Study On Method For Production Planning And Scheduling In Reheating Furnace&Hot Rolling Processes

In this dissertation, the characters of the steel production are analyzed and the sate ofproduction planning and scheduling in reheating furnace&hot rolling is also reviewed. Theresearch of this dissertation is supported by National Natural Science Foundation of China underGrant No.71172219and Research Foundation of Education Department of Anhui Province inChina under Grant No. SK2012B578. Taking a national production line of reheating furnace&hotrolling as study object and aiming at increasing the productivity of hot rolling and reducing energyconsumption of reheating furnace, three special planning and scheduling problems of reheatingfurnace&hot rolling are solved by considering the actual production process in Iron&Steelenterprises. The three problems are hot-rolling batch plan, production scheduling of heatingfurnace&hot rolling and rescheduling of heating furnace&hot rolling. This dissertation hasmainly carried on the following research.(1) For the hot-rolling batch plan problem. The target of the problem is thatminimize the penalty cost, which is caused by temperature jump and specification changesbetween batches. Therefore, rolling plan problem is analyzed and a VRP model of rolling batchplan is proposed in this dissertation, and the rolling temperature constraint of the slabs should beto focus on the model. For this model, a new algorithm of fish seeking particle swarm is designedbased on particle swarm algorithm and artificial fish swarm algorithm. The steps are as follows.To begin with, the initial population is created by particle swarm optimization algorithm andevolved by the iterative formula. Then the fitness value is put into artificial fish swarm algorithmforforaging, assemblage, following and random operation.Finally, the reservations of the elitepopulation are compared, obtains the optimization results.(2) For the production scheduling of heating furnace&hot rolling problem. In orderto guarantee the smooth implementation of hot rolling batch planning, a new productionscheduling model is proposed for reheating furnace and hot rolling, and the target of the model isto realize continuous production, cost reduction, and other optimal objectives. In this process,production process demands, customer demand, and productivity of heating furnace and hot millare taken into account. In order to solve the multi-objective optimization problem, the Paretosorting strategy of non dominated sorting genetic algorithm (NSGA-II) is studied. The nondominated sorting genetic algorithm is mainly improved in two aspects. On the one hand is theindividual coding problem, the chromosome is used by the two sections and multilayer realcoding form. The first section real on the first layer represents the number of heating furnace, andthe second section real on the first layer represents the number of hot rolling mill. The first sectionreal on the second layer represents the beginning time of billets for heating process, and the second section real on the second layer represents the beginning time of billets for rolling process.On the other hand is mutation operation. Two pairs of mutation individuals are randomly selectedfrom the first section real and the second section real on the first layer according to the mutationprobability. The positions of the two pairs of individuals are exchanged. Then the two mutationindividuals are respectively selected from the two sections real of the second layer, and theresults of each mutation individual follows the uniform distribution on the interval so that thebegging time for processing to fit reality after mutation.(3) For the production rescheduling of heating furnace&hot rolling problem. When theproduction scheduling plan still can not meet the demand of actual production, it is necessary toadjust scheduling scheme based on the original production scheduling. In order to solvethe conflict problem in operation, a rescheduling mathematical model for reheating furnace andhot rolling is established, and the target of the model is to minimize the total penalty cost of theheating furnace fault, rolling mill fault and waiting time between processing equipment. In orderto make the simulation results closer to the production practice, the non-dominated sortinggenetic algorithm based on simulation optimization. Using the optimizer to solve the problem onupper strata, applying the simulator to evaluate optimal solution on lower strata, and theevaluation result will come back to the optimizer to guide the optimization. To reduce influence ofrandom factors, the chromosome should be simulated repeatedly and independently, and thenaverage value will be returned to the optimizer. When the heating furnace or hotrolling machine fails, using local rescheduling strategy can realize the rescheduling for theproduction of heating furnace and hot rolling.(4) In order to make the proposed model and algorithm to get the promotion and application,they can be integrated into the scheduling system.Developers, who need to understand what thesystem needs to do in order to develop it.The unified modeling language is used to establish thecase model, the static model and the dynamic mode for the system of heating furnace and hotrolling in this dissertation. It proves that the models make certain guidance for the development ofEMS. Therefore, the management level is improved in Iron and Steel enterprises, because it canhelp the managers to make a decision as a tool.All the formulated mathematical models for production planning and scheduling and thecorresponding algorithms are tested by using practical data. The numerical analysis has shown thatthe proposed model and algorithm are feasible and effective.