Dynamic Scheduling Of Mold Design Tasks With Rework

In the mold production process,first the mold design is required,and then the parts are manufactured and assembled.Since the initial product requirements of the customer are often unclear,the design task needs to be submitted to the customer fo r confirmation.The result of the customer confirmation may be a one-time pass,or it may be necessary to modify the design plan several times.In other words,the completion time of the design task is highly uncertain,and this uncertainty will be spread to subsequent manufacturing tasks,thus causing a great impact on the production management of the entire mold.In addition,mold companies are usually several designers with different capabilities within a design team to complete multiple concurrent desig n tasks.At present,companies mainly rely on experience to make design task scheduling decisions.Usually,it is impossible to reasonably arrange limited designers to complete these urgent delivery design tasks in a timely manner,causing severe task dela ys.In order to solve this problem,this paper explores a dynamic scheduling method based on the Markov decision process theory for the design task of repair with a mold.The specific study is as follows:(1)Build a discrete event simulation model for the single-machine scheduling problem considering the random repair mold design task,and embed a variety of classic priority rules in the simulation model.Using different rules to simulate the examples,the total weighted tardiness cost of the task is the evaluation criterion,and the advantages and disadvantages of these rules under different random rework environments are compared and analyzed.(2)Under the assumption that the task can not preempt resources,based on discrete-time Markov decision process theory to establish a dynamic scheduling model considering the random repair and multi-mode mold design tasks,describes the optimal strategy solution method based on stochastic dynamic programming,And an approximate strategy to solve action sets using priority rules.(3)Under the premise of allowing rework tasks to seize resources,the two modes of task interruption-non-recoverable and interrupt-recoverable are considered,and the discrete-time Markov decision process theory is used to model and solve r espectively,and the tasks can be preempted.Dynamic scheduling problem.(4)Developed an information management system for dynamic scheduling of mold design tasks.By embedding the scheduling model and method constructed in this paper into the system,the optimal scheduling decision of the mold design task can be achieved,which can be connected with the enterprise ERP system as part of the enterprise information management system.In addition,this paper also designed a corresponding calculation experimen t to verify the proposed model and algorithm.Through the above research,not only the practical problems of the dynamic scheduling of mold design tasks are solved,but also the methods proposed by the thesis can play an auxiliary role for the enterprises in task scheduling.