Optimization Scheduling Method Of Hot Metal Between Iron-Making Plants And Steel-Making Plants And Its Application

In modern large iron and steel enterprises, hot metal logistics process (HMLP) between iron-making plants and steel-making plants consists of multiple blast furnaces, hot metal pretreatment stations and reladling stations. The hot metal pretreatment station includes multiple pre-processing equipments (PR), desulphurization and dephosphorization equipments (SP), and post-processing equipments (PO). Hot metal is firstly poured into a pot carried on torpedo car (TPC). The hot metal in the same TPC is called a 'pot' or job. Then the pot is transported by locomotive to a pretreatment station for PR, SP and PO. Finally, the pot is driven by locomotive through a rail-track network to reladling station and the hot metal in pot is dumped into ladles.The hot metal scheduling between iron-making plants and steelmaking plants firstly aims at ensuring the supply and demand balance of hot metal logistics. Secondly, on the basis of the blast furnace tapping plan and steel tapping plan, hot metal scheduling needs to be done the correspondence between the pots and pre-pots according to the requirements of cast plan for the time, weight, composition of hot metal, which is in the form of the iron-steel correspondence plan. Thirdly, on the basis of the iron-steel correspondence plan, the hot metal scheduling aims at determining the processing machines of pots in every process and the starting process time and the end process time of pots in each machine, which is in the form of the hot metal scheduling plan. Usually, the hot metal scheduling plan needs to be adjusted in practical production due to the fluctuation of the processing time of pots on the machine, the second receiving iron of pots in blast furnace, the time delay of transfer facilities and the fluctuation of transportation time. If the original scheduling is still performed, the processing time conflict of two adjacent pots will occur on the same machine. Therefore, it needs to the rescheduling of hot metal. Thus it can be seen, the hot metal scheduling between iron-making plants and steelmaking plants includes the hot metal logistics balance (HMLB), the iron-steel correspondence scheduling (ISCS), the hot metal static scheduling (HMSS) and the hot metal rescheduling (HMRS).The maximum capacity of the blast furnaces and the converters, the demand weight for iron of the steel plants, the weight of hot metal being processed, the maximum and minimum safety stock weight of hot metal being processed and the transport costs from all blast furnaces to all steel plants are known in practice. The optimization objectives are to maximize the production capacity of every blast furnace and to minimize the costs of transporting hot metal from every blast furnace to every steel plant. The problem of HMLB aims at determining the weight of hot metal transported from every blast furnace to every steel plant within each sub-cycle of the planning cycle, which is in the form of the plan table of HMLB.On the basis of the blast furnace tapping plan and steel tapping plan, the ISCS aims at determining the correspondence relationship between the pots and the pre-pots transformed by the cast plans. That is ensuring the pot whose the end time of receiving iron is earlier have preferential correspondence relationship with the pre-pot whose the target time of reladling is earlier, which is in the form of the iron-steel correspondence plan.In the HMLP, the hot metal pretreatment station and the reladling station that the pots be processed, the processing time of all the pots on all the machines and the transportation time between different machines are known. The performance indexes of the HMSS are that the hot metal in each pot is dumped into ladles punctually at reladling station and the processing time conflict of two adjacent pots cannot occur on the same machine. On the basis of the iron-steel correspondence plan, the HMSS aims at determining the processing machines of the pots in every process and the starting process time and the end process time of pots in each machine, which is in the form of the hot metal scheduling plan. The HMRS is similar to the HMSS except that its optimization objective and constraints refer to the process status parameters of pots such as the performed pot, the performing pot and the unperformed pot.The hot metal scheduling between iron-making plants and steelmaking plants has a very important role in the production of iron and steel enterprises. If the result of hot metal scheduling is not satisfactory, the production safety of iron-making will be affected or the production time of steel-making will be delayed. In the HMSS problem and the HMRS problem, the constraints of the processing time of different pots cannot overlap on the same machine and the processing sequence of the pots in different processes have many kinds of possibility description, and they can be expressed as multiple constraint equations. So the HMSS problem and the HMRS problem cannot be solved using existing optimization method. In the existing research literature, the hot metal pretreatment station is seen as a whole, which ignores that the hot metal pretreatment station includes a lot of process equipments, such as PR, SP and PO. In addition, the fact that many similar machines exist in many processes has also been neglected. Therefore, the existing research methods are difficult to be applied effectively to the hot metal scheduling. As a result, manual scheduling is taken to tackle the problem, which often results in the long redundant waiting time between different machines or affecting the utilization of the bottleneck equipments, such as PR and PO.This dissertation takes a large iron and steel enterprises as the research background, whose hot metal logistics process consists of four blast furnaces, two hot metal pretreatment stations (which including one PR, two desulphurization equipments at the No.1 steelmaking plant and one PR, three SP, two PO at the No.2 steelmaking plant), two reladling stations (which including one work position at the No.1 steelmaking plant and four work positions at the No.2 steelmaking plant). The optimization scheduling method for hot metal between iron-Making plants and steel-making plants and its application are studied in this dissertation, where the major contributions are listed as follows:1. First of all, on the basis of describing the HMLP between iron-Making plants and steel-making plants, the overall structure of hot metal scheduling is presented, which includes the HMLB, the ISCS, the HMSS and the HMRS. The constraint conditions, the optimization objectives and decision variables of above four problems are depicted. While, the reason that the hot metal scheduling cannot be solved using existing optimization method is analyzed. And on the basis of describing the manual scheduling process of the production site, the shortcoming of the manual scheduling is analyzed.2. The overall scheduling strategy of the hot metal scheduling is proposed. And the optimization scheduling method for hot metal based on rules, heuristic algorithm and linear programming is suggested, which mainly consists of the following four parts.?The hot metal logistics balance method:Through the analysis of the technological requirements of the HMLP, this dissertation develops a mathematical model for the HMLB problem, which consists of some objective functions and some constrain equations. The objective functions are to maximize the production capacity of every blast furnace and to minimize the costs of transporting hot metal from every blast furnace to every steel plant. The constraint equations are that the limit of the maximum capacity of the blast furnaces and the converters, and the weight of the hot metal being processed between the minimum and the maximum safety stock weight. The decision variable is that the weight of hot metal transported from every blast furnace to every steel plant within each sub-cycle of the planning cycle. And the model for the HMLB can be solved by using linear programming method.?The iron-steel correspondence scheduling method:On the basis of analyzing the factors (including the weight, component, time, temperature etc) that be considered in manual scheduling, the performance indexes and constrains of the ISCS problem are described. Considering the difficulties in accurately modeling of the ISCS problem, seven rules are abstracted from the experience of scheduling experts and technological requirements of the production site. And we use the heuristic algorithm based on rules to solve the problem.?The hot metal static scheduling method:For the HMSS problem, the objective functions are to minimize the total waiting time in every process of the pots and the total earliness and tardiness time of the pots at reladling station, and to maximize the production capacity of the bottleneck machines, such as PR and PO. The constraint equations are that the processing time of different pots cannot overlap on the same machine, and the processing sequence of the pots in different processes and the limit of the maximum capacity of the machines of PR and PO. The decision variables are that the allocation machine to every process of each pot and the start time of every process of each pot. Since the problem cannot be solved by existing optimization methods, this dissertation presented a scheduling strategy as follows. Firstly, the pots are distinguished between the pots of hot metal of special type (HMST) and the pots of hot metal of normal type (HMNT) by using rules. Secondly, we determine the allocation machines in every process for the pots of HMST by using heuristic algorithm. Thirdly, the start time in each process of the pots of HMST is figured out by using linear programming method. Finally, we schedule the pots of HMNT by using the heuristic algorithm based on the rule of first come first service (FCFS).?The hot metal rescheduling method:For the HMRS problem of a large delay caused by the delay of the start time of a pot, the objective functions are to minimize the total waiting time of the unperformed pots and the total earliness and tardiness time of the unperformed pots at reladling station, and to maximize the production capacity of the bottleneck machines, such as PR and PO. The constraint equations are that the processing time of two adjacent pots can not conflict on the same machine, and the processing sequence of the unperformed pots in different processes and the limit of the maximum capacity of the machines of PR and PO. The decision variables are that the allocation machine to the unperformed process of each pot and the start time of the unperformed process of each pot. Because the HMRS problem can not be solved by the existing optimization methods, this dissertation proposes a rescheduling strategy including five steps. Step one is to distinguish the unperformed pots and the performed pots based on rules. Step two is to partition the unperformed pots into the unperformed pots of HMST and the unperformed pots of HMNT by using rules. Step three is to determine the allocation machines in every process for the unperformed pots of HMST by using the heuristic algorithm. Step four is to determine the start time in each process of the unperformed pots of HMST by using linear programming method. The last step is to reschedule the unperformed pots of HMNT using the heuristic algorithm based on the rule of FCFS.3. The scheduling methods above-mentioned have been coded to a scheduling software. And we have applied them to the aforementioned iron and steel enterprise successfully. The practical running results show that the various plans all can be made by the presented methods within 10 seconds, which is much shorter than the time needed by the manual making mode. And the requirements of production site for real-time and rapidity has been satisfied. Compared with the results of the manual scheduling, the scheduling results of the average daily 150 pots reduce the waiting time in every process of the pots and decrease the earliness and tardiness time of the pots at reladling station. While, the utilization rate of the bottleneck equipments (i.e. PR and PO) is raised. And the production efficiency has been improved.