Intelligent linkage decision-making method and service platform for production and transportation

By applying collaborative optimization CO algorithm and dynamic tolerance in the production and operation collaboration system, and updating the expected values ​​of production and distribution variables, the difficulties of gap-based production enterprises in dynamic collaborative decision-making are solved, and efficient linkage decision-making is achieved.

CN118428652BActive Publication Date: 2024-11-08JINAN UNIVERSITY
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202410501296.9
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-04-25
Publication Date
2024-11-08
Estimated Expiration
2044-04-25

AI Technical Summary

Technical Problem

Dynamic collaborative decision-making for production and distribution is difficult to find the best in gap-based production enterprises, real-time decision-making is difficult under dynamic interference, and existing technologies are difficult to effectively solve.

Method used

Coordinated optimization CO algorithm and dynamic tolerance are used for collaborative optimization. By updating the expected value of production and distribution variables in the production and operation collaboration system, until the production and operation consistency constraints are met, linkage decision results are generated, and coupling functions are used for consistent convergence and dynamic updates are updated. Relax the factor to improve decision efficiency.

Benefits of technology

It improves the speed of decision-making calculation, reduces the time-consuming of collaborative optimization, improves the linkage decision-making efficiency between production and transportation, and solves the problem of real-time decision-making under dynamic interference.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118428652B_ABST
    Figure CN118428652B_ABST
Patent Text Reader

Abstract

The present application relates to an intelligent linkage decision-making method and service platform for production and transportation, the method comprising: sending updated expected values ​​of production variables and expected values ​​of distribution variables to the production planning subsystem and the transportation planning subsystem respectively; receiving the production variable return value returned by the production planning subsystem in response to the expected value of the production variable and the distribution variable return value returned by the transportation planning subsystem in response to the expected value of the distribution variable, and judging whether the production variable return value and the distribution variable return value satisfy the preset production and transportation consistency constraint; when judging that the production variable return value and the distribution variable return value do not satisfy the preset production and transportation consistency constraint, repeatedly performing collaborative optimization using the preset collaborative optimization CO algorithm and dynamic tolerance until the new production variable return value and the distribution variable return value satisfy the production and transportation consistency constraint, and obtaining the linkage decision result.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present application relates to the field of smart manufacturing technology, and in particular to an intelligent linkage decision-making method and service platform for production and transportation. Background Art

[0002] Driven by customer demand for multiple varieties and small batches, traditional manufacturing companies such as coatings and petrochemicals that mainly engage in mass production have been greatly impacted. In order to pursue large-scale production and improve economic benefits, these companies tend to use large-capacity equipment for batch production and entrust third-party logistics companies to dispatch vehicles to transport retail orders in batches to distribution locations.

[0003] In the relevant technologies, for the production and transportation decisions under the entire batch mode, the splitting of customer orders and the reorganization of production orders are particularly critical, which involves a series of factors such as customer order product category, size, delivery period, delivery location, etc. At the same time, the previous and subsequent transformations of the corresponding batches in the two coupled links of production and distribution must also be taken into account. Under this circumstance, there are huge challenges in searching for the global optimal solution of the entire system; in the relevant technologies, the scheme for production and transportation decision-making of production and distribution using heuristic optimization algorithms has the problems of low decision-making planning efficiency, high production and distribution costs corresponding to the decided schemes, difficulty in finding the best solution in the linkage decision-making process, and difficulty in making real-time decisions under dynamic interference.

[0004] At present, the solutions for dynamic collaborative decision-making for production and transportation in related technologies make it difficult for intermittent production enterprises to find the best solution in the process of production and transportation linkage decision-making, and the problem of difficulty in real-time decision-making under dynamic interference has not yet been proposed. An effective solution. Summary of the invention

[0005] The embodiments of the present application provide an intelligent linkage decision method for production and transportation, a production scheduling decision method, a transportation planning decision method and a service platform, so as to at least solve the dynamic collaborative decision-making solutions for production and distribution in the related technologies, so as to make it difficult for intermittent production enterprises to find the best solution in the production and transportation linkage decision-making process and make real-time decisions under dynamic interference.

[0006] In the first aspect, an embodiment of the present application provides an intelligent linkage decision-making method for production and transportation, which is applied to a production-transportation collaboration system, including: after completing the current collaborative optimization, the updated production variable expected value and the distribution variable expected value are respectively sent to the production planning subsystem and the transportation planning subsystem; the production variable return value returned by the production planning subsystem in response to the production variable expected value and the distribution variable return value returned by the transportation planning subsystem in response to the distribution variable expected value are received, and it is determined whether the production variable return value and the distribution variable return value meet the preset production-transportation consistency constraints, wherein the production variable return value is used to represent the production scheduling parameters for batch production planning of multiple production orders, and the distribution variable return value is used to represent the distribution parameters for combined delivery planning of multiple production orders to be delivered. number; in the case of judging that the production variable return value and the distribution variable return value do not satisfy the preset production-operation consistency constraint, repeatedly executing the collaborative optimization using the preset collaborative optimization CO algorithm and the dynamic tolerance to update the production variable expected value and the distribution variable expected value, and judging whether the production variable return value in response to the updated production variable expected value and the distribution variable return value in response to the updated distribution variable expected value satisfy the production-operation consistency constraint, until the new response production variable return value and the distribution variable return value satisfy the production-operation consistency constraint, and obtaining a linkage decision result, wherein the CO algorithm uses a dynamic relaxation factor to update the dynamic tolerance, and the linkage decision result includes the production variable return value and the distribution variable return value that satisfy the production-operation consistency constraint.

[0007] In a second aspect, an embodiment of the present application provides a production scheduling planning method, which is applied to a production planning subsystem and is used to return a variable value, including: obtaining production order information of the current corresponding batch production, and performing initial population coding on the production order information to generate multiple first double-coded individuals, wherein the first double-coded individual includes multiple first double-coded sub-units, and one of the first double-coded sub-units includes multiple double codes in the same production workshop, and each of the double codes includes an order code and a batch code that constitute a code pair, and the order code is used to characterize the order to which the corresponding product belongs, and the batch code is used to characterize the batching of the products corresponding to the order code; based on the first double-coded individual corresponding to Individual fitness and a preset selection strategy, a set number of second dual-coded individuals are selected from multiple first dual-coded individuals; the order codes and batch codes corresponding to all the second dual-coded individuals are subjected to genetic evolution iteration operations until a target dual-coded individual is generated, and the target dual-coded individual is decoded to generate the production variable return value, wherein the genetic evolution operation includes a crossover operation, a mutation operation, and a perturbation operation, and the individual fitness is determined by solving a preset production objective function, and the production objective function is constructed by weighting the difference between the received expected value of the production variable and the production variable return value generated by decoding the first dual-coded individual.

[0008] In a third aspect, an embodiment of the present application provides a transportation planning decision method, which is applied to a transportation planning subsystem and is used to generate a distribution variable return value, including: obtaining the incoming order information corresponding to the current delivery, and encoding the incoming order information to generate an initial double code body, wherein the initial double code body includes an order code sequence having multiple distribution order coders and a distribution code sequence having multiple vehicle group coders, one distribution order coder corresponds to one vehicle group coder, and forms a distribution double coder, the distribution order coder is used to characterize the code of the corresponding incoming order, and the order of the distribution order coder in the corresponding order code sequence is used to characterize the priority of the distribution order, the vehicle group coder is used to characterize the type of distribution vehicle selected when the corresponding incoming order is distributed, and one vehicle group coder also maps at least one distribution path, and the nodes of the distribution path include the initial position of the distribution vehicle and the storage location of the product corresponding to the incoming order; using linear sorting to select the LRS strategy and the control line CL strategy, respectively from a variety of destruction operations, a variety of repair operations and a variety of disturbance operations corresponding to the adaptive large neighborhood search ALNS algorithm. , select a target destruction operation, a target repair operation and a target disturbance operation, and use the target destruction operation, the target repair operation and the target disturbance operation in sequence to destroy, repair and disturb all the delivery double coders and the delivery order coders and the vehicle assembly coders corresponding to the delivery double coders to generate an alternative delivery double coder, wherein the destruction operation includes one of the following: head and tail truncation, random removal, random block cutting, random excision, the repair operation includes one of the following: random block filling, random gap filling, head and tail gap filling, greedy gap filling, the disturbance The dynamic operation includes one of the following: transposition, block shifting, flipping, order change, and greedy replacement; decoding the alternative delivery double code body to obtain alternative delivery information, and obtaining the delivery variable value corresponding to the alternative delivery double code body from the alternative delivery information, and according to the deviation between the delivery variable value and the received delivery variable expected value, destroying, repairing and perturbing the alternative delivery double code body iteratively until a target delivery double code body is generated; decoding the target delivery double code body to obtain the target delivery information, and generating the delivery variable return value according to the target delivery information.

[0009] In a fourth aspect, an embodiment of the present application provides a service platform, comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, wherein when the processor executes the computer program, it implements the steps of the intelligent linkage decision-making method for production and transportation as described in the first aspect, or implements the production scheduling planning method as described in the second aspect, or implements the distribution planning decision-making method as described in the third aspect.

[0010] Compared with the related art, the intelligent linkage decision method, production scheduling decision method, transportation planning decision method and service platform for production and transportation provided by the embodiments of the present application adopt the method of, after completing the current collaborative optimization, sending the updated production variable expected value and distribution variable expected value to the production planning subsystem and the transportation planning subsystem respectively; receiving the production variable return value returned by the production planning subsystem in response to the production variable expected value and the distribution variable return value returned by the transportation planning subsystem in response to the distribution variable expected value, and judging whether the production variable return value and the distribution variable return value meet the preset production and transportation consistency constraints, the production variable return value is used to represent the production scheduling parameters for batch production planning of multiple production orders, and the distribution variable return value is used to represent the distribution parameters for combined transportation planning of multiple production orders to be delivered; when it is judged that the production variable return value and the distribution variable return value do not meet the preset production and transportation consistency constraints, repeatedly executing the collaborative optimization CO algorithm and dynamic tolerance to perform collaborative optimization. Optimize to update the expected value of the production variable and the expected value of the distribution variable, and judge whether the return value of the production variable in response to the updated expected value of the production variable and the return value of the distribution variable in response to the updated expected value of the distribution variable meet the production and operation consistency constraint, until the new response of the production variable return value and the return value of the distribution variable meet the production and operation consistency constraint, and obtain the linkage decision result. The CO algorithm adopts a dynamic relaxation factor to update the dynamic tolerance. The linkage decision result includes the production variable return value and the distribution variable return value that meet the production and operation consistency constraint. The multi-decision unit problem is solved in a distributed manner by adopting a collaborative optimization algorithm, and a coupling function is used for consistency convergence, so that the decision calculation speed is greatly improved. Then, by adopting a dynamically updated relaxation factor, the time consumption of collaborative optimization is reduced, the decision efficiency is improved, and the solution of dynamic collaborative decision-making for production and distribution in the relevant technology is solved, so that the intermittent production enterprises are difficult to find the best in the process of production and operation linkage decision-making, and the problem of difficulty in real-time decision-making under dynamic interference.

[0011] Details of one or more embodiments of the present application are set forth in the following drawings and description to make other features, objects, and advantages of the present application more readily apparent. BRIEF DESCRIPTION OF THE DRAWINGS

[0012] The drawings described herein are used to provide a further understanding of the present application and constitute a part of the present application. The illustrative embodiments of the present application and their descriptions are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings:

[0013] Figure 1 is a flow chart of an intelligent linkage decision-making method for production and transportation according to an embodiment of the present application;

[0014] Figure 2 is a flowchart of a production scheduling planning method according to an embodiment of the present application;

[0015] Figure 3 is a flow chart of a transportation planning decision method according to an embodiment of the present application;

[0016] Figure 4 A structural block diagram of an intelligent linkage decision-making device for production and transportation according to an embodiment of the present application;

[0017] Figure 5 This is a structural block diagram of a production scheduling planning device used in a production planning subsystem according to an embodiment of the present application.

[0018] Figure 6 It is a structural block diagram of a transportation planning decision device used in a transportation planning subsystem according to an embodiment of the present application. DETAILED DESCRIPTION

[0019] In order to make the purpose, technical solutions and advantages of the present application clearer, the present application is described and illustrated in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not intended to limit the present application. Based on the embodiments provided in the present application, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present application. In addition, it can also be understood that although the efforts made in this development process may be complex and lengthy, for ordinary technicians in the field related to the contents disclosed in the present application, some changes such as design, manufacturing or production based on the technical contents disclosed in the present application are only conventional technical means, and should not be understood as insufficient contents disclosed in the present application.

[0020] Reference to "embodiments" in this application means that a particular feature, structure, or characteristic described in conjunction with the embodiments may be included in at least one embodiment of the present application. The appearance of the phrase in various locations in the specification does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment that is mutually exclusive with other embodiments. It is explicitly and implicitly understood by those of ordinary skill in the art that the embodiments described in this application may be combined with other embodiments without conflict.

[0021] Unless otherwise defined, the technical terms or scientific terms involved in this application should be understood by people with ordinary skills in the technical field to which this application belongs. The words "one", "a", "a", "the" and the like involved in this application do not indicate a quantitative limitation, and may represent the singular or plural. The terms "include", "comprise", "have" and any of their variations involved in this application are intended to cover non-exclusive inclusions; for example, a process, method, system, product or device that includes a series of steps or modules (units) is not limited to the listed steps or units, but may also include steps or units that are not listed, or may also include other steps or units inherent to these processes, methods, products or devices. The "multiple links" involved in this application refer to links greater than or equal to two. "And / or" describes the association relationship of associated objects, indicating that there may be three relationships, for example, "A and / or B" may represent: A exists alone, A and B exist at the same time, and B exists alone. The terms "first", "second", "third" and the like involved in this application are only used to distinguish similar objects, and do not represent a specific ordering of objects.

[0022] This embodiment provides an intelligent linkage decision-making method for production and transportation running in a production-transportation collaboration system. Figure 1 is a flow chart of an intelligent linkage decision-making method for production and transportation according to an embodiment of the present application, such as Figure 1 As shown, the process includes the following steps:

[0023] Step S101, after completing the current collaborative optimization, the updated expected values ​​of production variables and the expected values ​​of distribution variables are sent to the production planning subsystem and the transportation planning subsystem respectively.

[0024] In this embodiment, after executing a collaborative optimization, the production and transportation collaboration system will generate new production variable expected values ​​and new distribution variable expected values, and issue them to the production planning subsystem and the transportation planning subsystem accordingly, so that the production planning subsystem generates a new production scheduling plan (corresponding to the new production variable return value) according to the new production variable expected value, and the transportation planning subsystem generates a new transportation plan (corresponding to the distribution variable return value) according to the new distribution variable expected value; in this embodiment, the production variable return value corresponding to the production scheduling plan includes at least production variables in two dimensions of production cost and related production time, and the corresponding production variable expected value also includes at least production variable expected values ​​in two dimensions of production cost expected value and related production time expected value; the distribution variable return value corresponding to the transportation plan includes at least production variables in two dimensions of distribution cost and related distribution time, and the corresponding distribution variable expected value also includes at least production variables in two dimensions of distribution cost expected value and related distribution time expected value.

[0025] Step S102, receiving the production variable return value returned by the production planning subsystem in response to the expected value of the production variable and the distribution variable return value returned by the transportation planning subsystem in response to the expected value of the distribution variable, and judging whether the production variable return value and the distribution variable return value satisfy the preset production and transportation consistency constraints, wherein the production variable return value is used to represent the production scheduling parameters for batch production planning of multiple production orders, and the distribution variable return value is used to represent the distribution parameters for combined delivery planning of multiple production orders to be delivered.

[0026] Step S103, when it is determined that the production variable return value and the distribution variable return value do not satisfy the preset production and operation consistency constraint, repeatedly execute collaborative optimization using the preset collaborative optimization CO algorithm and dynamic tolerance to update the production variable expected value and the distribution variable expected value, and determine whether the production variable return value in response to the updated production variable expected value and the distribution variable return value in response to the updated distribution variable expected value satisfy the production and operation consistency constraint, until the new response production variable return value and the distribution variable return value satisfy the production and operation consistency constraint, and obtain a linkage decision result, wherein the CO algorithm uses a dynamic relaxation factor to update the dynamic tolerance, and the linkage decision result includes the production variable return value and the distribution variable return value that satisfy the production and operation consistency constraint.

[0027] In this embodiment, after the newly generated production variable return value and distribution variable return value are obtained, a judgment will be made whether the production and transportation consistency constraints are met. If the newly generated production variable return value and distribution variable return value meet the production and transportation consistency constraints, the corresponding production variable return value and distribution variable return value will be output as the collaborative optimization result. If not, the collaborative optimization will continue to be performed. In the process of continuing to perform the collaborative optimization, the dynamic tolerance is first updated according to the preset formula, and the collaborative optimization is performed based on the updated dynamic tolerance to generate new production variable expected values ​​and distribution variable expected values, and they are issued to the production planning subsystem and the transportation planning subsystem accordingly. Then, before the production and transportation consistency constraints are met, the issuance of the corresponding production variable expected values ​​and distribution variable expected values, the receipt of the production variable return value and distribution variable return value generated according to the issued production variable expected values ​​and distribution variable expected values, and the judgment of whether the generated production variable return value and distribution variable return value meet the production and transportation consistency constraints are repeated.

[0028] In this embodiment, whether the production variable return value and the distribution variable return value meet the production and operation consistency constraint is judged, and the deviation between the result of the current collaborative optimization and the result of the previous collaborative optimization is less than the set production and operation consistency constraint threshold as the judgment criterion.

[0029] Step S104, when it is determined that the production variable return value and the distribution variable return value satisfy the preset production and transportation consistency constraint, the currently received production variable return value and distribution variable return value are used as the linkage decision result.

[0030] In this embodiment, after completing the judgment of the production-operation consistency constraint and determining that the newly generated production variable return value and distribution variable return value meet the production-operation consistency constraint, the corresponding production variable return value and distribution variable return value are output as collaborative optimization results.

[0031] In some optional implementations, before collaborative planning is performed, initial parameter settings are performed, including: the goal of the production and transportation collaboration system layer, the initial solution setting, the dynamic tolerance, and the production and transportation consistency constraint. Specifically, the goal of the production and transportation collaboration system layer is first determined, that is, the optimization goal of the production and distribution linkage link, such as: minimizing the total cost; then, after determining the objective function, mathematical models corresponding to production scheduling and transportation planning are established respectively and their constraints are determined; the two corresponding mathematical scheduling models are integrated into the collaborative optimization CO algorithm model. In this embodiment, the objective function of the production and transportation collaboration system layer is: minTC 1 +TC 2 , the constraints are: , as follows: TI 1 -TI 2 =0,TO 1 -TO 2 =0, where TC 1 is the expected total production cost assigned to the production planning subsystem by the production-operation collaboration system layer, TI 1 is the expected value set of offline time assigned to the production planning subsystem by the production-operation collaboration system layer, TO 1 is the expected value of the delivery time set assigned to the production planning subsystem by the production-transportation collaboration system layer, TC ps is the total production cost return value uploaded by the production planning subsystem, TI ps It is the return value of the offline time set uploaded by the production planning subsystem. ps It is the return value of the delivery time set uploaded by the production planning subsystem. 2 is the expected value of the total distribution cost assigned to the transportation planning subsystem by the production-transportation collaboration system layer, TI 2 is the expected value set of the storage time assigned to the transportation planning subsystem by the production and transportation collaboration system layer. 2 is the expected value set of delivery time assigned to the transportation planning subsystem by the production-transportation collaboration system layer, TC ts is the total delivery cost return value uploaded by the transportation planning subsystem, TI ts It is the return value of the storage time set uploaded by the transportation planning subsystem.ts It is the return value of the delivery time set uploaded by the transportation planning subsystem. It is the variable tolerance of the production-transportation collaboration system layer for the consistency constraints of the coupled variables total cost, offline time / warehousing time, and outbound time / shipping time of the production planning subsystem and the transportation planning subsystem.

[0032] In this embodiment, the production-transportation collaboration system layer allocates the expected values ​​of the corresponding coupling variables to the production planning subsystem and the transportation planning subsystem. The coupling variable between the production-transportation collaboration system layer and the production planning subsystem is the total production cost TC 1 , offline time collection TI 1 , delivery time collection TO 1 ; The coupling variable between the production and transportation collaboration system layer and the transportation planning subsystem is the total distribution cost TC 2 , offline time collection TI 2 , delivery time collection TO 2 ,For the CO algorithm, the design of coupling variables is also crucial, which is a prerequisite for ensuring that the system layer can be solved and that the consistency between systems can be met; at the same time, in this embodiment, the total cost corresponding to the production and transportation collaboration system layer is obtained by adding the total production cost and the total distribution cost, which requires obtaining the return value of the production variable and the return value of the distribution variable, so the total production cost TC 1 And the total delivery cost TC 2 Set as coupled variable; take the related time set as coupled variable.

[0033] Through the above steps S101 to S104, after completing the current collaborative optimization, the updated expected values ​​of production variables and distribution variables are sent to the production planning subsystem and the transportation planning subsystem respectively; the production variable return values ​​returned by the production planning subsystem in response to the expected values ​​of production variables and the distribution variable return values ​​returned by the transportation planning subsystem in response to the expected values ​​of distribution variables are received, and it is judged whether the production variable return values ​​and the distribution variable return values ​​meet the preset production and transportation consistency constraints, the production variable return values ​​are used to characterize the production scheduling parameters for batch production planning of multiple production orders, and the distribution variable return values ​​are used to characterize the distribution parameters for combined transportation planning of multiple production orders to be distributed; when it is judged that the production variable return values ​​and the distribution variable return values ​​do not meet the preset production and transportation consistency constraints, the collaborative optimization using the preset collaborative optimization CO algorithm and dynamic tolerance is repeatedly performed to update the production variable expected values ​​and the distribution variable expected values, and to judge the production variables in response to the updated production variable expected values. Whether the return value and the distribution variable return value in response to the updated distribution variable expected value meet the production and transportation consistency constraint, until the new response production variable return value and distribution variable return value meet the production and transportation consistency constraint, the linkage decision result is obtained, and the CO algorithm adopts dynamic relaxation factor to update the dynamic tolerance. The linkage decision result includes the production variable return value and distribution variable return value that meet the production and transportation consistency constraint; when it is judged that the production variable return value and the distribution variable return value meet the preset production and transportation consistency constraint, the currently received production variable return value and distribution variable return value are used as the linkage decision result, and the multi-decision unit problem is solved in a distributed manner by adopting a collaborative optimization algorithm, and a coupling function is used for consistency convergence, so that the decision calculation speed is greatly improved, and then by adopting a dynamically updated relaxation factor, the time consumption of collaborative optimization is reduced, the decision efficiency is improved, and the dynamic collaborative decision-making solution for production and distribution in the relevant technology is solved, so that intermittent production enterprises are difficult to find the best in the process of production and transportation linkage decision-making, and the problem of difficulty in real-time decision-making under dynamic interference.

[0034] In some embodiments, collaborative optimization is performed using a preset collaborative optimization CO algorithm and a dynamic tolerance to update the expected value of the production variable and the expected value of the distribution variable, including the following steps:

[0035] Step 111, after receiving the current corresponding production variable return value and distribution variable return value, determine the production sub-variable group and distribution sub-variable group corresponding to the production variable return value and the distribution variable return value respectively, wherein the production sub-variable group includes the total production cost, the production offline time set and the production outbound time set, and the distribution sub-variable group includes the total distribution cost, the distribution inbound time set and the distribution delivery time set, the production offline time set is used to measure the total production cost, the distribution delivery time set is used to determine the total distribution cost, and the production outbound time set is used to cooperate with the distribution inbound time to constrain the production, delivery and inbound of the corresponding products.

[0036] In this embodiment, the corresponding production variable return value and expected value will map the total production cost / total production cost expected value, production offline time set / production offline time expected value set and production outbound time set / production outbound time expected value set corresponding to the corresponding production scheduling arrangement. The corresponding offline time set is composed of the offline times of multiple batch orders, that is, multiple offline time values, and the corresponding outbound time set is composed of the outbound times of multiple batch orders, that is, multiple outbound time values; similarly, the corresponding distribution variable return value and expected value will map the corresponding distribution total cost / total distribution cost expected value, distribution entry time set / distribution entry time expected value set and distribution shipment time set / distribution shipment time expected value set corresponding to the corresponding transportation planning. It should be understood that the corresponding distribution entry time is sent to the transportation planning subsystem by the production and transportation collaboration system and the same distribution variable expected value. In this embodiment, the distribution entry time set is equivalent to the distribution entry time expected value set, which is also That is, the warehousing time of the products of a distribution batch is equal to the expected value of the corresponding warehousing time. At the same time, the products of the distribution batch can only be shipped after they are put into storage, that is, the delivery and shipment time of the products of each distribution batch is after the delivery warehousing time, and earlier than the delivery deadline (customer demand time) of the products of the distribution batch; it should be understood that when the distribution vehicle delivers products of more than two distribution batches, the delivery and shipment time of the corresponding distribution batch is the latest delivery and shipment time among the delivery and shipment times of the more than two distribution batches; in this embodiment, since all production orders are released at time 0, it is only necessary to know the production offline time of the product to infer the production cost, and the batch order is used for the warehousing of the distribution batch only after it is shipped out of the warehouse. Therefore, the production and distribution stages are coupled through the production outbound time and the distribution warehousing time; for the distribution process, when the product delivery warehousing time is known, it is only necessary to know the distribution and shipment time of the product to reflect the distribution cost.

[0037] In this embodiment, the production sub-variable group represents the relevant variables of the cost dimension and the time dimension in the batch production stage of the order group, and the distribution sub-variable group represents the relevant variables of the cost dimension and the time dimension in the batch distribution stage of the production order; in this embodiment, the received production variable return value includes the total production cost return value, the production offline time set return value and the production warehouse time set return value, and the received distribution variable return value includes the total distribution cost return value, the distribution warehouse time set return value and the distribution delivery time set return value.

[0038] Step 112, respectively determine the dynamic relaxation factors corresponding to each production sub-variable in the production sub-variable group and each distribution sub-variable in the distribution sub-variable group, and determine the dynamic tolerance corresponding to each production sub-variable and each distribution sub-variable based on the dynamic relaxation factors.

[0039] In this embodiment, the corresponding dynamic tolerance of each production sub-variable or each distribution sub-variable in the linkage decision-making process is different. Therefore, each dynamic relaxation factor is a production sub-variable or a distribution sub-variable. It should be understood that the dynamic relaxation factor is different in different linkage decision-making processes, but a set of dynamic relaxation factors are updated before the start of the current linkage decision and remain unchanged during the current linkage decision-making process. The dynamic relaxation factor and the corresponding dynamic tolerance will only change during repeated iterative linkage decision-making processes.

[0040] Step 113, based on each production sub-variable and the corresponding dynamic tolerance, the production sub-variable expected values ​​corresponding to all production sub-variables are corrected to obtain updated production variable expected values, and based on each distribution sub-variable and the corresponding dynamic tolerance, the distribution sub-variable expected values ​​corresponding to all distribution sub-variables are corrected to obtain updated distribution variable expected values.

[0041] In this embodiment, after determining the corresponding dynamic tolerance, since the value of the production sub-variable or the distribution sub-variable is known, the corresponding expected value of the coupling variable can be corrected according to the value of the production sub-variable or the distribution sub-variable, the dynamic tolerance and the corresponding constraints. For example: the expected value of the coupling variable of the current production sub-variable or the distribution sub-variable before correction is a, the dynamic tolerance before update is b, and the dynamic tolerance after update is c, then the minimum correction value corresponding to a is cd.

[0042] Through steps 111 to 113, the CO algorithm and dynamic tolerance are used to make joint decisions to update the expected values ​​of production variables and distribution variables, so that the production planning subsystem and the transportation planning subsystem can be locally optimized respectively until the solution of the entire production and transportation decision is optimal.

[0043] In some embodiments, determining whether the corresponding production variable return value and delivery variable return value meet the preset production and delivery consistency constraint includes the following steps:

[0044] Step 121, determine the expected total production cost, the expected production offline time, and the expected production delivery time corresponding to the expected production variable values, and determine the expected total distribution cost, the expected distribution entry time, and the expected distribution delivery time corresponding to the expected distribution variable values.

[0045] Step 122, calculate the deviation values ​​corresponding to the total production cost and the expected value of the total production cost, the production off-line time set and the production off-line time expected value set, the production outbound time set and the production outbound time expected value set, the total distribution cost and the expected value of the total distribution cost, the distribution entry time set and the distribution entry time expected value set, and the distribution shipment time set and the distribution shipment time expected value set, and after normalizing the calculated first deviation, second deviation, third deviation, fourth deviation, fifth deviation and sixth deviation values, weight all normalized deviations to obtain the first total production and transportation consistency difference.

[0046] In this embodiment, in order to eliminate the dimensional and order of magnitude differences between different deviation values, each deviation value is standardized. It can be understood that the standardization processing in the embodiment of the present application can be any standardization processing method in the prior art, but is not limited to the normalization processing adopted in the embodiment of the present application; in this embodiment, after the multiple deviation values ​​are standardized, the multiple standardized parameters are weighted to obtain a quantitative value of the linkage decision result with the production variable return value and the distribution variable return value, that is, the first total production and transportation consistency difference.

[0047] Step 123, based on the expected value of the production variable, the expected value of the distribution variable and the received production variable return value and distribution variable return value updated by the collaborative optimization before the current one, calculate the second total production-operation consistency difference corresponding to the received production variable return value and the distribution variable return value, and determine whether the difference between the first total production-operation consistency difference and the second total production-operation consistency difference is less than the production-operation consistency constraint threshold.

[0048] In this embodiment, the second total production and transportation consistency difference is the quantitative value of the linkage decision result corresponding to the production variable return value and the distribution variable return value corresponding to the current one; in this embodiment, the following formula is used to judge the consistency constraint: zz′<δ, where z represents the first total production and transportation consistency difference, z′ represents the second total production and transportation consistency difference, and δ is the production and transportation consistency constraint threshold, and δ is set to 10 -4 .

[0049] Step 124, when it is determined that the difference between the first total production and operation consistency difference and the second total production and operation consistency difference is less than the production and operation consistency constraint threshold, determine whether the corresponding production variable return value and distribution variable return value meet the preset production and operation consistency constraint.

[0050] Through the above steps 121 to 124, by comparing the size of the total production-transportation consistency difference corresponding to the two corresponding production variable return values ​​and distribution variable return values, it is determined whether the corresponding production variable return value and distribution variable return value meet the production-transportation consistency constraint, and the linkage decision-making results are quantified in terms of cost and time, thereby improving the efficiency of linkage decision-making and reducing the search for non-optimal solutions in the linkage decision-making process.

[0051] This embodiment provides a production scheduling planning method running in a production planning subsystem. Figure 2 is a flowchart of a production scheduling planning method according to an embodiment of the present application, such as Figure 2 As shown, the process includes the following steps:

[0052] Step S201, obtain the production order information of the current corresponding batch production, and perform initial population coding on the production order information to generate multiple first double-coded individuals, wherein the first double-coded individual includes multiple first double-coded sub-units, and one first double-coded sub-unit includes multiple double codes in the same production workshop, and each double code includes an order code and a batch code that constitute a code pair, the order code is used to characterize the order to which the corresponding product belongs, and the batch code is used to characterize the batching of products corresponding to the order code.

[0053] In this embodiment, the production planning subsystem performs a solution based on the genetic GA algorithm according to the received expected values ​​of the production variables and the production orders of the batch production required for each local optimization. In this embodiment, the first dual-coded individual is the dual-coded individual corresponding to the local solution at that time; in this embodiment, before performing the solution based on the GA algorithm, the parameters are initialized first, for example: setting the population size P, the iteration number G, the crossover probability Pc, the mutation probability Pm, and the disturbance probability Pb; the weight factor: Customer orders (including but not limited to product types, demand, and delivery dates); equipment information (including but not limited to the number of equipment, size, and unit cost) and production and warehousing information; In this embodiment, the initial solution generated is to encode the customer order and batch to generate the chromosomes corresponding to the corresponding double-coded individuals, for example: The first line of the double-coded individual represents the order code, the second line represents the batch code, and the partition in the middle represents different production workshops. For example, the left side represents production workshop 1, and the right side represents production workshop 2. The order code represents different customer orders, and the batch code represents the combined batch. Starting from the first digit, if the next digit is the same, a batch is formed. For example, [1 1] represents the same batch. If the next digit is different, different batches are formed. For example, [1 0] represents a new batch starting from 0. For example, [1 1 0 1] means [1 1] constitutes batch 1, [0] constitutes batch 2, and [1] constitutes batch 3. Represents a double code, for a double-coded individual Converted to batch form, Among them, it means that in production workshop 1, order 1 and order 3 form the first batch, and order 2 and order 4 form the second batch; in production workshop 2, order 2 forms the first batch, order 3 forms the second batch, and order 4 and order 1 form the third batch.

[0054] Step S202: selecting a set number of second dual-coded individuals from a plurality of first dual-coded individuals based on the individual fitness corresponding to the first dual-coded individuals and a preset selection strategy.

[0055] In this embodiment, each first double-coded individual is individually evaluated, and the corresponding double-coded individual is decoded according to the set operation table, and the total production cost, production offline time set, and production outbound time set of the target function are obtained from the decoded production information in a simulated production manner, and the individual fitness corresponding to the first double-coded individual is obtained according to the corresponding target function; in this embodiment, the individual fitness is used as the screening parameter, and the selection strategy of elite retention ES-tournament selection TS is adopted to select a set number of second double-coded individuals from multiple first double-coded individuals, that is, firstly, a part (the number is less than the total number of first double-coded individuals) of the first double-coded individuals are selected from multiple first double-coded individuals using the ES strategy for retention, and then, multiple optimal first double-coded individuals are selected based on the TS strategy, so that the population size corresponding to the second double-coded individuals is made consistent with the population size corresponding to the first double-coded individuals in a complementary manner; in this embodiment, a part is selected from multiple first double-coded individuals using the standard roulette method, and the optimal individual is selected from the part for retention, and then a part of individuals is selected and the optimal individual is selected from the part for retention repeatedly until a set number of second double-coded individuals is obtained, and the so-called set number refers to the population size initially set.

[0056] Step S203, performing genetic evolution iteration operation on the order codes and batch codes corresponding to all second dual-coded individuals until a target dual-coded individual is generated, and decoding the target dual-coded individual to generate a production variable return value, wherein the genetic evolution operation includes a crossover operation, a mutation operation, and a disturbance operation, and the individual fitness is determined by solving a preset production objective function, and the production objective function is constructed by weighting the difference between the received expected value of the production variable and the production variable return value generated by decoding the first dual-coded individual.

[0057] In this embodiment, the crossover mutation of the double code of the second double code individual is performed with a certain crossover probability P c , mutation probability P m , determine whether to perform crossover and mutation operations on the double codes of the second double-coded individual. In this embodiment, the crossover probability p c Take a fixed value, and the mutation probability p m It starts large and then iterates and gradually shrinks.

[0058] In this embodiment, the crossover operation on the dual codes of the second dual code individual is as follows:

[0059]

[0060]

[0061] Among them, P is the parent generation, V is the offspring generation, and the offspring individuals are obtained by exchanging the black parts of the two parents and correcting the duplication value of the order chromosomes of the intermediate individuals after the exchange. It should be noted that only chromosomes in the same production workshop will undergo mutual crossover operations.

[0062] In this embodiment, the double code of the second double code individual is mutated as follows:

[0063]

[0064]

[0065] Among them, P is the parent generation and V is the offspring generation. The offspring individuals are obtained by changing the black part of the parent generation and correcting the repetition value of the order chromosome of the intermediate individual after mutation.

[0066] In this embodiment, with a certain probability p b , respectively, the product code of the second double-coded individual is inverted, the block is swapped, the first position is swapped, and the batch code is flipped 0-1, inverted, greedily swapped, and the like; in this embodiment, the disturbance probability p b It starts large and then iterates and gradually shrinks.

[0067] In this embodiment, the flipping operation of the product code or batch code is as follows:

[0068]

[0069]

[0070] Among them, P is the parent generation and V is the offspring generation. The black part of the parent generation is selected randomly, and then the offspring individuals are generated by reversing the order, for example, reversing [1 3 2 4] to [4 2 3 1].

[0071] In this embodiment, the product code or batch code is interchanged as follows:

[0072]

[0073]

[0074] Among them, P is the parent generation, V is the offspring generation, and the black part of the parent generation is selected by random selection. For example, the two coding genes (numbers) of the order chromosome and the batch chromosome are selected respectively, and they are directly exchanged to generate offspring individuals.

[0075] In this embodiment, the movement operation of the product code or batch code is as follows:

[0076]

[0077]

[0078] Among them, P is the parent generation and V is the offspring generation. The black part of the parent generation is selected by random selection. For example, [2 4] in the parent generation is selected, and then the whole is moved one position to the left to generate the offspring individuals.

[0079] In this embodiment, the area change operation for the product code or batch code is as follows:

[0080]

[0081]

[0082] Among them, P is the parent generation and V is the offspring generation. The black part of the parent generation is selected by random selection. For example, for two different regions of a batch of chromosomes, coding genes with the same length are selected and then exchanged as a whole to generate offspring individuals.

[0083] In this embodiment, the reversal operation of the product code or batch code is as follows:

[0084]

[0085]

[0086] Among them, P is the parent generation and V is the offspring generation. The black part of the parent generation is selected randomly. Since the operation bit 0-1 is reversed, it will only operate on batches of chromosomes containing only "0" and "1". By selecting certain genes of the batch coding sub-genes and exchanging their "1" and "0", the offspring individuals are formed.

[0087] In this embodiment, the greedy operation on the product code or batch code is as follows:

[0088]

[0089]

[0090] Among them, P is the parent generation and V is the offspring generation. The black part of the parent generation, such as [1; 1], is selected randomly, and the offspring individuals are generated by greedily moving it to all other positions and leaving it at the position with the lowest objective function.

[0091] Through the above steps S201 to S203, the production order information of the current corresponding batch production is obtained, and the initial population encoding is performed on the production order information to generate multiple first double-coded individuals; based on the individual fitness corresponding to the first double-coded individuals and the preset selection strategy, a set number of second double-coded individuals are selected from the multiple first double-coded individuals; the genetic evolution iterative operation is performed on the order codes and batch codes corresponding to all the second double-coded individuals until the target double-coded individual is generated, and the target double-coded individual is decoded to generate a production variable return value, so as to generate the corresponding production variable return value in response to the expected value of the production variable.

[0092] In some embodiments, the selection strategy includes an ES-TS hybrid strategy, which selects a set number of second dual-coded individuals from a plurality of first dual-coded individuals based on the individual fitness corresponding to the first dual-coded individuals and a preset selection strategy, and is implemented by the following steps:

[0093] Step 211, determine the individual fitness corresponding to each first dual-coded individual, and based on the individual fitness, use the elite retention ES strategy to select a set number of elite dual-coded individuals from multiple first dual-coded individuals, where the elite dual-coded individuals include first dual-coded individuals whose individual fitness is less than the fitness threshold.

[0094] In this embodiment, the ES strategy refers to elite selection, that is, directly selecting the best part of the population from the parent generation for retention. It should be noted that elite retention will greatly increase the result and prevent it from falling into the dilemma of local optimality. In this embodiment, only individuals from no more than 3 different parent generations are selected for retention.

[0095] Step 212, based on individual fitness, using a standard roulette algorithm and a set selection probability, select a set number of candidate dual-coded individuals from multiple first dual-coded individuals, and select the best dual-coded individual from the selected candidate dual-coded individuals based on the tournament TS strategy.

[0096] In this embodiment, the TS strategy refers to tournament selection, that is, a part of the population is selected according to the selection probability through the standard roulette method, and a tournament is held in the population to select the best individual for retention. It can be understood that the TS strategy can well avoid the shortcomings of the traditional roulette method, so that the results of individual screening are more in line with expectations. In this embodiment, the standard roulette method is used to select two individuals for the tournament; in some optional implementations, the selection probability is calculated according to the following formula: Among them, p(x i ) represents the selection probability, F(x i ) indicates the corresponding individual.

[0097] Step 213, repeatedly selecting a set number of candidate dual-coded individuals from multiple first dual-coded individuals and selecting the best dual-coded individual from the selected candidate dual-coded individuals to obtain multiple best dual-coded individuals, and using all the selected elite dual-coded individuals and multiple best dual-coded individuals as second dual-coded individuals.

[0098] In some embodiments, generating a corresponding production variable return value includes the following steps:

[0099] Step 221, obtaining a preset work table, wherein the work table includes batch information, batch order information and production information as well as the correspondence information among the three, the batch order information includes at least one order code, and the production information includes the production batch of the batch-to-order corresponding to the batch order information and the delivery time of the batch order.

[0100] In this embodiment, the job table is as follows:

[0101] Batch information Batch order information Production batch Delivery time 1 [1 3] 10 19:00 2 [3 2] 7 17:00 .

[0102] Step 222, based on all order codes and all batch codes corresponding to each first double-coded sub-unit of the current double-coded individual to be decoded, generate the batch order corresponding to each first double-coded sub-unit and the batch information corresponding to each batch pair, and query the production information corresponding to the batch order and batch information corresponding to each first double-coded sub-unit in the job table, wherein the current double-coded individual includes one of the first double-coded individual and the target double-coded individual.

[0103] Step 223, based on the batch information corresponding to each first double-coded sub-item, the batch order information, the production batch and delivery time corresponding to the batch order information, simulated production deduction is performed to deduce the completion time and single-batch production cost corresponding to each batch order, and the completion time and single-batch production cost corresponding to all batch orders corresponding to each current double-coded individual are counted separately to obtain the production offline time set and total production cost corresponding to each current double-coded individual.

[0104] Step 224, from the time interval determined by the completion time and delivery time corresponding to each batch order, select the time with the smallest difference from the target outbound time expectation value, obtain the outbound time corresponding to each batch order, count the outbound times corresponding to all batch orders corresponding to each current double-coded individual, and obtain the production outbound time set corresponding to each current double-coded individual cost, wherein the target outbound time expectation value is an outbound time expectation value corresponding to the corresponding batch order in the outbound time expectation value set of the production variable expectation value, and the corresponding production variable return value includes the total production cost, production offline time set and production outbound time set corresponding to the current double-coded individual.

[0105] In some optional implementations, the simulated production simulation includes the following steps:

[0106] Step 1: Get a corresponding dual-coded individual and set time t to 0, then execute step 2.

[0107] Step 2: Determine whether there are any corresponding double-coded individuals that have not completed the batch order. If so, execute step 3 and jump to the next double-coded individual.

[0108] Step 3: Determine whether the next batch of batch orders can be arranged for machine production. If so, proceed to step 4; otherwise, proceed to step 5.

[0109] Step 4: Arrange the next batch of batch orders to the machine for production.

[0110] Step 5: Jump time t to the node where the next event is completed, that is, the completion time of the most recent equipment, and execute step 2, and then execute step 6.

[0111] Step 6: When all batch orders of the corresponding double-coded individuals are completed, time t jumps to the node where the next event is completed, that is, the completion time of all equipment.

[0112] Through the simulation production deduction from step 1 to step 6, the total production cost, production offline time set, and production delivery time set used to calculate the objective function can be obtained.

[0113] In some embodiments, determining the individual fitness corresponding to each first dual-coded individual comprises the following steps:

[0114] Step 231, from the production variable return value corresponding to the first double-coded individual, obtain the corresponding total production cost, production offline time set and production delivery time set.

[0115] Step 232, respectively calculate the deviation values ​​corresponding to the total production cost corresponding to the first double-coded individual and the corresponding expected value of the total production cost, the production off-line time set and the corresponding expected value set of the production off-line time, and the production delivery time set and the expected value set of the production delivery time, and obtain the seventh deviation value, the eighth deviation value and the ninth deviation value.

[0116] Step 233, after normalizing the seventh deviation value, the eighth deviation value and the ninth deviation value, weighting the normalized deviation values ​​to obtain the individual fitness of the first dual-coded individual.

[0117] In this embodiment, by simulating production deduction, the total production cost, production offline time set, and production delivery time set used to calculate the objective function are obtained, and the following objective function is constructed based on the expected value of the production variable as the individual fitness:

[0118]

[0119] Among them, TC 1 is the expected total production cost, TI 1 is the expected value set of offline time, TO 1 is the expected value set of the outbound time, TC ps is the total production cost, TI ps is the offline time set, TO ps Is the collection of outbound time, Represents the weight of the target corresponding to each coupling variable.

[0120] This embodiment provides a transportation planning decision method running in the transportation planning subsystem. Figure 3 is a flow chart of a method for transport planning decision making according to an embodiment of the present application. Figure 3 As shown, the process includes the following steps:

[0121] Step S301, obtain the incoming order information corresponding to the current delivery, and encode the incoming order information to generate an initial double code body, wherein the initial double code body includes an order code sequence with multiple delivery order coders and a delivery code sequence with multiple vehicle assembly coders, one delivery order coder corresponds to one vehicle assembly coder, and forms a delivery double coder, the delivery order coder is used to characterize the code of the corresponding incoming order, and the order of the delivery order coder in the corresponding order code sequence is used to characterize the priority of the delivery order, the vehicle assembly coder is used to characterize the type of delivery vehicle selected when the corresponding incoming order is delivered, and one vehicle assembly coder also maps at least one delivery path, and the nodes of the delivery path include the initial position of the delivery vehicle and the storage location of the product corresponding to the incoming order.

[0122] In this embodiment, after receiving the corresponding expected value of the distribution variable, the transportation planning subsystem performs initial parameter settings, including the expected value of the distribution variable, the temperature coefficient T max 、T min , f; evaluation coefficient Sigma i , W max , W min ; Weight factor: Customer order, location information, vehicle information, and then generate an initial solution including the initial dual encoding body according to the rules.

[0123] In this embodiment, the initial double code body P is represented as follows: Among them, the first row represents the order coding sequence, and each column of the first row represents a delivery order code, that is, the delivery order corresponding to each incoming order. The order of the delivery order code in the corresponding order coding sequence corresponds to the priority of the delivery order of the incoming order. For example: the number in the first column is 1, which means that the priority of incoming order 1 in the delivery order is 1, the number in the second column is 5, which means that the priority of incoming order 5 in the delivery order is 2, and the number in the third column is 2, which means that the priority of incoming order 2 in the delivery order is 3; if incoming order 1 and incoming order 3 are delivered at the same time, since the priority of incoming order 1 is 1, which is higher than the priority of incoming order 3, incoming order 1 will be delivered first and then Delivery inbound order 3; the second row represents the delivery code sequence, and each column in the second row represents the vehicle group code. For example, inbound order 1 and inbound order 3 will select the same delivery vehicle type 1; inbound order 5 and inbound order 2 will select the same delivery vehicle type 2; inbound order 6 and inbound order 4 will select the same delivery vehicle type 3, and so on; according to the principle of grouping as much as possible, the inbound orders are combined under the condition of vehicle capacity restrictions. For example, if delivery vehicle type 3 can load inbound order 6 and inbound order 4 at one time, they will be delivered uniformly to form a delivery batch. For another example, if delivery vehicle type 1 cannot load inbound order 1 and inbound order 3 at one time, two vehicles will be sent for separate transportation. The vehicle group code is as follows:

[0124] .

[0125] Step S302, using the linear sorting selection LRS strategy and the control line CL strategy, respectively select the target destruction operation, the target repair operation and the target disturbance operation from the multiple destruction operations, the multiple repair operations and the multiple disturbance operations corresponding to the adaptive large neighborhood search ALNS algorithm, and use the target destruction operation, the target repair operation and the target disturbance operation in turn to destroy, repair and disturb all the delivery double coders of the initial double coder and the delivery order coders and the vehicle assembly coders corresponding to the delivery double coders to generate an alternative delivery double coder, wherein the destruction operation includes one of the following: head and tail truncation, random removal, random block cutting, random excision, the repair operation includes one of the following: random block filling, random gap filling, head and tail gap filling, greedy gap filling, and the disturbance operation includes one of the following: transposition, block shifting, flipping, order change, greedy replacement.

[0126] In this embodiment, the LRS-CL hybrid strategy is used to select the target destruction operation, the target repair operation and the target disturbance operation. It can be understood that during the execution of the adaptive large neighborhood search (Adaptive Large Neighborhood Search, referred to as ALNS) algorithm, it is easy to cause the evaluation coefficients between operators to be extremely different. Therefore, the control line CL strategy is first used to control the upper and lower limits of the operation evaluation index, and then the linear ranking selection (Linear rank-selection, referred to as LRS) strategy is used for operation selection. Among them, the LRS strategy means that each operation is sorted from small to large according to the score, and then the operation is selected according to the following formula: i represents the order of operations, N represents the total number of operations, and n + Represents the score of the highest-rated operation, n - Represents the score of the operation with the lowest score. The CL strategy sets upper and lower limits for the scores of operations so that the score gap is not too large. When the score of an operation reaches the threshold, the scores of all operations are shortened to half.

[0127] In this embodiment, all the delivery double coders of the initial double coder are destroyed to generate an intermediate solution. The corresponding destruction operations include: head and tail truncation, random removal, random slicing, and random excision. Among them, head and tail truncation means selecting the initial double coder to remove its first and last delivery double coders, random removal means randomly selecting one or more positions of the initial double coder and removing them, random slicing means randomly selecting one or more areas of the initial double coder and removing them; random excision means randomly selecting the delivery order coder and the vehicle assembly coder of the initial double coder, and selecting a certain position to remove it. Remove; In this embodiment, a repair operation is performed on the intermediate solution formed by the initial double-encoded body to generate a complete solution (new solution); the repair operations for repairing the intermediate solution include: random block filling, random gap filling, head and tail gap filling, and greedy gap filling, wherein random block filling means that the missing part (damaged part) is filled as a whole in a purely random manner, random gap filling means that the missing parts are filled one by one in a purely random manner; head and tail gap filling means that the missing parts are filled in the order of the first and last positions in a purely random manner; greedy gap filling means that the missing parts are inserted in a greedy manner and the position with the optimal objective function is selected for filling.

[0128] In this embodiment, a target perturbation operation is used to perform perturbation operations on the delivery order encoder and the vehicle assembly encoder in turn, and greedily select; in this embodiment, the perturbation operations that can be used as the target perturbation operation include transposition, block shifting, flipping, order change, and greedy replacement.

[0129] In this embodiment, the transposition operation of the delivery order code and the vehicle assembly code corresponding to the double code body that has been destroyed and repaired is as follows:

[0130]

[0131]

[0132] Among them, P is the parent generation, V is the offspring generation, and the black part of the parent generation is selected by random selection, and then the offspring individuals are generated by exchange.

[0133] In this embodiment, the block shifting operation is performed on the delivery order code and the vehicle assembly code corresponding to the double code body that has been destroyed and repaired as follows:

[0134]

[0135]

[0136] Among them, P is the parent generation, V is the offspring generation, and the offspring individuals are generated by randomly selecting the black part of the parent generation and moving it two positions to the right as a whole.

[0137] In this embodiment, the delivery order code and vehicle assembly code corresponding to the damaged and repaired double code bodies are flipped as follows:

[0138]

[0139]

[0140] Among them, P is the parent generation, V is the offspring generation, and the black part of the parent generation is selected randomly, and then the offspring individuals are generated by reversing the order, for example, reversing [1 2 2 3 3 1] to [1 3 3 2 2 1].

[0141] In this embodiment, the order exchange operation for the delivery order code and the vehicle group code corresponding to the damaged and repaired double code body is as follows:

[0142]

[0143]

[0144] Among them, P is the parent generation, V is the offspring generation, and the black part of the parent generation is selected by random selection. Only two codes above the delivery order code are selected, and the offspring individuals are generated by exchanging their respective positions.

[0145] In this embodiment, the greedy replacement operation is performed on the delivery order code and the vehicle assembly code corresponding to the double code body that has been destroyed and repaired as follows:

[0146]

[0147]

[0148] Among them, P is the parent generation and V is the offspring generation. The black part of the parent generation is selected by random selection, and then all possible results are tested and the individual with the best objective function is greedily selected to generate the offspring individuals.

[0149] Step S303, decode the alternative delivery double code body to obtain the alternative delivery information, and obtain the delivery variable value corresponding to the alternative delivery double code body from the alternative delivery information, and according to the deviation between the delivery variable value and the received expected value of the delivery variable, iterate the destruction, repair and disturbance operations on the alternative delivery double code body until the target delivery double code body is generated.

[0150] Step S304, decode the target delivery double code body to obtain the target delivery information, and generate a delivery variable return value based on the target delivery information.

[0151] Through the above steps S301 to S304, the warehouse order information corresponding to the current delivery is obtained and the warehouse order information is encoded to generate an initial double code body; the linear sorting selection LRS strategy and the control line CL strategy are used to select the target destruction operation, the target repair operation and the target disturbance operation from the multiple destruction operations, the multiple repair operations and the multiple disturbance operations corresponding to the adaptive large neighborhood search ALNS algorithm, and the target destruction operation, the target repair operation and the target disturbance operation are used in turn to destroy, repair and disturb all the delivery double coders of the initial double coder body and the delivery order coders and the vehicle assembly coders corresponding to the delivery double coders. The method comprises the following steps: performing a dynamic operation to generate an alternative delivery double code body; decoding the alternative delivery double code body to obtain the alternative delivery information, and obtaining the delivery variable value corresponding to the alternative delivery double code body from the alternative delivery information, and iterating the destruction, repair and disturbance operations on the alternative delivery double code body according to the deviation between the delivery variable value and the received expected value of the delivery variable, until a target delivery double code body is generated, decoding the target delivery double code body to obtain the target delivery information, and generating a delivery variable return value according to the target delivery information, and using the ALNS algorithm for local solution based on the received expected value of the delivery variable to generate delivery variable values ​​for batch transportation of products in the incoming order.

[0152] In some embodiments, decoding the corresponding delivery double code includes the following steps:

[0153] Step 311, obtaining a distribution parameter table, wherein the distribution parameter table includes distribution batch information, vehicle type information and distribution information as well as the correspondence information among the three, and the distribution information includes distribution route information, distribution batch quantity and delivery deadline of the distribution batch corresponding to the distribution batch information.

[0154] In this embodiment, the delivery parameter table is as follows:

[0155] Delivery batch information Vehicle type information Delivery route information Delivery batch Delivery deadline 1 1 [0]-[1]-[0] 7 19:00 2 1 [0]-[3]-[0] 10 17:00 3 2 [0]-[5]-[2]-[0] 18 19:00|20:00 .

[0156] Step 312, based on all the delivery order codes and vehicle group coders corresponding to the current delivery double code body to be decoded, obtain the delivery batch information of all delivery batches corresponding to the current delivery double code body and the vehicle type information corresponding to each delivery batch, and in the delivery parameter table, query the delivery information corresponding to the corresponding delivery batch based on the delivery batch information and vehicle type information corresponding to each delivery batch, wherein the current delivery double code body includes one of the alternative delivery double code body and the target delivery double code body.

[0157] Step 313, obtain a set of expected delivery entry time values ​​from the received delivery variable expected values, and use the set of expected delivery entry time values ​​as the set of delivery entry time corresponding to the current delivery double code body, wherein each delivery batch corresponds to a delivery entry time in the set of delivery entry time.

[0158] In this embodiment, the corresponding delivery warehousing time is sent to the transportation planning subsystem by the production and transportation collaboration system and the expected value of the same delivery variable. In this embodiment, the delivery warehousing time set is equivalent to the delivery warehousing time expected value set, that is, the warehousing time of a delivery batch of products is equal to the corresponding warehousing time expected value.

[0159] Step 314, based on all the delivery batch information, vehicle type information and the delivery path information, delivery batch size and delivery deadline corresponding to each delivery batch corresponding to the current delivery dual-code body, a delivery deduction is performed to deduce the delivery delivery time and unit delivery cost corresponding to each delivery batch, and the delivery delivery time and unit delivery cost corresponding to all delivery batches are counted to obtain the delivery delivery time set and the total delivery cost corresponding to the current delivery dual-code body, wherein the delivery delivery time is selected from the time interval determined by the delivery warehousing time and the delivery deadline corresponding to each delivery batch, and the difference with the target delivery time expectation value is the smallest, and the target delivery time expectation value includes the delivery delivery time expectation value corresponding to the corresponding delivery batch in the delivery variable expectation value set.

[0160] In this embodiment, the products in the delivery batch can only be shipped after they are put into storage, that is, the delivery and shipment time of the products in each delivery batch is after the delivery and warehousing time, and at the same time earlier than the delivery deadline of the products in the delivery batch (the time required by the customer); it should be understood that when a delivery vehicle delivers products from more than two delivery batches, the delivery and shipment time of the corresponding delivery batch is the latest delivery and shipment time among the delivery and shipment times of the more than two delivery batches.

[0161] Step 315, the total distribution cost, distribution warehousing time set and distribution delivery time set corresponding to the current distribution double code body are used as the distribution variable values ​​corresponding to the corresponding distribution double code, wherein the distribution variable return value includes the distribution variable value corresponding to the target distribution double code body.

[0162] Through steps 311 to 315, the delivery double code is decoded to obtain the corresponding delivery variable value.

[0163] In some of the embodiments, according to the deviation of the delivery variable value from the received delivery variable expected value, the alternative delivery dual encoding body is iteratively destroyed, repaired and disturbed, including the following steps:

[0164] Step 321, obtain at least the corresponding total delivery cost and delivery delivery time set from the delivery variable value corresponding to the current candidate delivery dual code body, and obtain the total delivery cost expected value and the delivery delivery time expected value set from the delivery variable expected value.

[0165] Step 322, normalize the deviation between the total distribution cost and the expected value of the total distribution cost and the deviation between the distribution delivery time set and the expected value set of distribution delivery time, and weight the normalized deviation values ​​to obtain the distribution fitness of the alternative distribution dual code body corresponding to the current one.

[0166] Step 323, determine the delivery fitness of the candidate delivery double-encoded body corresponding to the current one, and select a new delivery double-encoded body solution from the candidate delivery double-encoded body corresponding to the current one and the candidate delivery double-encoded body before the current one according to the delivery fitness and the Metropolis reception criteria, and update the target destruction operation, target repair operation and target disturbance operation.

[0167] In this embodiment, the Metropolis criterion is used to determine whether to accept the new solution. If the new solution is better than the current solution, the new solution is accepted; otherwise, the Metropolis criterion is used to determine whether to accept the new solution, where the acceptance probability is: In this embodiment, initially, all operations have the same evaluation and score. In each iteration, scores are given in a stepped manner according to the different performances of the operations. The higher the score, the better the operator performance. The evaluation coefficient of the operation is updated according to the set formula, and the corresponding target operation is reselected based on the evaluation coefficient.

[0168] Step 324, using the updated target destruction operation, target repair operation and target disturbance operation, the candidate delivery double code body as the new solution of the delivery double code is subjected to iterative destruction, repair and disturbance operations.

[0169] Through the above steps 321 to 324, the iterative update of the candidate delivery dual code body based on the ALNS algorithm is achieved.

[0170] This embodiment also provides an intelligent linkage decision device for production and transportation, a production scheduling planning device, and a transportation planning decision device. The above devices are used to implement the above embodiments and preferred implementation modes, and those that have been described will not be repeated. As used below, the terms "module", "unit", "subunit", etc. can implement a combination of software and / or hardware for predetermined functions. Although the devices described in the following embodiments are preferably implemented in software, the implementation of hardware, or a combination of software and hardware, is also possible and conceivable.

[0171] Figure 4 is a structural block diagram of an intelligent linkage decision-making device for production and transportation according to an embodiment of the present application, such as Figure 4 As shown, the device includes an acquisition module 41, an update module 42 and a processing module 43, wherein:

[0172] The sending module 41 is used to send the updated expected values ​​of production variables and distribution variables to the production planning subsystem and the transportation planning subsystem respectively after completing the current collaborative optimization.

[0173] The judgment module 42 is coupled to the sending module 41, and is used to receive the production variable return value returned by the production planning subsystem in response to the production variable expected value and the distribution variable return value returned by the transportation planning subsystem in response to the distribution variable expected value, and judge whether the production variable return value and the distribution variable return value meet the preset production and transportation consistency constraint, wherein the production variable return value is used to represent the production scheduling parameters for batch production planning of multiple production orders, and the distribution variable return value is used to represent the distribution parameters for combined order transportation planning of multiple production orders to be distributed;

[0174] The processing module 43 is coupled to the judgment module 42, and is used to repeatedly execute the collaborative optimization using the preset collaborative optimization CO algorithm and dynamic tolerance to update the production variable expected value and the distribution variable expected value, and to judge whether the production variable return value responding to the updated production variable expected value and the distribution variable return value responding to the updated distribution variable expected value meet the production and operation consistency constraints, until the new response production variable return value and distribution variable return value meet the production and operation consistency constraints, and obtain the linkage decision result, wherein the CO algorithm uses a dynamic relaxation factor to update the dynamic tolerance, and the linkage decision result includes the production variable return value and the distribution variable return value that meet the production and operation consistency constraints, and is used to use the currently received production variable return value and the distribution variable return value as the linkage decision result when it is judged that the production variable return value and the distribution variable return value meet the preset production and operation consistency constraints.

[0175] In some embodiments, the processing module 43 further includes:

[0176] A first determination unit is used to determine the production sub-variable group and the distribution sub-variable group corresponding to the production variable return value and the distribution variable return value respectively after receiving the current corresponding production variable return value and the distribution variable return value, wherein the production sub-variable group includes the total production cost, the production offline time set and the production warehouse time set, and the distribution sub-variable group includes the total distribution cost, the distribution warehouse time set and the distribution delivery time set, the production offline time set is used to measure the total production cost, the distribution delivery time set is used to determine the total distribution cost, and the production warehouse time set cooperates with the distribution warehouse time to constrain the production, delivery and storage of the corresponding product;

[0177] A second determining unit is coupled to the first determining unit and is used to respectively determine a dynamic relaxation factor corresponding to each production sub-variable in the production sub-variable group and each distribution sub-variable in the distribution sub-variable group, and to respectively determine a dynamic tolerance corresponding to each production sub-variable and each distribution sub-variable according to the dynamic relaxation factor;

[0178] The first correction unit is coupled to the second determination unit, and is used to correct the production sub-variable expected values ​​corresponding to all production sub-variables according to each production sub-variable and the corresponding dynamic tolerance to obtain updated production variable expected values, and to correct the distribution sub-variable expected values ​​corresponding to all distribution sub-variables according to each distribution sub-variable and the corresponding dynamic tolerance to obtain updated distribution variable expected values.

[0179] In some embodiments, the determination module 42 further includes:

[0180] The third determination unit is used to determine the expected value of the total production cost, the expected value set of the production offline time, and the expected value set of the production delivery time corresponding to the expected value of the production variable, and determine the expected value of the total distribution cost, the expected value set of the distribution entry time, and the expected value set of the distribution delivery time corresponding to the expected value of the distribution variable;

[0181] The first calculation unit is coupled to the third determination unit and is used to calculate the deviation values ​​corresponding to the total production cost and the expected value of the total production cost, the production offline time set and the production offline time expected value set, the production warehouse time set and the production warehouse time expected value set, the total distribution cost and the expected value of the total distribution cost, the distribution warehouse time set and the distribution warehouse time expected value set, and the distribution shipment time set and the distribution shipment time expected value set, and after normalizing the calculated first deviation, second deviation, third deviation, fourth deviation, fifth deviation and sixth deviation values, all the normalized deviations are weighted to obtain a first total production and transportation consistency difference;

[0182] a first judgment unit, coupled to the first calculation unit, for calculating a second total production-operation consistency difference corresponding to the received production variable return value and the distribution variable return value according to the production variable expected value and the distribution variable expected value updated by the collaborative optimization before the current collaborative optimization, and the received production variable return value and the distribution variable return value, and determining whether the difference between the first total production-operation consistency difference and the second total production-operation consistency difference is less than the production-operation consistency constraint threshold;

[0183] The first processing unit is coupled to the first judgment unit, and is used to determine whether the corresponding production variable return value and distribution variable return value satisfy the preset production and operation consistency constraint when the first judgment unit determines that the difference between the first total production and operation consistency difference and the second total production and operation consistency difference is less than the production and operation consistency constraint threshold.

[0184] Figure 5 is a structural block diagram of a production scheduling planning device used in a production planning subsystem according to an embodiment of the present application, such as Figure 5 As shown, the production scheduling planning device includes:

[0185] The first coding module 51 is used to obtain the production order information of the current corresponding batch production, and perform initial population coding on the production order information to generate multiple first double-coded individuals, wherein the first double-coded individual includes multiple first double-coded sub-codes, and one first double-coded sub-code includes multiple double codes in the same production workshop, and each double code includes an order code and a batch code that constitute a coding pair, the order code is used to represent the order to which the corresponding product belongs, and the batch code is used to represent the batching of products corresponding to the order code.

[0186] The selection module 52 is coupled to the first coding module 51 and selects a set number of second dual-coded individuals from a plurality of first dual-coded individuals based on the individual fitness corresponding to the first dual-coded individuals and a preset selection strategy.

[0187] The iteration module 53 is coupled to the selection module 52, and is used to perform genetic evolution iteration operations on the order codes and batch codes corresponding to all the second dual-coded individuals until a target dual-coded individual is generated, and the target dual-coded individual is decoded to generate a production variable return value, wherein the genetic evolution operation includes a crossover operation, a mutation operation, and a disturbance operation, and the individual fitness is determined by solving a preset production objective function, and the production objective function is constructed by weighting the difference between the received expected value of the production variable and the production variable return value generated by decoding the first dual-coded individual.

[0188] In some embodiments, the selection strategy includes an ES-TS hybrid strategy, and the selection module 52 further includes:

[0189] A first selection unit is used to determine the individual fitness corresponding to each first dual-coded individual, and based on the individual fitness, select a set number of elite dual-coded individuals from multiple first dual-coded individuals using an elite retention ES strategy, wherein the elite dual-coded individuals include first dual-coded individuals whose individual fitness is less than a fitness threshold;

[0190] a second selection unit coupled to the first selection unit, configured to select a set number of candidate dual-coded individuals from the plurality of first dual-coded individuals based on individual fitness, using a standard roulette algorithm and a set selection probability, and select the best dual-coded individual from the selected candidate dual-coded individuals based on a tournament TS strategy;

[0191] The third selection unit is coupled to the second selection unit and is used to repeatedly select a set number of candidate dual-coded individuals from multiple first dual-coded individuals and select the best dual-coded individual from the selected candidate dual-coded individuals to obtain multiple best dual-coded individuals, and use all the selected elite dual-coded individuals and multiple best dual-coded individuals as second dual-coded individuals.

[0192] In some of the embodiments, the production scheduling and planning device is also used to obtain a preset job table, wherein the job table includes batch information, batch order information and production information as well as the correspondence information among the three, the batch order information includes at least one order code, and the production information includes the production batch of the batch-to-order and the delivery time of the batch order corresponding to the batch order information; according to all order codes and all batch codes corresponding to each first double-coded sub-sub-of the current double-coded individual to be decoded, the batch order corresponding to each first double-coded sub-sub-and the batch information corresponding to each batch-to-order are generated, and the production information corresponding to the batch order and the batch information corresponding to each first double-coded sub-sub-is queried in the job table, wherein the current double-coded individual includes one of the first double-coded individual and the target double-coded individual; based on the batch information corresponding to each first double-coded sub-sub-, the batch order information, the production batch and the delivery time corresponding to the batch order information, a simulated production is performed. Deduction, deduce the completion time and single-batch production cost corresponding to each batch order, respectively count the completion time and single-batch production cost corresponding to all batch orders corresponding to each current double-coded individual, and obtain the production offline time set and total production cost corresponding to each current double-coded individual; from the time interval determined by the completion time and delivery time corresponding to each batch order, select the time with the smallest difference from the target delivery time expectation value, and obtain the delivery time corresponding to each batch order, count the delivery time corresponding to all batch orders corresponding to each current double-coded individual, and obtain the production delivery time set corresponding to each current double-coded individual cost, where the target delivery time expectation value is an delivery time expectation value corresponding to the corresponding batch order in the delivery time expectation value set of the production variable expectation value, and the corresponding production variable return value includes the total production cost, the production offline time set and the production delivery time set corresponding to the current double-coded individual.

[0193] In some of the embodiments, the production scheduling and planning device is also used to obtain the corresponding total production cost, production offline time set and production delivery time set from the production variable return value corresponding to the first dual-coded individual; respectively calculate the deviation values ​​corresponding to the total production cost corresponding to the first dual-coded individual and the corresponding expected value of the total production cost, the production offline time set and the corresponding expected value set of the production offline time, and the production delivery time set and the expected value set of the production delivery time to obtain the seventh deviation value, the eighth deviation value and the ninth deviation value; after normalizing the seventh deviation value, the eighth deviation value and the ninth deviation value, weight the normalized deviation values ​​to obtain the individual fitness of the first dual-coded individual.

[0194] Figure 6 is a structural block diagram of a transportation planning decision device used in a transportation planning subsystem according to an embodiment of the present application, such as Figure 6 As shown, the transportation planning decision device includes:

[0195] The second encoding module 61 is used to obtain the inbound order information corresponding to the current delivery, and encode the inbound order information to generate an initial double encoding body, wherein the initial double encoding body includes an order encoding sequence having multiple delivery order encoding sub-codes and a delivery encoding sequence having multiple vehicle assembly encoding sub-codes, one delivery order encoding sub-code corresponds to one vehicle assembly encoding sub-code, and forms a delivery double encoding sub-code, the delivery order encoding sub-code is used to characterize the encoding of the corresponding inbound order, and the order of the delivery order encoding sub-code in the corresponding order encoding sequence is used to characterize the priority of the delivery order, the vehicle assembly encoding sub-code is used to characterize the type of delivery vehicle selected when the corresponding inbound order is delivered, and one vehicle assembly encoding sub-code also maps at least one delivery path, and the nodes of the delivery path include the initial position of the delivery vehicle and the storage position of the product corresponding to the inbound order;

[0196] The operation module 62 is coupled to the second encoding module 61, and is used to use the linear sorting to select the LRS strategy and the control line CL strategy, and respectively select the target destruction operation, the target repair operation and the target disturbance operation from the multiple destruction operations, the multiple repair operations and the multiple disturbance operations corresponding to the adaptive large neighborhood search ALNS algorithm, and use the target destruction operation, the target repair operation and the target disturbance operation in turn to destroy, repair and disturb all the delivery double encoders and the delivery order encoders and the vehicle assembly encoders corresponding to the delivery double encoders to generate alternative delivery double encoders, wherein the destruction operation includes one of the following: head and tail truncation, random removal, random block cutting, random excision, the repair operation includes one of the following: random block filling, random gap filling, head and tail gap filling, greedy gap filling, and the disturbance operation includes one of the following: transposition, block shifting, flipping, order change, greedy replacement.

[0197] The first decoding module 63 is coupled to the operation module, and is used to decode the alternative delivery double code body to obtain the alternative delivery information, and obtain the delivery variable value corresponding to the alternative delivery double code body from the alternative delivery information, and according to the deviation between the delivery variable value and the received expected value of the delivery variable, iteratively destroy, repair and disturb the alternative delivery double code body until the target delivery double code body is generated.

[0198] The second decoding module 64 is coupled to the first decoding unit 63, and is used to decode the target delivery double code body to obtain the target delivery information, and generate a delivery variable return value according to the target delivery information.

[0199] In some of the embodiments, the transportation planning decision device is also used to obtain a distribution parameter table, wherein the distribution parameter table includes distribution batch information, vehicle type information and distribution information as well as the correspondence information among the three, and the distribution information includes the distribution path information, distribution batch size and delivery deadline of the distribution batch corresponding to the distribution batch information; according to all the distribution order codes and vehicle group codes corresponding to the current distribution double code body to be decoded, the distribution batch information of all distribution batches corresponding to the current distribution double code body and the vehicle type information corresponding to each distribution batch are obtained, and in the distribution parameter table, according to the distribution batch information and vehicle type information corresponding to each distribution batch, the distribution information corresponding to the corresponding distribution batch is queried, wherein the current distribution double code body includes one of the alternative distribution double code body and the target distribution double code body; from the received distribution variable expected values, a distribution entry time expected value set is obtained, and the distribution entry time expected value set is used as the distribution entry time set corresponding to the current distribution double code body, wherein each distribution batch corresponds to the distribution entry time. a delivery entry time in the time set; based on all delivery batch information, vehicle type information and delivery path information, delivery batch and delivery deadline corresponding to each delivery batch corresponding to the current delivery double-coded body, delivery deduction is performed to deduce the delivery delivery time and single delivery cost corresponding to each delivery batch, and the delivery delivery time and single delivery cost corresponding to all delivery batches are counted to obtain the delivery delivery time set and total delivery cost corresponding to the current delivery double-coded body, wherein the delivery delivery time is selected from the time interval determined by the delivery entry time and delivery deadline corresponding to each delivery batch, and the time with the smallest difference from the target delivery time expectation value is selected, and the target delivery time expectation value includes the delivery time expectation value corresponding to the corresponding delivery batch in the delivery time expectation value set of the delivery variable expectation value; the total delivery cost, delivery entry time set and delivery delivery time set corresponding to the current delivery double-coded body are used as the delivery variable value corresponding to the corresponding delivery double code, wherein the delivery variable return value includes the delivery variable value corresponding to the target delivery double-coded body.

[0200] In some of the embodiments, the transportation planning decision device is also used to obtain at least the corresponding total delivery cost and delivery delivery time set from the delivery variable value corresponding to the current corresponding alternative delivery double code body, and obtain the expected value of the total delivery cost and the expected value set of the delivery delivery time from the expected value of the delivery variable; normalize the deviation value between the total delivery cost and the expected value of the total delivery cost and the deviation value between the delivery delivery time set and the expected value set of the delivery delivery time, respectively, and weight the normalized deviation value to obtain the delivery fitness of the alternative delivery double code body corresponding to the current one; determine the delivery fitness of the alternative delivery double code body corresponding to the current one, and select a new solution of the delivery double code body from the current corresponding alternative delivery double code body and the alternative delivery double code individuals before the current one according to the delivery fitness and the Metropolis acceptance criterion, and update the target destruction operation, target repair operation and target disturbance operation; use the updated target destruction operation, target repair operation and target disturbance operation to iterate the destruction, repair and disturbance operations on the alternative delivery double code body as the new solution of the delivery double code.

[0201] This embodiment further provides a service platform, including a memory and a processor, wherein a computer program is stored in the memory, and the processor is configured to run the computer program to execute the steps in any one of the above method embodiments.

[0202] Optionally, the service platform may further include a transmission device and an input / output device, wherein the transmission device is connected to the processor, and the input / output device is connected to the processor.

[0203] It should be noted that the specific examples in this embodiment can refer to the examples described in the above embodiments and optional implementation modes, and this embodiment will not be described in detail here.

[0204] In addition, in combination with the intelligent linkage decision-making, production scheduling planning method and transportation planning method for production and transportation in the above embodiments, the embodiments of the present application can provide a storage medium for implementation. The storage medium stores a computer program; when the computer program is executed by the processor, any one of the intelligent linkage decision-making, production scheduling planning method and transportation planning method for production and transportation in the above embodiments is implemented.

[0205] Those skilled in the art should understand that the technical features of the above embodiments may be arbitrarily combined. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0206] The above embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the invention patent. It should be pointed out that, for a person of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the patent of the present application shall be subject to the attached claims.

Claims

1. An intelligent linkage decision-making method for production and transportation, applied to the production and transportation collaboration system, characterized in that: include: After the collaborative optimization is completed, the updated expected values ​​of production variables and distribution variables are sent to the production planning subsystem and the transportation planning subsystem respectively; Receive the production variable return value returned by the production planning subsystem in response to the production variable expected value and the delivery variable return value returned by the delivery planning subsystem in response to the delivery variable expected value, and determine whether the production variable return value and the delivery variable return value meet the preset production and transportation consistency constraint, wherein the production variable return value is used to represent the production scheduling parameters for batch production planning of multiple production orders, and the delivery variable return value is used to represent the delivery parameters for combined delivery planning of multiple production orders to be delivered; In the case where it is determined that the production variable return value and the distribution variable return value do not satisfy the preset production-operation consistency constraint, the collaborative optimization using the preset collaborative optimization CO algorithm and dynamic tolerance is repeatedly performed to update the production variable expected value and the distribution variable expected value, and to determine whether the production variable return value in response to the updated production variable expected value and the distribution variable return value in response to the updated distribution variable expected value satisfy the production-operation consistency constraint, until the new response production variable return value and the distribution variable return value satisfy the production-operation consistency constraint, and a linkage decision result is obtained, wherein the CO algorithm uses a dynamic relaxation factor to update the dynamic tolerance, and the linkage decision result includes the production variable return value and the distribution variable return value that satisfy the production-operation consistency constraint, wherein, The collaborative optimization is performed using a preset collaborative optimization CO algorithm and a dynamic tolerance to update the expected value of the production variable and the expected value of the distribution variable, including: After receiving the currently corresponding production variable return value and distribution variable return value, determine the production sub-variable group and distribution sub-variable group corresponding to the production variable return value and the distribution variable return value respectively, wherein the production sub-variable group includes the total production cost, the production offline time set and the production outbound time set, and the distribution sub-variable group includes the total distribution cost, the distribution inbound time set and the distribution delivery time set, the production offline time set is used to measure the total production cost, the distribution delivery time set is used to determine the total distribution cost, and the production outbound time set is used to cooperate with the distribution inbound time to constrain the production, delivery and inbound of the corresponding product; Determine respectively the dynamic relaxation factors corresponding to each production sub-variable in the production sub-variable group and each delivery sub-variable in the delivery sub-variable group, and determine respectively the dynamic tolerance corresponding to each production sub-variable and each delivery sub-variable according to the dynamic relaxation factors; According to each of the production sub-variables and the corresponding dynamic tolerance, the production sub-variable expected values ​​corresponding to all the production sub-variables are corrected to obtain updated expected values ​​of the production variables; and according to each of the distribution sub-variables and the corresponding dynamic tolerance, the distribution sub-variable expected values ​​corresponding to all the distribution sub-variables are corrected to obtain updated expected values ​​of the distribution variables.

2. The method according to claim 1, characterized in that: When it is determined that the production variable return value and the distribution variable return value satisfy the preset production and transportation consistency constraint, the currently received production variable return value and the distribution variable return value are used as the linkage decision result.

3. The method according to claim 1, characterized in that Determining whether the corresponding production variable return value and the delivery variable return value meet the preset production and transportation consistency constraints includes: Determine the expected value of total production cost, the expected value set of production offline time, and the expected value set of production delivery time corresponding to the expected value of the production variable, and determine the expected value of total distribution cost, the expected value set of distribution entry time, and the expected value set of distribution delivery time corresponding to the expected value of the distribution variable; Calculate the deviation values ​​corresponding to the total production cost and the expected value of the total production cost, the production offline time set and the expected value set of the production offline time, the production warehouse time set and the expected value set of the production warehouse time, the total distribution cost and the expected value of the total distribution cost, the distribution warehouse time set and the expected value set of the distribution warehouse time, and the distribution shipment time set and the expected value set of the distribution shipment time, and after normalizing the calculated first deviation, second deviation, third deviation, fourth deviation, fifth deviation and sixth deviation values, weight all the normalized deviations to obtain the first total production and transportation consistency difference; According to the production variable expected value, the distribution variable expected value and the received production variable return value and the distribution variable return value updated by the collaborative optimization before the current one, calculate the second total production-operation consistency difference corresponding to the received production variable return value and the distribution variable return value, and determine whether the difference between the first total production-operation consistency difference and the second total production-operation consistency difference is less than the production-operation consistency constraint threshold; When it is determined that the difference between the first total production and operation consistency difference and the second total production and operation consistency difference is less than the production and operation consistency constraint threshold, it is determined that the corresponding production variable return value and the distribution variable return value meet the preset production and operation consistency constraint.

4. A production scheduling planning method, applied to a production planning subsystem, for generating the production variable return value in any one of claims 1 to 3, characterized in that: include: Acquire the production order information of the current corresponding batch production, and perform initial population coding on the production order information to generate multiple first double-coded individuals, wherein the first double-coded individual includes multiple first double-coded sub-units, one of the first double-coded sub-units includes multiple double codes in the same production workshop, each of the double codes includes an order code and a batch code constituting a code pair, the order code is used to characterize the order to which the corresponding product belongs, and the batch code is used to characterize the batching of the products corresponding to the order code; Based on the individual fitness corresponding to the first dual-coded individuals and a preset selection strategy, selecting a set number of second dual-coded individuals from a plurality of the first dual-coded individuals; Perform genetic evolution iteration operation on the order codes and batch codes corresponding to all the second dual-coded individuals until a target dual-coded individual is generated, and decode the target dual-coded individual to generate the production variable return value, wherein the genetic evolution operation includes a crossover operation, a mutation operation and a perturbation operation, and the individual fitness is determined by solving a preset production objective function, and the production objective function is constructed by weighting the difference between the received expected value of the production variable and the production variable return value generated by decoding the first dual-coded individual.

5. The method according to claim 4, characterized in that The selection strategy includes an ES-TS hybrid strategy, which selects a set number of second dual-coded individuals from a plurality of the first dual-coded individuals based on the individual fitness corresponding to the first dual-coded individuals and a preset selection strategy, including: Determine the individual fitness corresponding to each of the first dual-coded individuals, and based on the individual fitness, select a set number of elite dual-coded individuals from a plurality of the first dual-coded individuals using an elite retention ES strategy, wherein the elite dual-coded individuals include the first dual-coded individuals whose individual fitness is less than a fitness threshold; Based on the individual fitness, a set number of candidate dual-coded individuals are selected from the plurality of first dual-coded individuals using a standard roulette algorithm and a set selection probability, and the best dual-coded individual is selected from the selected candidate dual-coded individuals based on a tournament TS strategy; Repeat the steps of selecting a set number of candidate dual-coded individuals from the plurality of the first dual-coded individuals and selecting the best dual-coded individual from the selected candidate dual-coded individuals to obtain a plurality of best dual-coded individuals, and use all the selected elite dual-coded individuals and the plurality of the best dual-coded individuals as the second dual-coded individuals.

6. The method according to claim 5, characterized in that Generate corresponding production variable return values, including: Obtaining a preset operation table, wherein the operation table includes batch information, batch order information and production information as well as corresponding relationship information among the three, the batch order information includes at least one order code, and the production information includes a batch-to-order production batch and a delivery time of the batch order corresponding to the batch order information; According to all the order codes and all the batch codes corresponding to each of the first double-encoded sub-units of the current double-encoded individual to be decoded, a batch order corresponding to each of the first double-encoded sub-units and batch information corresponding to each of the batch order lists are generated, and the production information corresponding to the batch order and the batch information corresponding to each of the first double-encoded sub-units is searched in the operation table, wherein the current double-encoded individual includes one of the first double-encoded individual and the target double-encoded individual; Based on the batch information corresponding to each of the first dual-coded sub-units, the batch order information, the production batch corresponding to the batch order information, and the delivery time, a simulated production deduction is performed to deduce the completion time and single-batch production cost corresponding to each of the batch orders, and the completion time and single-batch production cost corresponding to all the batch orders corresponding to each of the current dual-coded individuals are respectively counted to obtain a set of production offline time and total production cost corresponding to each of the current dual-coded individuals; From the time interval determined by the completion time and the delivery time corresponding to each of the batch orders, the time with the smallest difference from the target delivery time expectation is selected to obtain the delivery time corresponding to each of the batch orders, and the delivery times corresponding to all the batch orders corresponding to each of the current double-coded individuals are counted to obtain the set of production delivery times corresponding to the cost of each of the current double-coded individuals, wherein the target delivery time expectation is an expected delivery time value corresponding to the corresponding batch order in the set of expected delivery time values ​​of the production variable expected value, and the corresponding production variable return value includes the total production cost, the production offline time set and the production delivery time set corresponding to the current double-coded individual.

7. The method according to claim 6, characterized in that Determining the individual fitness corresponding to each of the first dual-coded individuals includes: From the production variable return value corresponding to the first double-coded individual, obtain the corresponding total production cost, the production offline time set and the production delivery time set; Calculate the deviation values ​​corresponding to the total production cost and the corresponding expected value of the total production cost, the production offline time set and the corresponding expected value set of the production offline time, and the production delivery time set and the production delivery time expected value set respectively, to obtain a seventh deviation value, an eighth deviation value, and a ninth deviation value; After normalizing the seventh deviation value, the eighth deviation value, and the ninth deviation value, weighting is performed on the normalized deviation values ​​to obtain the individual fitness of the first dual-coded individual.

8. A method for transport planning decision-making, applied to a transport planning subsystem, for generating the delivery variable return value of any one of claims 1 to 3, characterized in that: include: Obtain the inbound order information corresponding to the current delivery, and encode the inbound order information to generate an initial double code body, wherein the initial double code body includes an order code sequence having multiple delivery order coders and a delivery code sequence having multiple vehicle group coders, one delivery order coder corresponds to one vehicle group coder, and forms a delivery double coder, the delivery order coder is used to characterize the code of the corresponding inbound order, and the order of the delivery order coder in the corresponding order code sequence is used to characterize the priority of the delivery order, the vehicle group coder is used to characterize the type of delivery vehicle selected when the corresponding inbound order is delivered, one vehicle group coder also maps at least one delivery path, and the nodes of the delivery path include the initial position of the delivery vehicle and the storage position of the product corresponding to the inbound order; The linear sorting selection LRS strategy and the control line CL strategy are used to select target destruction operations, target repair operations and target disturbance operations from a variety of destruction operations, a variety of repair operations and a variety of disturbance operations corresponding to the adaptive large neighborhood search ALNS algorithm, respectively, and the target destruction operations, the target repair operations and the target disturbance operations are used in turn to destroy, repair and disturb all the delivery double coders and the delivery order coders and the vehicle assembly coders corresponding to the delivery double coders to generate alternative delivery double coders, wherein the destruction operation includes one of the following: head and tail truncation, random removal, random block cutting, random excision, the repair operation includes one of the following: random block filling, random gap filling, head and tail gap filling, greedy gap filling, and the disturbance operation includes one of the following: transposition, block shifting, flipping, order change, greedy replacement; Decoding the candidate delivery double code body to obtain the candidate delivery information, and obtaining the delivery variable value corresponding to the candidate delivery double code body from the candidate delivery information, and iterating the destruction, repair and disturbance operations on the candidate delivery double code body according to the deviation between the delivery variable value and the received delivery variable expected value, until the target delivery double code body is generated; The target delivery double code body is decoded to obtain the target delivery information, and the delivery variable return value is generated according to the target delivery information.

9. The method according to claim 8, characterized in that Decode the corresponding delivery double code, including: Obtaining a delivery parameter table, wherein the delivery parameter table includes delivery batch information, vehicle type information, and delivery information as well as corresponding relationship information therebetween, and the delivery information includes delivery path information, delivery batch size, and delivery deadline of a delivery batch corresponding to the delivery batch information; According to all the delivery order codes and the vehicle group codes corresponding to the current delivery double code body to be decoded, the delivery batch information of all delivery batches corresponding to the current delivery double code body and the vehicle type information corresponding to each delivery batch are obtained, and in the delivery parameter table, according to the delivery batch information and the vehicle type information corresponding to each delivery batch, the delivery information corresponding to the corresponding delivery batch is queried, wherein the current delivery double code body includes one of the candidate delivery double code body and the target delivery double code body; Obtain a delivery entry time expected value set from the received delivery variable expected values, and use the delivery entry time expected value set as the delivery entry time set corresponding to the current delivery double code body, wherein each delivery batch corresponds to a delivery entry time in the delivery entry time set; Based on all the delivery batch information, the vehicle type information, the delivery path information, the delivery batch size and the delivery deadline corresponding to the current delivery dual-code body, delivery deduction is performed to deduce the delivery delivery time and the single delivery cost corresponding to each delivery batch, and the delivery delivery time and the single delivery cost corresponding to all the delivery batches are counted to obtain the delivery delivery time set and the total delivery cost corresponding to the current delivery dual-code body, wherein the delivery delivery time is the time with the smallest difference from the target delivery delivery time expected value selected from the time interval determined by the delivery warehousing time and the delivery deadline corresponding to each delivery batch, and the target delivery delivery time expected value includes the delivery delivery time expected value corresponding to the corresponding delivery batch in the delivery variable expected value set; The total distribution cost, the distribution warehousing time set and the distribution shipment time set corresponding to the current distribution double code body are used as the distribution variable values ​​corresponding to the corresponding distribution double code, wherein the distribution variable return value includes the distribution variable value corresponding to the target distribution double code body.

10. The method according to claim 9, characterized in that According to the deviation between the delivery variable value and the received delivery variable expected value, the candidate delivery dual encoding body is subjected to iterative destruction, repair and disturbance operations, including: From the distribution variable value corresponding to the currently corresponding candidate distribution dual encoding body, at least the corresponding distribution total cost and the distribution delivery time set are obtained, and the distribution total cost expected value and the distribution delivery time expected value set are obtained from the distribution variable expected value; Normalizing the deviation between the total delivery cost and the expected value of the total delivery cost and the deviation between the delivery delivery time set and the expected value set of the delivery delivery time, respectively, and weighting the normalized deviation values ​​to obtain the delivery fitness of the candidate delivery dual encoding body corresponding to the current one; Determine the delivery fitness of the candidate delivery double-encoded body corresponding to the current one, and select a new delivery double-encoded body solution from the candidate delivery double-encoded body corresponding to the current one and the candidate delivery double-encoded body before the current one according to the delivery fitness and Metropolis acceptance criteria, and update the target destruction operation, the target repair operation and the target disturbance operation; The updated target destruction operation, target repair operation and target disturbance operation are used to iterate the destruction, repair and disturbance operations on the candidate delivery double-encoding body as the new solution of the delivery double-encoding.

11. A service platform, characterized in that: It includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the intelligent linkage decision method for production and transportation as described in any one of claims 1 to 3, or implements the production scheduling planning method as described in any one of claims 4 to 7, or implements the transportation planning decision method as described in any one of claims 8 to 10.