Search

CN-116484997-B - Method and device for collaborative optimization of production scheduling and distribution under large-scale customization

CN116484997BCN 116484997 BCN116484997 BCN 116484997BCN-116484997-B

Abstract

The invention discloses a method and a device for collaborative optimization of production scheduling and distribution under large-scale customization, and relates to the technical field of collaborative optimization of production scheduling and distribution. The method comprises the steps of obtaining production data to be optimized, inputting the production data to be optimized into a constructed collaborative optimization model of production scheduling and distribution, and solving the collaborative optimization model of production scheduling and distribution by adopting a multi-objective evolutionary algorithm IMOEA/COD based on collaboration and decomposition to obtain an optimized production scheduling and distribution scheme. The invention can solve the problem of how to cooperatively produce and distribute two stages in a large-scale customization environment, and a reasonable production scheduling and distribution scheme is formulated so as to realize the reduction of the maximum finishing moment, distribution cost and advance/delay cost. The invention considers the mutual influence of two stages of production and distribution, can fully utilize the existing resources, and effectively solves the problem of collaborative optimization of production scheduling and distribution under large-scale customization.

Inventors

  • WU XIULI
  • CUI JIANJIE

Assignees

  • 北京科技大学

Dates

Publication Date
20260512
Application Date
20221230

Claims (7)

  1. 1. A method for collaborative optimization of production scheduling and distribution under large-scale customization, the method comprising: S1, obtaining production data to be optimized, wherein the production data comprises order demand information, workshop processing information and delivery vehicle information; s2, inputting the production data to be optimized into a constructed collaborative optimization model for production scheduling and distribution; s3, solving the production scheduling and distribution collaborative optimization model by adopting a multi-objective evolutionary algorithm IMOEA/COD based on collaborative and decomposition to obtain an optimized production scheduling and distribution scheme; the construction process of the production scheduling and distribution collaborative optimization model in the S2 comprises the following steps: S21, setting parameters of a production scheduling and distribution collaborative optimization model, wherein the parameters comprise population number PN, iteration number Iter, crossover probability P c , variation probability P m and neighborhood updating threshold T; S22, setting an objective function and constraint conditions to obtain a production scheduling and distribution collaborative optimization model, wherein the objective function is the minimization of the maximum completion time, the minimization of distribution cost and the minimization of early or late cost; The constraint conditions comprise that any operation can be only allocated to one machine at a certain moment, the same machine can start to carry out next operation after finishing the previous operation, the finishing moment of any working procedure of parts depends on starting time and processing time, the finishing moment of order assembly depends on starting assembly time and assembly time, the next working procedure can be started after the same part is required to finish the previous working procedure, the assembly of custom parts required by order assembly can be started after all finishing, each part starts to process from 0 moment, orders loaded in each vehicle cannot exceed capacity limit, the starting moment of each vehicle cannot be earlier than the production finishing moment of the order in the vehicle, the order delivering moment is equal to the delivering moment of the previous order in the vehicle or the vehicle starting moment, the running time is added, each vehicle starts from a workshop, the vehicle returns to the workshop after the task is executed, each order is responsible for delivery, the decision variable value is 0 or 1, and the constraint variable value is non-negative; And in the step S3, a multi-objective evolutionary algorithm IMOEA/COD based on synergy and decomposition is adopted to solve the production scheduling and distribution synergy optimization model to obtain an optimized production scheduling and distribution scheme, which comprises the following steps: S31, initializing a population; S32, assigning a weight vector of the sub-problem to each individual in the initial population; S33, selecting two parent individuals in the population to perform cross operation to generate an individual x cross ; S34, performing mutation operation on the individual x cross to generate a child individual x child ; S35, performing collaborative decoding operation on the child individuals x child to obtain a production scheduling and distribution scheme; S36, updating the neighborhood of the individual x child ; S37, judging whether the preset iteration times Iter is met, if yes, outputting an external archive, namely, a non-dominant solution set obtained after iteration, to obtain an optimized production scheduling and distribution scheme, and if not, turning to execute S33.
  2. 2. The method of claim 1, wherein the order demand information in S1 includes a time window, a demand, a volume, a geographic location of the customer, and a cost per unit time of advance or retard delivery for each order; The workshop processing information comprises workshop geographic position, order processing information and standard component processing information, wherein the order processing information comprises order assembly time, fixed components required by order assembly and process information of each customized component; the delivery vehicle information includes a capacity limit, a travel speed, a fixed use cost, and a delivery cost per unit distance of the delivery vehicle.
  3. 3. The method of claim 1, wherein initializing the population in S31 comprises: s311, adopting a genetic algorithm to respectively aim at the delivery cost and the early or late cost, and sequentially solving the delivery stages to obtain delivery schemes P1 and P2; S312, generating a random number r between 0 and 1, if r is less than or equal to 0.25, executing S313, if r is less than or equal to 0.25 and less than or equal to 0.5, selecting P1 as a distribution scheme P c of a current individual, executing S314, if r is less than or equal to 0.5 and less than or equal to 0.75, selecting P2 as a distribution scheme P c of the current individual, executing S314, and if r is less than or equal to 0.75 and less than or equal to 1, executing S315; S313, randomly generating individual codes, storing the individual codes into an initial population P init , and executing S316, wherein the individual codes comprise an individual production part code I 1 and an individual distribution part code I 2 ; S314, taking the code of the P c as an individual distribution part code I 2 , randomly sequencing custom parts required for assembling orders on the premise of meeting the prior production of orders in vehicles with early departure according to the departure time and order allocation information of the vehicles in the P c , and storing the custom parts into an individual production part code I 1 ; S315, randomly generating an individual production part code I 1 , generating an individual distribution part code I 2 on the premise of meeting the requirement of priority distribution of orders corresponding to the custom parts which are produced preferentially and the limitation of vehicle capacity according to the order of the custom parts in the I 1 ; s316, judging whether the population number of the initial population P init is smaller than the preset population number PN, if yes, turning to execute S312, otherwise, outputting the initial population P init .
  4. 4. The method of claim 1, wherein the two parent individuals in the selected population in S33 perform a crossover operation to generate individual x cross , comprising: S331, generating a random number r2 between 0 and 1, randomly selecting two parent individuals x 1 、x 2 from the population if r2 is less than or equal to 0.5, otherwise, randomly selecting two parent individuals x 1 、x 2 from the neighborhood of the individuals x; S332, generating a random number r3 between 0 and 1, executing crossover operation on the parent individual x 1 、x 2 only when r3 is less than or equal to preset crossover probability P c , and randomly selecting one individual from crossed individuals as an individual x cross .
  5. 5. The method of claim 1, wherein the performing a mutation operation on the individual x cross in S34 generates a child individual x child , comprising: Generating a random number r4 between 0 and 1, and executing mutation operation on the individual x cross only when r4 is less than or equal to a preset mutation probability P m to generate a child individual x child .
  6. 6. The method of claim 1, wherein the performing a co-decoding operation on the child individual x child in S35 results in a production scheduling and distribution scheme, comprising: S351, acquiring a delivery part code I 2 of the child individual x child , solving by adopting a nearest insertion method with the delivery cost as a target, and adjusting the departure time of a vehicle to generate a delivery scheme D; S352, taking the vehicle departure time of the distribution scheme D as the optimal production completion time of the in-vehicle order, and acquiring a production part code I 1 of the child individual x child , wherein i=0; S353, if I is smaller than the total number of genes of I 1 , executing S354, otherwise, generating a production scheduling and distribution scheme S pd , and executing S358; S354, acquiring an ith gene g i of the I 1 , if g i is a standard component, executing S355, otherwise, executing S356; S355, if g i does not start scheduling, the earliest starting time of the current process is 0, otherwise, the earliest starting time of the current process is the finishing time of the preceding process, forward scheduling is performed by a gap extrusion method, i=i+1 is performed, and S353 is executed; s356, if the order of g i is not assembled, obtaining the optimal production completion time of the order of g i , performing reverse scheduling by using a gap extrusion method, and executing S357, otherwise, executing S357; S357, if the scheduling is not started by g i , the optimal finishing time of the current working procedure is the order starting assembly time of g i , otherwise, the optimal finishing time of the current working procedure is the working starting time of the subsequent working procedure, the reverse scheduling is carried out by a gap extrusion method, i=i+1 is carried out, and S353 is executed; S358, if the starting time of the production part in S pd is not negative, executing S359, otherwise, sequentially adjusting the production part and the distribution part in S pd to enable S pd to be feasible, and executing S359; S359, outputting a production scheduling and distribution scheme S pd .
  7. 7. A production scheduling and distribution co-optimizing device under large scale customization, the device comprising: The system comprises an acquisition module, a distribution module and a control module, wherein the acquisition module is used for acquiring production data to be optimized, and the production data comprises order demand information, workshop processing information and distribution vehicle information; The input module is used for inputting the production data to be optimized into the constructed collaborative optimization model for production scheduling and distribution; The output module is used for solving the production scheduling and distribution collaborative optimization model by adopting a multi-objective evolutionary algorithm IMOEA/COD based on collaborative and decomposition to obtain an optimized production scheduling and distribution scheme; the construction process of the production scheduling and distribution collaborative optimization model comprises the following steps: S21, setting parameters of a production scheduling and distribution collaborative optimization model, wherein the parameters comprise population number PN, iteration number Iter, crossover probability P c , variation probability P m and neighborhood updating threshold T; S22, setting an objective function and constraint conditions to obtain a production scheduling and distribution collaborative optimization model, wherein the objective function is the minimization of the maximum completion time, the minimization of distribution cost and the minimization of early or late cost; The constraint conditions comprise that any operation can be only allocated to one machine at a certain moment, the same machine can start to carry out next operation after finishing the previous operation, the finishing moment of any working procedure of parts depends on starting time and processing time, the finishing moment of order assembly depends on starting assembly time and assembly time, the next working procedure can be started after the same part is required to finish the previous working procedure, the assembly of custom parts required by order assembly can be started after all finishing, each part starts to process from 0 moment, orders loaded in each vehicle cannot exceed capacity limit, the starting moment of each vehicle cannot be earlier than the production finishing moment of the order in the vehicle, the order delivering moment is equal to the delivering moment of the previous order in the vehicle or the vehicle starting moment, the running time is added, each vehicle starts from a workshop, the vehicle returns to the workshop after the task is executed, each order is responsible for delivery, the decision variable value is 0 or 1, and the constraint variable value is non-negative; The method for solving the production scheduling and distribution collaborative optimization model by adopting a multi-objective evolutionary algorithm IMOEA/COD based on collaboration and decomposition to obtain an optimized production scheduling and distribution scheme comprises the following steps: S31, initializing a population; S32, assigning a weight vector of the sub-problem to each individual in the initial population; S33, selecting two parent individuals in the population to perform cross operation to generate an individual x cross ; S34, performing mutation operation on the individual x cross to generate a child individual x child ; S35, performing collaborative decoding operation on the child individuals x child to obtain a production scheduling and distribution scheme; S36, updating the neighborhood of the individual x child ; S37, judging whether the preset iteration times Iter is met, if yes, outputting an external archive, namely, a non-dominant solution set obtained after iteration, to obtain an optimized production scheduling and distribution scheme, and if not, turning to execute S33.

Description

Method and device for collaborative optimization of production scheduling and distribution under large-scale customization Technical Field The invention relates to the technical field of collaborative optimization of production scheduling and distribution, in particular to a method and a device for collaborative optimization of production scheduling and distribution under large-scale customization. Background With the rapid development of economy and technology, the variety of commodities is increasingly abundant, the demands of customers tend to be diversified and personalized, the traditional large-scale production mode is difficult to adapt to the market development, and the manufacturing industry faces new challenges. The large-scale customization can realize the high-speed and low-cost production of products with customized production characteristics in large-scale production, provides a low-cost diversified solution for the survival of manufacturing enterprises in strong market competition, and has a plurality of manufacturing enterprises to implement the production mode. The mass-customized production process mainly consists of two stages of part processing and product assembly. The part processing stage is mixed line production of standard parts and fixed parts, the standard parts are required by all products, and a mass production mode is adopted. The customized piece embodies the personalized characteristics of the product and needs to be produced according to the order. Product assembly is driven by customer demand. Meanwhile, production and distribution are two important value activities for manufacturing enterprises, and the operation efficiency and cost of the two important value activities affect the overall benefit of the supply chain. In the conventional "production-distribution" operation mode, production and distribution are independently operated by two departments, and thus, there is a lack of timely information sharing and operation coordination, and optimization of overall operation is difficult to achieve. In particular, in a large-scale custom model, inefficient decisions on production and distribution will directly impact customer satisfaction and enterprise operating benefits. Therefore, the production and distribution are necessary to be combined, a reasonable and efficient collaborative optimization method is formulated, and the maximization of the overall benefits of enterprises and clients is realized. Disclosure of Invention The invention provides the method for reasonably formulating the collaborative optimization scheme for the two stages of production and distribution under large-scale customization so as to improve the overall benefits of clients and enterprises. In order to solve the technical problems, the invention provides the following technical scheme: in one aspect, the invention provides a method for collaborative optimization of production scheduling and distribution under large-scale customization, which is implemented by electronic equipment and comprises the following steps: s1, obtaining production data to be optimized, wherein the production data comprise order demand information, workshop processing information and delivery vehicle information. S2, inputting production data to be optimized into the constructed collaborative optimization model for production scheduling and distribution. And S3, solving a production scheduling and distribution collaborative optimization model by adopting a multi-objective evolutionary algorithm IMOEA/COD based on collaborative and decomposition, so as to obtain an optimized production scheduling and distribution scheme. Optionally, the order demand information in S1 includes time window, demand, volume, customer geographic location, and cost per unit time of lead or lag delivery for each order. The workshop processing information comprises workshop geographic position, order processing information and standard component processing information, wherein the order processing information comprises order assembly time, fixed components required by order assembly and process information of each of the fixed components, and the standard component processing information comprises process information of each of the standard components, batch times and sub-batch numbers. The delivery vehicle information includes a capacity limit of the delivery vehicle, a traveling speed, a fixed use cost, and a delivery cost per unit distance. Optionally, the process of constructing the collaborative optimization model for production scheduling and distribution in S2 includes: S21, setting parameters of a production scheduling and distribution collaborative optimization model, wherein the parameters comprise population number PN, iteration number Iter, crossover probability P c, variation probability P m and neighborhood updating threshold T. S22, setting an objective function and constraint conditions to obtain a production scheduling and distributio