Research On Management Optimizing Method For Polyester Staple Fibre Production Processes

Polyester staple fiber production is a typical processing production process. Recent years, with the global market situation changes, especially, the price changes in the upstream raw material oil and the downstream products, the higher requirement is presented for the production management of chemical fiber production plants. The production styles of the plants are changed from the past single production into the multiple products and small quantities. The higher demands are put forward for manage of the production process and control of production cost. These require the plants quickly adapt to the environment, enhance the management of production procedure and production cost for competition advantage.The optimization problem for Polyester Staple Fibre Production Processes is an important topic. There exist many problems worthwhile to be studied. The problem of this dissertation comes from the typical large-scale Petrochemical enterprises of China. In the dissertation some optimization problems for Polyester Staple Fibre Production Processes are proposed andstudied.The dissertation reviews the production situation of polyester staple fiber plants. The focuses are on the sticking point problem, such as spinpacks replacing cycle optimization and production process scheduling problem for polyester staple fiber. First it studies the spinpacks replacing cycle optimization by analyzing the factor correlation to the production cost. Secondly it studies the production scheduling of two kinds of production line, i.e., direct spin production scheduling and indirect spin production scheduling. Finally, a production planning and scheduling decision support system based on the researches above is designed and developed.The main works are as follows:(1) The spinpacks replacing cycle scheduling problem for polyester staple fiber Based on the production process of polyester staple fiber, considering the production technology and characteristics of polyester staple fiber, this paper presents a new production process optimization method to decrease the materials consumption and energy consumption. The problem is to determine when the spinpacks to be replaced so as to minimize the total cost over a given planning horizon. The total cost consisting of the operating cost and the changeover cost for the spinpacks accounts for an important fraction in a polyester staple fiber production plant. Two methods based on the optimization to used to solve the replacing cycle scheduling problem for the spinpacks in a typical polyester staple fiber plant. The economic order quantity (EOQ) method is for long planning horizon. The dynamic programming (DP) method is for short planning horizon. The model and algorithm are applied to SINOPEC Tianjin Company. The computational results verify the effectiveness of the proposed methods.(2) The scheduling problems consisting in direct spin of the polyester staple fiber plantsThe scheduling problems consisting in direct spin of the polyester staple fiber plants, which is using the same kinds of raw material or different kinds of raw material with close price, producing products with different specification and close price, production process scheduling problem is characterized by Traveling Salesman Problem(TSP). A mixed-integer programming (MILP) model for cyclic schedules in polyester staple fiber plants is formulated. The objective function of the problem is minimization of sum of costs. Considering the complexity of the problem and model, a taboo search (TS) algorithm and a hybrid intelligent optimization algorithm based on taboo search (TS) which embedded with variable depth search (VDS) are used to solve the problem in order to gain better results. The TS is good at the local search while VDS is capable to expend the searching space, and the advantages of these two methods are fully combined in our algorithm. The computation results validate the model and algorithm.(3) The scheduling problems consisting in indirect spin of the polyester staple fiber plantsWhich is using different kinds of raw material with different price, producing products with different specification and different price, production process scheduling problem is characterized by traveling salesman problem (TSP). A mixed-integer programming (MILP) model for cyclic schedules in polyester staple fiber plants is formulated. The objective function of the problem is maximization of the profit. The model introduces the raw material cost into the objective function and considers raw material cost; in contrast none of the similar models considers the influence of raw material cost. Considering the complexity of the problem and model, a TS algorithm and a hybrid intelligent optimization algorithm are used to solve the problem in order to gain better results. In order to improve the algorithm, three different neighborhoods are used alternately which increase the diversification of VDS. The effectiveness of the proposed model and methods are verified with the data collected from a real-world manufacturing facility cf SINOPEC Tianjin Company.(4) The integrated optimization problems for spinpacks cyclic replacing and the production scheduling of the polyester staple fiber plantsThe spinpacks cyclic replacing interferes with whole production scheduling of multi-product. The study is carried on integrated optimization for the decision of the spinpacks cyclic replacing and the production scheduling of multi-product. A MILP is presented. The objective function includes the product changeover cost, spinpacks cyclic replacing cost and etc. The restrains include time and spinpacks cyclic replacing and etc. The near-optimum solution is found by heuristic method. For the requirement of production for the polyester staple fiber plants, an integrated optimization decision support system for spinpacks cyclic replacing and the production scheduling is designed and developed based on the model and algorithm of the above researches and employed the production data of SINOPEC Tianjin Company.