Research On Cooperative Manufacturing And Transportation Scheduling Based On Approximate Dynamic Programming Algorithm

In recent years,with the continuous innovation of many manufacturing enterprises,a multi-manufacturer collaborative manufacturing mode combining intelligence and digitalization has gradually evolved,which makes full use of the advantages of manufacturers through collaborative way to achieve the purpose of reducing production time,improving product quality and reducing product time to market.However,in the actual production scenarios of enterprises,how to deploy and utilize the superior resources of many cooperative enterprises to complete tasks with low cost and high efficiency becomes the difficulty of collaborative manufacturing.The research content of this paper is mainly divided into two parts,considering the production problem and the combination of production and transportation in the collaborative manufacturing process.Collaborative production problems in many manufacturing enterprises for many process product collaborative manufacturing process between the production plans and manufacturer of choice,on the basis of markov decision process,the order after the completion of the same process of processing and transportation process is divided into a stage of the manufacturers processing action and node transport action,In order to minimize production cost,waiting cost,penalty cost of late delivery and inventory cost of early delivery,a model was established based on the status of waiting and completion of orders at the manufacturer and the processing and arrival time of nodes.According to the problems only in order to complete the process that the characteristics of its delivery,put forward a kind of heuristic rules give orders in production time without waiting for the upper bound and lower bound to approximate delivery status and cost,at the same time,the method of approximate linear function approximation order delivery status and cost,finally combining with two kinds of approximate method the approximation of dynamic programming algorithm for solving problems.At the end of the problem,a static cooperative production model was established,and cplex was used to solve the model accurately.Collaborative production transportation problem based on the background of collaborative production problems,to join the order in the merger of the transport behavior between manufacturers,to consider the order transport costs between the manufacturers,the use of combined transport can save the traffic cost,build to minimize the cost of production,waiting for the cost,transportation cost,delay penalty cost and inventory cost of delivery in advance in order to establish the model optimization goal.Finally,the method of co-production problem is used to solve the problem.At the end of the problem,a static cooperative production model was established,and cplex was used to solve the model accurately.In the numerical example part of the paper,the production related parameters are simulated in line with the actual situation,and the two problems are solved by the approximate dynamic programming under the two approximate methods respectively,and the results are compared with those solved by the static model.Finally,it can be found that in the comparison process,the cost of the dynamic programming scheme under the linear approximation method is better than the result obtained by the heuristic rule approximation.The error between the two and the exact solution of the cooperative manufacturing problem is about 5%,and the error between the two and the exact solution of the cooperative manufacturing transportation problem is about7% and 10%,respectively.