Research On The Self-Designed Order Plan In The Steel Industry

In order to reduce open-order slabs quickly,self-designed order planning problems are investigated in this thesis.The research can significantly improve the open-order slab utilization,reduce the production and inventory cost,as well as improve the enterprise benefit.Self-designed order planning problems include self-designed order designing problem and self-designed order matching problem.The self-designed orders designing problem is clustering the redundant open-order slabs into several batches,each corresponding to a self-designed order.The self-designed order matching problem is to allocate the open-order slabs to self-designed order.Two integer programming models are established,and two improved differential evolution algorithms are designed to solve the two problems.The specific work is as follows:1)The self-designed order designing problem in multiple parallel production lines is investigated.The task of the problem is clustering open-order slabs into batches,each corresponding to a self-designed order,and to allocate the formed orders self-designed to hot rolling production line,with the constraints of load balancing and order capacity.Compared with the previous research,this thesis considers the self-designed order clustering and line distribution simultaneously.Considering actual production process,an integer programming model is built.The optimal objective is to minimize the self-designed order number,the difference cost among slabs in the same order,production load deviation,and the waiting time in stock of slabs,as well as maximize the number of used open-order slabs.In addition,an improved differential evolution algorithm based on HDE is designed to solve the problem quickly.Finally,the proposed algorithm is compared with CPLEX by computational experiments,the comparison result the proposed model and algorithm is valid for solving the self-designed order designing problem.2)The self-designed order matching problem is investigated.The task of the problem is to allocating open-order slabs to self-designed order with the constraints of matching process rules and order capacity.Considering actual production process,an integer programming model is built.The optimal objective is to minimize the difference cost between demands and dispensed amount of orders,open cost.In addition,an improved differential evolution algorithm is designed to solve the problem quickly.Based on the structure characteristics,a customized integer encoding,decoding is designed.Two kinds of heuristic algorithm and random strategy are combined together to get better quality initial population.At the same time,based on the variation operation of the mixed strategy,the definition of the special symbol in the discrete difference is proposed,which ensures the diversity and convergence rate of the solution.The infeasible solutions are repaired generated in the process of differential mutation.Finally,the proposed algorithm is compared with CPLEX by computational experiments,the comparison result the proposed model and algorithm is valid for solving the self-designed order matching problem.3)Self-designed order making model and self-designed order matching model are embedded into the desktop applications.This thesis designs self-designed order plan decision support system to realize the making and matching of self-designed order.The system improves enterprise performance and reduces production cost.