In the implementation process of the above solution, by using the real data of historical logistics orders to solve the pre-built optimal matching algorithm model, it can ensure that the model performs more reliably in actual scenarios, so that the model can learn and adapt to different actual situations, and using real-time logistics orders for running verification can further verify the performance of the model in the current market environment, which ensures that the model not only performs well on historical data, but also can cope with the latest market changes. This logistics order allocation method based on real data, real-time logistics orders and optimal matching algorithm model is more scientific and objective than manually maintaining weights, and effectively avoids deviations caused by personal experience or subjective judgment. Therefore, when the running results meet the constraints and are pushed to the business production environment, the allocation of logistics orders can minimize the total freight cost under the business constraints, and maximize the utilization of logistics resources under the constraints.

[0044] Optionally, the above-mentioned historical logistics orders can directly adopt the logistics orders of the previous preset period (such as the previous month). For example, if you want to predict the logistics order volume on a certain transportation route in the current month, you can let long-term trends, seasonal changes, periodic changes and other factors as the influencing factors of the order volume, and use the time series prediction algorithm to fit the logistics order data of the previous month, so as to fit the logistics order volume of the transportation route in the current month. In addition, the logistics order data of the previous month can also be counted to calculate the carrier's transportation capacity, overtime rate, customer complaint rate, assessment results, etc., and the origin city of each logistics order and the order volume proportion of each carrier can also be counted. The system can also obtain the constraints on the carrier's order volume ratio of the big market dimension, city origin dimension, and transportation route dimension of the business configuration.

[0045] It is understandable that the transportation route dimension of the above-mentioned historical logistics orders can be divided into two categories: algorithm-adjustable routes and unadjustable routes. Among them, the adjustable routes are the routes that the algorithm can select and recommend carriers and their transportation service products according to the rules. Unadjustable routes refer to excluded routes (such as too few monthly orders), excluded carriers' transportation service products (such as exclusive contracts), or routes that have already occurred (such as when calculated in the middle of the month).

[0046] It can be understood that the ideal effect of allocating logistics orders closer to each other can be to minimize the total freight cost of logistics orders under the constraints, that is, to maximize the utilization of logistics resources under the constraints, that is, to make the total freight cost under the constraints lower. The calculation process of the total freight cost here can be to subtract accounts payable from accounts receivable. Among them, the above-mentioned accounts receivable refers to the freight settlement amount between the e-commerce platform and the buyer user. This freight calculation method can be calculated using the standard freight template of the carrier; the above-mentioned accounts payable refers to the freight settlement amount between the e-commerce platform and the carrier. Since the carrier has business discounts or tiered price discounts for customers with a large number of logistics orders that need to be carried (for example, there is a discount for freight tiered settlement, or there is a discount for single-volume tiered settlement), etc., therefore, through the above-mentioned optimal allocation algorithm model, a logistics order allocation strategy with a larger business discount or tiered price discount can be found, thereby achieving the ideal effect of making the total freight cost under the constraints lower, thereby effectively maximizing the utilization of logistics resources under the constraints.

[0047] As an optional implementation of the above-mentioned logistics order allocation method, if the total transportation cost of the transportation service product is settled in a freight ladder, then it means that the ladder of the transportation service product is a freight ladder. Before using historical logistics orders to solve the pre-built optimal allocation algorithm model, a freight ladder model can also be performed. This implementation may include:

[0048] Step S101: Use a piecewise linear function to perform freight ladder modeling on the transportation service product to obtain a first linear expression, which is used to construct an optimal matching algorithm model.

[0049] The above freight ladder means that when the total logistics freight is in a certain range, you can enjoy the corresponding freight discount for that range. For example, if the total transportation cost of a transportation service product is settled through the freight ladder, and when the monthly freight is in [20 million, 30 million), the discount value is 0.9; when the monthly freight is in [30 million, 40 million), the discount value is 0.8; when the monthly freight is in [40 million, 50 million), the discount value is 0.7; when the monthly freight is in [50 million, 99999999 million], the discount value is 0.6. It can be understood that the freight ladder is usually a full discount mode, and is usually calculated based on the sum of the freight of all routes.

[0050] The implementation of the above step S101 can be, for example: introducing a 0-1 variable μ k To indicate whether the freight falls within the kth interval, a continuous variable μ is introduced k The discount value is σ kCorresponding freight, get the total freight after discount y and the freight before discount c for the transportation service product a ia And the number of orders x ia After the piecewise linear function relationship between them is used to model the freight ladder of the transportation service product, the first linear expression obtained can be expressed as follows:

[0051]

[0052] Among them, μ k ∈{0,1} indicates whether the freight falls within the kth interval. The discount value is σ k The corresponding freight, A k-1 represents the lower limit of the kth interval, A k represents the upper limit of the kth interval, x ia represents the number of orders assigned to transport service product a on route i, c ia represents the price calculated based on the normal first and subsequent weight when route i is assigned to a carrier's transport service product a, σ j represents the discount value of the kth interval, and y represents the first linear expression of the total freight after discount. The above first linear expression is used to build the optimal matching algorithm model. It can be understood that the above first formula indicates that only one μ value is equal to 1, and the others are 0; the second formula indicates that only one π value is greater than 0, and the others are 0, which ensures that the total freight will only be fully discounted at one price.

[0053] As an optional implementation of the above-mentioned logistics order allocation method, if the total transportation cost of the transportation service product is settled in an order quantity ladder, then it means that the ladder of the transportation service product is a single quantity ladder. Before using historical logistics orders to solve the pre-built optimal allocation algorithm model, single quantity ladder modeling can also be performed. This implementation may include:

[0054] Step S102: Use a nonlinear function to perform single-volume step modeling on the transportation service product to obtain a nonlinear expression, which is used to construct an optimal matching algorithm model.

[0055] The above-mentioned order quantity ladder means that when the total logistics order quantity (hereinafter referred to as order quantity) is within a certain range, you can enjoy the preset freight discount. The order quantity ladder can be expressed by the formula: [A 0 ,A 1 ),[A 1 ,A 2 ),…,[A n-1 ,A n). A specific example of the above order quantity ladder is as follows: assuming that the total transportation cost of a certain transportation service product is settled through the order quantity ladder, and when the monthly order quantity is [1 million, 3 million), the discount value is 0.8; when the monthly order quantity is [3 million, 4 million), the discount value is 0.7; when the monthly order quantity is [4 million, 5 million), the discount value is 0.6; when the monthly order quantity is [5 million, 99999999 million], the discount value is 0.5. It can be understood that the order quantity ladder is usually a full-volume discount mode, and the total order quantity is usually calculated according to all routes.

[0056] For example, in the implementation of step S102, for a transport service product a of a certain carrier, a 0-1 variable μ is introduced. k Indicates that the total number of orders allocated and following the order quantity ladder falls within the kth interval, and the continuous variable π k Indicates the corresponding order quantity. The above-mentioned constraints for the order quantity ladder modeling may include:

[0057]

[0058]

[0059] Among them, μ k Indicates that the total number of orders allocated and following the order quantity ladder falls within the kth interval, A k represents the upper limit of the order quantity in the kth interval, A k-1 represents the lower limit of the order quantity in the kth interval, x ia represents the number of orders assigned to transport service product a on route i, μ and x are both decision variables, π k The discount value is σ k The corresponding order quantity. It can be understood that the first formula above indicates that the number of orders carried will fall within a range, the second formula is the upper and lower limit constraints of the interval order quantity, and the third expression is the total order quantity undertaken by the transportation service product a.

[0060] However, at this time the interval variable π k It represents the order quantity rather than the freight, and a new auxiliary variable needs to be introduced to represent the freight after discount. In the total cost expression, the freight before discount of a carrier's transportation service product a is ∑ i x ia c ia , the discount factor is ∑ k μ k σ k Therefore, a nonlinear function is used to perform a single-volume step modeling on the transportation service product to obtain a nonlinear expression. The total transportation cost after the discount can be expressed as follows:

[0061]

[0062] Among them, x ia represents the number of orders assigned to transport service product a on route i, c ia represents the price calculated based on the normal first and subsequent weight when route i is assigned to a carrier's transport service product a, μ k Indicates that the total number of orders allocated and following the order quantity ladder falls within the kth interval, σ k Represents the discount value of the kth interval. Because in the above expression, μ and x are both decision variables, multiplying the two will introduce nonlinearity, so this expression is a nonlinear expression.

[0063] As an optional implementation of the above logistics order allocation method, the above nonlinear expression may further include a nonlinear term; after obtaining the nonlinear expression, it further includes:

[0064] Step S103: After introducing auxiliary variables and inequality constraints into the nonlinear expression, the nonlinear term is eliminated to obtain a second linear expression. The second linear expression and the inequality constraints are used to construct an optimal matching algorithm model.

[0065] For example, in the above nonlinear expression, μ and x are both decision variables, and multiplying them will introduce nonlinear terms. Note that μ∈[0,1], so an auxiliary continuous variable z can be introduced. iak =x ia μ k And the following constraints are used for linearization, that is, auxiliary variables and inequality constraints are introduced into the nonlinear expression as follows:

[0066] 0≤z iak ≤N i μ k

[0067] x ia +N i (μ k -1)≤z iak ≤x ia

[0068] Among them, the auxiliary continuous variable z iak and the decision variable x ia The domain of z is the same as iak =x ia μ k Indicates z iak Equal to x ia With μ k The product of; the interpretation is as follows, when μ k = 0, z iak =0; if μ k =1, then ziak =x ia .

[0069] It can be understood that after introducing auxiliary variables and inequality constraints into the nonlinear expression to eliminate the nonlinear term, the second linear expression can be obtained as follows:

[0070]

[0071] Among them, x ia represents the number of orders assigned to transport service product a on route i, c ia represents the price calculated based on the normal first and subsequent weight when route i is assigned to a carrier's transport service product a, μ k Indicates that the total number of orders allocated and following the order quantity ladder falls within the kth interval, σ k represents the discount value of the kth interval.

[0072] As an optional implementation of the logistics order allocation method, the above-mentioned optimal allocation algorithm model may include: multiple decision variables; before solving the pre-built optimal allocation algorithm model using historical logistics orders, it also includes:

[0073] Step S104: for each decision variable among the multiple decision variables, determine whether the decision variable is a known variable at the current moment.

[0074] Step S105: If the decision variable is a known variable at the current moment, the decision variable is pre-set to a fixed value.

[0075] The implementation method of the above steps S104 to S105 includes: according to the routes that cannot be controlled by the algorithm, the step intervals can be calculated, and the values ​​of some decision variables can be fixed in advance, thereby reducing the number of variables. For example: Assume that a certain transportation service product can be calculated by the order quantity step method, and when the monthly order quantity is [1 million, 3 million), the discount value is 0.8; when the monthly order quantity is [3 million, 4 million), the discount value is 0.7; when the monthly order quantity is [4 million, 5 million), the discount value is 0.6; when the monthly order quantity is [5 million, 99999999 million], the discount value is 0.5. At this time, in the uncontrollable routes, the order quantity allocated to a certain transportation service product has reached 4.5 million, so its order quantity step will at least fall above the second step, that is, the decision variable μ 0 (indicates whether the single quantity falls within the 0th interval) and μ 1 (indicates whether the single quantity falls within the first interval) is a known variable at the current moment, so the decision variable μ 0 and μ 1 Set to a fixed value (such as 0), then μ 0 =μ1 =0.

[0076] Optionally, constraint enhancement can be used to accelerate the solution process of the above-mentioned optimal matching algorithm model, for example: using the upper and lower limits of the order quantity constraints of the originating city to tighten the upper and lower limits of the global order quantity constraints, etc.

[0077] Optionally, the solution process of the above-mentioned optimal matching algorithm model can be accelerated by feasibility verification. For example, according to the algorithm's uncontrollable routes, the number of constant waybills that have occurred is calculated; according to the candidate set of controllable routes, the upper and lower bounds that the carrier can control are calculated, and then it can be determined whether the upper and lower limit constraints of the order quantity (global, origin, route) can be met, and an exception can be thrown or the output causes explainable information that is infeasible. For example, assuming that the monthly market order volume is 10 million, and the global order volume upper and lower limits of a carrier m are 60%-70%, then the order volume that should be met is [6 million, 7 million], and 4.3 million orders have been allocated (already occurred or cannot be controlled) to carrier m; therefore, among the controllable orders, the range that needs to be allocated to carrier m is [1.7 million, 2.7 million]. Assuming that the total controllable order volume is 1 million, which is less than the minimum 1.7 million, even if all the controllable orders are allocated to carrier m, it still cannot meet the global order volume lower limit constraint, so it is not feasible.

[0078] As an optional implementation of the above step S110, the above implementation of solving the pre-built optimal matching algorithm model may include:

[0079] Step S111: Obtain all interval combinations of discount variables, and rearrange all interval combinations in reverse order to obtain interval combinations in reverse order.

[0080] All interval combinations of the above discount variables refer to interval combinations of all transportation service products participating in the step calculation, which may include freight intervals and order quantity intervals. It can be understood that the reason for traversing each interval combination in reverse order is that the larger the step of the interval, the greater the discount.

[0081] Step S112: During the traversal of each interval combination in the reversed interval combination, it is determined whether a feasible solution is successfully obtained by using the historical logistics orders to solve the optimal matching algorithm model. The feasible solution is an interval combination that satisfies the constraints of the optimal matching algorithm model.

[0082] For example, the implementation of the above steps S111 to S112 is as follows: the discount variable of transportation service products 1 and 2 is u 1 and u 2 , where u 1 The step interval is {0,1,2,3}, u 2 The step interval is {0,1}. Assume that the variable u1 The interval 0 (can be used To represent) is fixed, then the decision variable u 1 The interval to be determined is {1,2,3}, and the decision variable u 2 The pending interval is {0,1}. The pending interval here means that other intervals are pre-fixed (that is, it is known whether the interval will fall into the interval at the current moment). In this case, we can respectively 1 The undetermined interval {1,2,3} (arranged in descending order to {3,2,1}) and the decision variable u 2 The pending interval is {0,1} (in descending order, {1,0}), and all interval combinations are set as follows:

[0083]

[0084] From the above, we can know that for the decision variable u 1 The undetermined interval {1,2,3} and the decision variable u 2 There are a total of 6 interval combinations for all interval combinations set by the pending interval {0,1}. The historical logistics orders are used to solve the optimal matching algorithm model for these 6 interval combinations. If feasible, the order splitting ratio or order splitting quantity corresponding to the interval combination can be obtained.

[0085] Step S113: If a feasible solution is successfully obtained by using the historical logistics orders to optimize the matching algorithm model, the order splitting ratio or order splitting quantity corresponding to the interval combination is determined as the order splitting result.

[0086] For example, if the historical logistics order matching algorithm model successfully solves the first interval combination The corresponding split ratio or split quantity can be combined with the first interval The corresponding order splitting ratio or order splitting quantity is determined as the order splitting result. Because each interval combination is traversed in reverse order, and the larger the interval step, the greater the discount, in this case, there is no need to calculate other interval combinations. The current solution is the optimal solution, that is, the optimal interval combination that meets the constraints of the optimal matching algorithm model.

[0087] Optionally, as another optional implementation of the above step S110, for each interval combination in all interval combinations, the value range of the ladder can be narrowed by fixing the variables or by business experience, and then the total freight corresponding to the interval combination is calculated using the optimal matching algorithm model, and the interval combination with the smallest total freight is selected from all interval combinations. Finally, the order splitting ratio or order splitting quantity corresponding to the interval combination with the smallest total freight is obtained, and the order splitting ratio or order splitting quantity is determined as the order splitting result.

[0088] Optionally, as another optional implementation of the above-mentioned step S110, the above-mentioned implementation method of solving the pre-constructed optimal matching algorithm model can be solved by a global optimization method. For example, the global optimization method is that the optimal model takes all constraints into consideration at the same time, and the optimal solution to the problem can be obtained through a single linearized integer programming. This solution method is suitable for scenarios with fewer control lines such as in the later stage of intervention.

[0089] Optionally, after solving the optimal allocation algorithm model, the optimal allocation algorithm model can also issue a recommendation data table based on dimensions such as route, package weight, business type, whether there is an air embargo, recommended carriers, recommended transportation service products, weight priority, etc. This table is a strategic plan for the order splitting results of each transportation service product.

[0090] See also Figure 2 The structure diagram of the logistics order distribution device provided in the embodiment of the present application is shown; the embodiment of the present application provides a logistics order distribution device 200, including:

[0091] The order splitting result obtaining module 210 is used to solve the pre-built optimal matching algorithm model using the historical logistics orders to obtain the order splitting results of the historical logistics orders.

[0092] The accompanying result obtaining module 220 is used to use the order splitting results to accompany the real-time logistics orders in the online environment to obtain the accompanying results.

[0093] The accompanying running result judgment module 230 is used to judge whether the accompanying running result meets the constraint conditions of the optimal matching algorithm model.

[0094] The algorithm model pushing module 240 is used to push the optimal matching algorithm model to the online environment if the constraint conditions of the optimal matching algorithm model are met, so that the optimal matching algorithm model can optimize and allocate future real-time logistics orders in the online environment.

[0095] As an optional implementation of the above device, the logistics order distribution device further includes:

[0096] The cost ladder modeling module is used to use a piecewise linear function to perform freight ladder modeling on the transportation service product to obtain a first linear expression, and the first linear expression is used to construct an optimal matching algorithm model.

[0097] As an optional implementation of the above device, the logistics order distribution device further includes:

[0098] The single quantity step modeling module is used to use nonlinear functions to perform single quantity step modeling on the transportation service product to obtain a nonlinear expression, which is used to construct an optimal matching algorithm model.

[0099] As an optional implementation of the above device, the nonlinear expression includes a nonlinear term; the logistics order distribution device also includes:

[0100] The nonlinear term elimination module is used to eliminate the nonlinear term after introducing auxiliary variables and inequality constraints into the nonlinear expression to obtain a second linear expression. The second linear expression and the inequality constraints are used to construct an optimal matching algorithm model.

[0101] As an optional implementation of the above device, the optimal matching algorithm model includes: multiple decision variables; the logistics order allocation device also includes:

[0102] The decision variable judgment module is used to judge, for each decision variable among the multiple decision variables, whether the decision variable is a known variable at the current moment.

[0103] The variable fixed setting module is used to pre-set the decision variable to a fixed value if the decision variable is a known variable at the current moment.

[0104] As an optional implementation of the above device, the order splitting result obtaining module includes:

[0105] The combined reverse order rearrangement submodule is used to obtain all interval combinations of discount variables, and to rearrange all interval combinations in reverse order to obtain reversed interval combinations.

[0106] The interval combination judgment submodule is used to determine whether a feasible solution is successfully solved for the optimal matching algorithm model using historical logistics orders during the traversal of each interval combination in the reversed interval combination. The feasible solution is an interval combination that meets the constraints of the optimal matching algorithm model.

[0107] The order splitting result determination submodule is used to determine the order splitting ratio or order splitting quantity corresponding to the feasible solution as the order splitting result if a feasible solution is successfully solved by the optimal matching algorithm model using historical logistics orders.

[0108] It should be understood that the device corresponds to the above-mentioned logistics order allocation method embodiment and can execute the various steps involved in the above-mentioned method embodiment. The specific functions of the device can be found in the description above, and the detailed description is appropriately omitted here. The device includes at least one software function module that can be stored in a memory in the form of software or firmware or solidified in the operating system (OS) of the device.

[0109] See also Figure 3The electronic device 300 provided in the embodiment of the present application includes: a processor 310 and a memory 320, wherein the memory 320 stores machine-readable instructions executable by the processor 310, and when the machine-readable instructions are executed by the processor 310, the above method is executed.

[0110] The embodiment of the present application also provides a computer-readable storage medium 330, on which a computer program is stored, and the computer program is executed by the processor 310 to execute the above method. The computer-readable storage medium 330 can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, disk or optical disk.

[0111] The embodiment of the present application also provides a computer program product, including: a computer program or a computer instruction, and the computer program or the computer instruction executes the method described above when executed by a processor.

[0112] It should be noted that each embodiment in this specification is described in a progressive manner, and each embodiment focuses on the differences from other embodiments, and the same or similar parts between the embodiments can be referred to each other. For the device embodiment, since it is basically similar to the method embodiment, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiment.

[0113] In several embodiments provided by the embodiments of the present application, it should be understood that the disclosed device and method can also be implemented in other ways. The device embodiments described above are only schematic, for example, the flowcharts and block diagrams in the accompanying drawings show the possible implementation architecture, functions and operations of the devices, methods and computer program products according to the multiple embodiments of the embodiments of the present application. In this regard, each box in the flowchart or block diagram can represent a part of a module, a program segment or a code, and a part of a module, a program segment or a code includes one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions marked in the box can also be different from the order of occurrence marked in the accompanying drawings. For example, two consecutive boxes can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, which is mainly determined by the functions involved.

[0114] In addition, each functional module of each embodiment in the embodiment of the present application can be integrated together to form an independent part, or each module can exist separately, or two or more modules can be integrated to form an independent part. In addition, in the description of this specification, the description of reference terms "one embodiment", "some embodiments", "example", "specific example", "some examples", etc. means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the embodiment of the present application. In this specification, the schematic representation of the above terms does not necessarily target the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described can be combined in any one or more embodiments or examples in a suitable manner. In addition, those skilled in the art can combine and combine the different embodiments or examples described in this specification and the features of the different embodiments or examples without contradiction.

[0115] The above description is only an optional implementation manner of the embodiments of the present application, but the protection scope of the embodiments of the present application is not limited thereto. Any technician familiar with the technical field can easily think of changes or replacements within the technical scope disclosed in the embodiments of the present application, which should be covered within the protection scope of the embodiments of the present application.

Claims

1. A logistics order allocation method, characterized in that: include: Using historical logistics orders to solve the pre-built optimal allocation algorithm model, and obtaining the order splitting results of the historical logistics orders; Use the order splitting result to run the real-time logistics order in the online environment to obtain the running result; Determine whether the accompanying running result meets the constraint conditions of the optimal matching algorithm model; If so, the optimal matching algorithm model is pushed to the online environment so that the optimal matching algorithm model can optimize and allocate future real-time logistics orders in the online environment.

2. The method according to claim 1, characterized in that: The total transportation cost of the transportation service product is settled in a freight ladder manner; before solving the pre-built optimal allocation algorithm model using historical logistics orders, it also includes: The freight ladder modeling of the transportation service product is performed using a piecewise linear function to obtain a first linear expression, and the first linear expression is used to construct the optimal matching algorithm model.

3. The method according to claim 1, characterized in that The total transportation cost of the transportation service product is settled in a tiered manner based on the order volume; before solving the pre-built optimal allocation algorithm model using historical logistics orders, it also includes: The transport service product is modeled in single quantity steps using a nonlinear function to obtain a nonlinear expression, which is used to construct the optimal matching algorithm model.

4. The method according to claim 3, characterized in that The nonlinear expression includes a nonlinear term; after obtaining the nonlinear expression, the following further includes: After introducing auxiliary variables and inequality constraints into the nonlinear expression, the nonlinear term is eliminated to obtain a second linear expression, and the second linear expression and the inequality constraint are used to construct the optimal matching algorithm model.

5. The method according to claim 1, characterized in that: The optimal matching algorithm model includes: a plurality of decision variables; before solving the pre-built optimal matching algorithm model using the historical logistics orders, it also includes: For each decision variable among the multiple decision variables, determining whether the decision variable is a known variable at a current moment; If so, the decision variable is pre-set to a fixed value.

6. The method according to claim 1, characterized in that The use of historical logistics orders to solve the pre-built optimal matching algorithm model includes: Obtain all interval combinations of discount variables, and rearrange all the interval combinations in reverse order to obtain interval combinations in reverse order; During the traversal of each interval combination in the reversed interval combination, determining whether a feasible solution is successfully obtained for the optimal matching algorithm model using the historical logistics orders, wherein the feasible solution is an interval combination that satisfies the constraint conditions of the optimal matching algorithm model; If so, the order splitting ratio or order splitting quantity corresponding to the feasible solution is determined as the order splitting result.

7. A logistics order distribution device, characterized in that: include: An order splitting result obtaining module is used to solve a pre-built optimal matching algorithm model using historical logistics orders to obtain order splitting results of the historical logistics orders; A running-accompanying result obtaining module is used to use the order splitting result to run-accompanying the real-time logistics order in the online environment to obtain the running-accompanying result; A running-along result judgment module, used to judge whether the running-along result meets the constraint conditions of the optimal matching algorithm model; The algorithm model pushing module is used to push the optimal matching algorithm model to the online environment if the constraint conditions of the optimal matching algorithm model are met, so that the optimal matching algorithm model can optimize and allocate future real-time logistics orders in the online environment.

8. An electronic device, characterized in that: include: A processor and a memory, wherein the memory stores machine-readable instructions executable by the processor, and the machine-readable instructions are executed by the processor to perform any method according to claims 1 to 6.

9. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the method according to any one of claims 1 to 6 is executed.

10. A computer program product, characterized in that include: A computer program or a computer instruction, wherein when the computer program or the computer instruction is executed by a processor, the method according to any one of claims 1 to 6 is executed.

Images