JP2006113684A - Process plan creation apparatus, operation controller, method for creating process plan, operation control method, computer program and recording medium - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To optimize at high speeds an operation plan or operation control for one process that matches operation plans under the operational restrictions of the other process, relating to manufacturing processes for processing a plurality of products on a plurality of different process routes. <P>SOLUTION: A process plan creation apparatus is provided with a Petri net discrete event simulator 10 for estimating the current state of physical distribution that varies every time a tapping event occurs; a logical physical distribution model construction device 11 for constructing a model of physical distribution within a predetermined delimited range of future prospects; a linear physical distribution model construction device 12; a logical physical distribution model retaining device 17; a linear physical distribution model retaining device 15; an evaluation function setting device 14; and an optimization calculation device 13 for making optimization calculations by making evaluations using evaluation functions. Optimization calculations are made using linear programming or constraint logic programming, whereby, even if the number of processes to pass or restrictions on physical distribution vary, solutions can be provided at high speed. <P>COPYRIGHT: (C)2006,JPO&NCIPI

Description

  The present invention relates to a processing plan creation device, an operation control device, a processing plan creation method, an operation control method, a computer program, and a recording medium, and more particularly to a technique suitable for use in a manufacturing process for processing a plurality of products through different plurality of process paths.

  Conventionally, in manufacturing processes in many industries including steel, a plurality of products are processed by a plurality of different processes. For example, in a process from a converter to continuous casting performed at an ironworks, molten steel discharged from a plurality of operating furnaces is transported to a plurality of continuous casting facilities (hereinafter referred to as CC), and continuous casting is performed in each. Processing is performed.

  When making such a production plan in CC, it is necessary to take into account the operational restrictions of CC and the constraints caused by the steelmaking logistics process. That is, in a steelmaking factory, the steel out of a converter is performed in units of one molten steel pan (referred to as charge), and in CC, casting is performed in units of casts of the same type of charge (for example, 8) as a unit.

  When knitting a production plan at CC, that is, casting, as much charge as possible within the range that the equipment can withstand from the viewpoint of productivity improvement while minimizing the quality degradation of the seam by continuously casting different steel types. Is required to be continuously cast.

  In addition, due to the nature of continuous casting, molten steel must always be present in the CC. Specifically, in order to pour out the molten steel, it is necessary to carry the molten steel from the converter to a container called a tundish that has a buffer role provided at the CC inlet so that the molten steel is not interrupted at all times. is there.

  Further, the casting speed in the CC can be changed within a certain range, but the range of the processing speed that can be changed is narrow due to the quality problem of the product. Furthermore, it is necessary to take into account that the passage route and the processing time in each process differ for each charge.

  If steel is produced from one converter to one CC under such various restrictions, each charge may be simply carried to the CC in the order of the given cast composition. However, under the operating conditions where there are multiple converters and CCs in operation as in the steelmaking factory as described above, in order to improve the overall production efficiency, which charge is in what order and which CC. It is necessary to make a steel production plan in the converter that matches the CC operation plan, whether to produce steel.

  In the case of a more complicated manufacturing process, there is a degree of freedom in selecting a processing method and a processing apparatus in each process, and it may be possible that the processing time, the number of passing processes, and distribution restrictions to be considered change depending on the selection. For example, in the process from the converter to the continuous casting performed at the same ironworks, there are two types of treatment called normal blowing and ORP blowing in each converter, and this treatment method is designated. There are charges that are not and those that are not. The above normal blowing is a conventional blowing method in which decarburization is performed in a converter.

  On the other hand, ORP blow smelting first performs dephosphorization treatment in the converter, and after discharging, the molten steel is once discharged from the converter into a pan on the carriage (called tapping water) and poured again into the converter for decarburization treatment. It is blowing and aims to reduce costs by consolidating the dephosphorization treatment that has been performed in separate facilities in the converter. However, since this ORP blowing has a longer processing time in the converter than in the normal blowing, if the ORP blowing is performed more than necessary, the number of steel cups from the converter may decrease and productivity may decrease.

  Moreover, the furnace which performs these decarburization processes and dephosphorization processes has the charge beforehand decided by operation, and the thing which is not so. When the blowing method and the furnace to produce steel are specified, it is necessary to make a production plan by strictly adhering to the specification. Otherwise, it is necessary to determine them according to some evaluation index.

  In general, there is a degree of freedom in selecting the equipment to be used when manufacturing a product, and multiple products can be routed through different multi-process routes that change the processing time, the number of processes to be passed, and logistics restrictions depending on the selected equipment. In the manufacturing process to be processed, it is desired to create a processing plan in the process or to automate operation control. Conventionally, simulators and the like by various methods have been proposed as techniques for this automation.

JP 2000-172745 A

  However, as disclosed in the “distribution plan creation device” of the above-mentioned Patent Document 1, a method for creating a schedule based on a method that guarantees optimality, such as linear programming and mathematical programming. In this case, (1) when the scale for creating a processing plan is increased, the number of combinations becomes enormous and it becomes difficult to solve in a practical time. In addition, (2) errors due to constraints and conditions that cannot be described by mathematical formulas occur, so it has not been guaranteed whether the obtained processing plan can be executed.

  For this reason, as in the above operation example, in the operation consisting of the upper process (converter) to the lower process (CC), there are many restrictions on the lower process, and in the selection of the processing method and equipment to be used in a certain process There is a degree of freedom, and a method suitable for executing a processing plan or operation control in the upper process using the degree of freedom effectively while keeping the restrictions has not been proposed. In addition, in the method of making a plan based on a simulation that has been often used in the past, there is a problem that an executable schedule can be created, but its optimality is not considered. Furthermore, if the optimization is performed by a conventional method that repeats the simulation until a satisfactory result is obtained, there is a problem that the planning time becomes long.

  In view of the above-mentioned problems, the present invention has a degree of freedom in selecting equipment to be used when manufacturing a product, so that the processing time, the number of steps to be passed, and physical distribution constraints change depending on the selected processing method and equipment. In a manufacturing process that processes multiple products through different multiple process paths, an optimal and feasible processing plan is considered while simultaneously considering the items to be planned, such as the selection of processing methods and processing equipment, the order of processing, and the processing time. The purpose is to enable high-speed planning.

  The processing plan creation device of the present invention is a processing plan creation device for creating a processing plan in a processing process for manufacturing a plurality of products with different plurality of process paths, and represents the processing flow and processing constraints of the processing process as a discrete event model. Processing of the above processing process for a discrete event system simulator that detects a processing state and processing constraints when an event occurs, and a target period (plan creation period) preset from the planning start date and time of the above processing plan Of the logical logistics model holding device that holds the logical logistics model that combines the line formats obtained from the states and processing constraints with logical operators, and each linear logistics model included in the logical logistics model, a predetermined one is determined. A linear logistics model holding device that selects and holds a linear logistics model that satisfies the specified conditions, and an evaluation for evaluating the linear logistics model An evaluation function setting device that sets a number, an optimization calculation device that performs an optimization calculation process using the linear logistics model and the evaluation function to calculate a processing instruction for the discrete event simulator, A processing instruction is calculated for a period (instruction calculation period) preset from the present time by the optimization calculation device, and is given to the discrete event simulator to execute simulation, and only for a preset period (plan decision period). A series of simulation results obtained by repeating the process of finalizing the processing plan, setting the date and time immediately after the determined period as a new planning start date and planning the processing plan, as the processing plan in the processing process It is characterized by deciding.

  The operation control apparatus of the present invention is an operation control apparatus for creating an operation instruction in a processing process for manufacturing a plurality of products through different plurality of process paths, and represents a processing flow and processing constraints of the processing process in a discrete event model. Processing of the above processing process for a discrete event system simulator that detects a processing state and processing constraints when an event occurs, and a target period (plan creation period) preset from the planning start date and time of the above processing plan Of the logical logistics model holding device that holds the logical logistics model that combines the line formats obtained from the states and processing constraints with logical operators, and each linear logistics model included in the logical logistics model, a predetermined one is determined. A linear logistics model holding device that selects and holds a linear logistics model that satisfies the specified conditions and an evaluation function for evaluating the linear logistics model are installed. An optimization function setting device, an optimization calculation device that performs an optimization calculation process using the linear logistics model and the evaluation function to calculate a processing instruction for the discrete event simulator, and the optimization A processing instruction is calculated for a period (instruction calculation period) set in advance from the current time by a calculation device, is given to the discrete event simulator, and a simulation is executed, and a processing plan is calculated for a preset period (plan decision period). Determine a series of simulation results obtained by repeating the process of establishing a process plan by setting the date and time immediately after the determined period as a new planning start date and time, as operation instructions in the process It is characterized by.

  The processing plan creation method of the present invention is a processing plan creation method for creating a processing plan in a processing process for manufacturing a plurality of products with different plurality of process paths, wherein the processing flow and processing constraints of the processing process are represented by a discrete event model. Processing of the above processing process for a discrete event system simulator that detects a processing state and processing constraints when an event occurs, and a target period (plan creation period) preset from the planning start date and time of the above processing plan Of the logical logistics model holding device that holds the logical logistics model that combines the line formats obtained from the states and processing constraints with logical operators, and each linear logistics model included in the logical logistics model, a predetermined one is determined. A linear logistics model holding device that selects and holds a linear logistics model that satisfies the specified conditions, and an evaluation for evaluating the linear logistics model Using an evaluation function setting device that sets a number, an optimization calculation device that performs an optimization calculation process using the linear logistics model and the evaluation function, and calculates a processing instruction for the discrete event system simulator, A processing instruction is calculated for a period (instruction calculation period) preset from the present time by the optimization calculation device, and is given to the discrete event simulator to execute simulation, and only for a preset period (plan decision period). A series of simulation results obtained by repeating the process of finalizing the processing plan, setting the date and time immediately after the determined period as a new planning start date and planning the processing plan, as the processing plan in the processing process It is characterized by deciding.

  The operation control method of the present invention is an operation control method for creating an operation instruction in a processing process for manufacturing a plurality of products through different plurality of process paths, and represents a processing flow and processing constraints of the processing process by a discrete event model. Processing of the above processing process for a discrete event system simulator that detects a processing state and processing constraints when an event occurs, and a target period (plan creation period) preset from the planning start date and time of the above processing plan Of the logical logistics model holding device that holds the logical logistics model that combines the line formats obtained from the states and processing constraints with logical operators, and each linear logistics model included in the logical logistics model, a predetermined one is determined. A linear logistics model holding device that selects and holds a linear logistics model that satisfies the specified conditions and an evaluation function for evaluating the linear logistics model are installed. The optimization function using the evaluation function setting device, the linear logistics model, and the optimization calculation device that performs optimization calculation processing using the evaluation function and calculates processing instructions for the discrete event simulator A processing instruction is calculated for a period (instruction calculation period) set in advance from the current time by a calculation device, is given to the discrete event simulator, and a simulation is executed, and a processing plan is calculated for a preset period (plan decision period). Determine a series of simulation results obtained by repeating the process of establishing a process plan by setting the date and time immediately after the determined period as a new planning start date and time, as operation instructions in the process It is characterized by.

The computer program of the present invention is a program for causing a computer to execute a processing plan creation method for creating a processing plan in a processing process for manufacturing a plurality of products with different plurality of process paths. Represented by an event model, the discrete event system simulator that detects the processing state and processing constraints when an event occurs, and the target period (plan creation period) preset from the planning start date and time of the above processing plan Of the logical logistics model holding device that holds the logical logistics model that combines the line formats obtained from the processing state and processing constraints of the processing process with logical operators, and among the linear logistics models included in the logical logistics model , Linear logistics model holding device that selects and holds a linear logistics model that satisfies a predetermined condition , An evaluation function setting device for setting an evaluation function for evaluating the linear logistics model, an optimization calculation process using the linear logistics model and the evaluation function, and processing instructions for the discrete event system simulator Using the optimization calculation device to calculate, a processing instruction is calculated for a period (instruction calculation period) set in advance from the present time by the optimization calculation device, and is given to the discrete event simulator to execute a simulation, A series obtained by repeating the process of determining a processing plan for a preset period (plan determination period), setting a date immediately after the determined period as a new planning start date, and planning a processing plan Characterized in that a computer executes a process for determining a simulation result of the above as a processing plan in the above processing process. To have.
Another feature of the computer program of the present invention is a program for causing a computer to execute an operation control method for creating an operation instruction in a processing process for manufacturing a plurality of products through different plurality of process paths. The processing flow and processing constraints of a processing process are represented by a discrete event model, a discrete event system simulator that detects a processing state and processing constraints when an event occurs, and a target period set in advance from the planning start date and time of the processing plan ( For the plan creation period), a logical distribution model holding device for holding a logical distribution model in which a line form obtained from the processing state and processing constraints of the processing process is combined with a logical operator, and the logical distribution described above Of the linear logistics models included in the model, select and maintain a linear logistics model that satisfies a predetermined condition A linear logistics model holding device, an evaluation function setting device for setting an evaluation function for evaluating the linear logistics model, the linear logistics model, and the discrete event by performing optimization calculation processing using the evaluation function A processing instruction is calculated for a period (instruction calculation period) preset from the present time by the optimization calculation apparatus using the optimization calculation apparatus that calculates a processing instruction for the system simulator, and is given to the discrete event simulator The simulation is executed, the processing plan is determined for a preset period (plan determination period), and the process of planning the process plan by setting the date and time immediately after the determined period as a new planning start date and time is repeated. A process for determining a series of simulation results obtained as a result of operation as an operation instruction in the above processing process It is characterized in that to execute.

  The recording medium of the present invention is characterized by recording the computer program described above.

According to the present invention, there is a degree of freedom in selecting a processing method and equipment to be used, and the processing time, the number of steps to be passed, and distribution restrictions vary depending on the selected processing method and equipment. When a new event occurs in a manufacturing process that processes multiple products, a logical expression model is newly constructed as the mathematical expression model, and the logical expression model that cannot be executed by analyzing the constructed logical expression model. Optimized after detecting and deleting and constructing an executable logic model, optimizing the executable processing plan in the processing process of manufacturing multiple products with different multiple process paths at high speed can do.
In addition, according to another feature of the present invention, an operation plan or operation control of a process that matches the operation plan of each process can be optimized at high speed under an operation restriction in a given process.

(First embodiment)
Hereinafter, an embodiment of the present invention will be described with reference to the drawings.
In the “processing plan creation device in the manufacturing process” (hereinafter, simply referred to as “processing plan creation device”) of the present embodiment, in the operation process from the converter to the CC as described in the conventional example, the given cast The optimization problem of the converter operation plan that matches the CC operation plan is dealt with under the organization result (casting order) and the steelmaking logistics constraints. However, this is merely an example, and the processing plan creation apparatus of the present embodiment has a degree of freedom in selecting a processing method and equipment to be used in a certain process, and processing time and passage depending on the selected processing method and equipment. In manufacturing processes where the number of processes to be performed and logistics restrictions change, it is possible to apply when creating an operation plan in each process while keeping many restrictions imposed on each process. It is valid.

  In the operation plan here, the first goal is to maximize the throughput (production amount per hour) of the steelmaking factory by making the casting speed in CC as fast as possible. The second purpose is to make the blowing method in the converter as ORP as possible (maximizing the ORP rate). The third object is to minimize the molten steel residence time in the intermediate process.

  Here, the second purpose is that the converter ORP process takes longer than the normal blowing as described above, so the productivity of the converter decreases as the number of charges for performing the converter ORP increases. Therefore, since the throughput of the steelmaking factory is lowered, the second purpose is contrary to the first and third purposes in this sense, and it is necessary to balance both.

  In addition, it takes time for the molten steel (charge) produced from the converter to be carried to CC by a crane or a carriage. At that time, depending on the type of the product, a secondary refining process may be performed in the middle of the process from the converter to the CC to adjust the components. In addition, the processing time of each charge in the CC may be different for each charge.

  On the other hand, as described above, the next charge needs to be poured before the contents of the tundish at the entrance of the CC are completely interrupted. Therefore, when preparing an operation plan for a converter, considering the processing time for each charge and the necessity of continuation of each charge and minimization of residence time, etc., It is necessary to accurately determine not only the steel order but also the time of steel production.

  FIG. 1 is a block diagram showing a schematic configuration of a processing plan creation apparatus according to this embodiment, and FIG. 2 is a diagram for explaining an outline of processing performed by the processing plan creation apparatus according to this embodiment. is there. FIG. 3 is a diagram showing the positioning of the processing plan creation device according to the present embodiment in the production planning system.

First, the positioning of the processing plan creation apparatus according to the present embodiment will be described with reference to FIG.
As shown in FIG. 3, when creating a daily operation plan, first, the cast knitting / casting order creation unit 31 arranges and casts between casts based on a preset weekly schedule 30. Determine the order in which each charge is cast. In addition, if there is a designation for the blowing method in the converter, specify it. Here, at least which charge is processed by which CC is determined.

  The processing plan creation unit 32 of the present embodiment has a cast knitting result (casting order or blowing method in a converter if specified) created by the cast knitting / casting order creation unit 31 and various steelmaking logistics restrictions. The operation plan in the converter, that is, the order of steel output and the time of steel output in the converter are obtained from the information of the casting order given by the cast knitting / casting order creation unit 31.

  In this processing plan creation unit 32, as will be described in detail below, a discrete event simulator using a Petri net that graphically models a physical distribution structure (equipment arrangement in the factory and its connection relationship, equipment capacity, charge passage route, etc.) The converter is a completely new combination of LP (linear programming), which is often used as a solution for static programming problems, and CLP (constraint logic programming), which is known as a solution for combinatorial optimization problems. Simultaneous optimization of the blowing method in the converter, the processing furnace, the steeling sequence and the time of steeling.

  The operation plan in the converter (information on the blowing method, processing furnace, steeling sequence and steeling time in the converter) obtained by the processing plan creation unit 32 is given to the Gantt chart display unit 33, for example, Gantt Displayed in the form of a chart. The various evaluation units 34 evaluate the obtained operation plan from various viewpoints (for example, residual transition), and if necessary, correct the cast arrangement and the casting order of each charge as necessary. Then, the operation plan creation unit 32 creates an operation plan again.

  Next, an outline of processing performed by the processing plan creation unit 32 will be described with reference to FIG. In the example of FIG. 2, the case where the steel is discharged from two converters of No. 1 and No. 2 furnaces to three CCs of No. 1 CC, No. 2 CC and No. 3 CC is taken as an example.

  Further, in the example of FIG. 2, the cast knitting / casting order creation unit 31 of FIG. 3 applies the charge indicated by the letters “A, B, C, D, E, F” from the No. 1 furnace to the No. 1 CC. In addition to the steel output in order, the charge indicated by the numbers “1, 2, 3, 4, 5, 6” in No. 2 CC is output in this order, and from No. 1 or No. 2 furnace to No. 3 CC “O , P, Q, R, S, T ”is designated to be charged in this order. Charges “A, B, C, D, E, F, 1, 2, 3, 4, 5, 6” are designated as normal blowing in the converter, but charge “O , P, Q, R, S, T ”does not specify the blowing method in the converter.

  Under such a cast arrangement, the processing plan creation unit 32 waits for a charge on the turret that is a future physical distribution state at each judgment time of the steeling time simulation (every time a steeling event occurs). And the processing method of the converter that optimizes the predetermined evaluation function set for maximizing the total t / H and minimizing the casting dwell time after estimating the deceleration rate of the charge processing speed and the furnace used The order of steeling and the time of steeling are determined simultaneously. At this time, the estimated range (instruction calculation period) of the future physical distribution state is, for example, 5 charges scheduled for the next steel production of each CC.

  That is, for example, as shown in FIG. 2 (a), it is assumed that a charge 1 steeling event occurs at time t in the simulation. This charge 1 is delivered from the No. 1 furnace and then carried by a crane or a carriage, and after time t1 (during this time t1, the processing time for secondary refining and the shortest waiting time on the No. 2 CC turret Arrive at No. 2 CC. And the process of continuous casting is performed in this No. 2 CC over time t2. When such a steeling event of charge 1 occurs, first, in step S1, 5 charges are set for each CC as a prediction range (instruction calculation period) of the future physical distribution state. Here, charge “A, B, C, D, E” for No. 1 CC, charge “2, 3, 4, 5, 6” for No. 2 CC, “O, P, Q, R” for No. 3 CC , S ”is set as the prediction range.

  Next, in step S2, each charge “A, B, C, D, E”, “2, 3, 4, 5, 6”, “O, P, Q, R, S” within the set prediction range Formulate a logical logistics model based on the logistics constraints. Here, the logical distribution model is a combination of various distribution constraints and mathematical formulas with logical operators (logical sum, logical product, negation), and will be described in detail later.

  In this way, when a steeling event occurs at a certain time t, when a logistics model is constructed by setting a prediction range for 5 charges for each CC from there, in the subsequent step S3, the construction formula and logical expression Using the logistics model and a predetermined evaluation function set in advance, 15 charges within the forecast range (“A, B, C, D, E”, “2, 3, 4, 5, 6”, “O , P, Q, R, S ”), the optimization time of the steel production time, blowing method, and furnace to be used is calculated.

  Here, by this optimization calculation, in step S4, charge A → charge 2 → charge O → charge B → charge 3 → charge P → charge C → charge 4 → charge Q → charge D → charge 5 → charge R → charge For Charges O, P, Q, R, and S where the order and time of E → Charge 6 → Charge S and the blowing method are not specified, Charges O, P, and S are ORP blowing, Charges Q and R are Suppose that the result that the normal blowing is the best is obtained.

  Therefore, in the next step S5, the steelmaking times of charges A, 2, and O are output to the Petri net simulator as one plan charge period for each CC, for example, among the steeling times within the above instruction calculation period. Give as steel instructions. In response to this, in step S6, the simulator proceeds with the simulation until the time O + steeling time t + Δt, and determines the simulation result as a processing plan. This state is shown in FIG. Since a steeling event has occurred again in this way, in step S7, 5 charges are set for each CC from the time t + Δt of the steeling event occurrence time of charge O as the prediction range (instruction calculation period) of the future physical condition. To do. For this newly set prediction range, the construction of the physical distribution model and the optimization calculation are performed as in the previous case.

Next, a schematic configuration of the processing plan creation unit 32 that performs processing as shown in FIG. 2 will be described with reference to FIG.
In FIG. 1, reference numeral 10 denotes a discrete event simulator based on Petri nets, which includes a graphical physical distribution structure model based on Petri nets and a rule description that cannot be graphically expressed. Here, as an example of the rule, there is “early steelmaking earliest time” indicating the limit of the time at which the charge can be steeled earliest, “latest steeling time” indicating the time limit of the time at which the charge can be steeled latest. .

  As described above, the casting speed of each charge in the CC can be changed only within a narrow range, and the waiting time on the turret has upper and lower limits. “Latest steel output latest time” and “Latest steel output latest time” are set for each charge. In addition, in order to continuously output steel from the same converter, it is necessary to secure an interval at least for the processing time in the converter, and even when steel is output from different converters, molten steel is charged into the converter. It is necessary to secure the necessary time interval due to restrictions such as crane capacity. For this reason, the TAP restriction is imposed on the condition that, for example, 45 minutes (in the case of normal blowing) for the same furnace and 20 minutes (in the case of normal blowing) for another furnace must be kept apart. .

  Reference numeral 11 denotes a logical logistics model construction apparatus, which is simulated by the discrete event simulator 10 and information on the logistics constraints set in the discrete event simulator 10 (early steelmaking earliest, latest time, TAP constraints, etc.). Combined with the current physical distribution status given as a result (information such as which charge is the one that started casting in the past and the earliest time of completion of casting), the future prediction range for 5 charges is set for each CC Then build a logistics model within that range. This logistics model is formulated by a mathematical model as described below. Here, formulation of the physical distribution model will be described.

FIG. 4 is a diagram showing an outline of formulation of a physical distribution model.
First, in the first step 41, the target manufacturing process 40 is classified into elements called tasks representing processing steps and elements called resources representing processing facilities. In the case of a steel mill, charging process, dephosphorization process, dephosphorization process, dephosphorization process, hot water process, recharge process, decarburization process, decarburization process, decarburization process for each charge The processing, the steelmaking process, the secondary refining process, the waiting process on the turret, and the casting process at CC are the tasks, respectively, and the two furnaces, the secondary refining process, the turret, CC, and the scrap crane are resources. When the charge is normal blowing, the processing time of the charging process, the dephosphorization process, the dephosphorization sedation process, the dephosphorization waste process, and the hot water process is treated as “0”.

  Next, in a second step 42, the prior relationship between the classified tasks and the resources necessary for executing the task are defined. Similarly, in the case of a steel factory, each task must be executed in the order of dephosphorization, decarburization, secondary refining, waiting, and casting. It must be carried out according to a defined casting sequence. In addition, the charging process requires a scrap crane, the dephosphorization process and decarburization process, if specified, the specified furnace, and if not specified, any furnace is required. Processing equipment is required, the waiting process requires a turret, and the casting process requires CC.

  After such classification, in step 43, the relationship between the task and the resource is converted into a logical expression, and a logical distribution model is constructed. This logical expression is obtained by setting the start and end times of each task as variables and connecting linear constraint expressions between the variables by logical operators such as logical sum, logical product, and logical negation. This logical logistics model includes information on various combinations, such as the charge blowing method for which the blowing method is not specified, the charge furnace number for which the converter is not specified, and the steeling sequence between CCs, and the target process. The distribution model includes all the information on the distribution constraints such as the processing order and the TAP interval, and a linear distribution model can be generated from the logical distribution model for each combination.

  However, the number of possible combinations increases exponentially as the number of target charges increases, and the number of generated mathematical formulas also increases. Therefore, in this embodiment, a constraint logic programming engine (hereinafter referred to as a CLP (Constraint Logic Programming Method) engine) is used to violate constraints from among the enormous number of combinations included in the above-described logical logistics model. A combination that is not feasible in actual operation is deleted, and a linear logistics model is generated only for the feasible combination. This way of thinking makes it possible to greatly reduce the scale of the problem and to significantly reduce the time required for solving.

  Returning to FIG. 1 again, reference numeral 13 denotes an optimization calculation apparatus. With this optimization calculation device 13, a logical distribution model 11 a is built by the logical distribution model construction device 11 and held by the logical distribution model holding device 17. Further, each linear logistics model 12 a generated by the linear logistics model construction device 12 is held by the linear logistics model holding device 15. And by evaluating with the evaluation function 14a as will be described later, which is set by the evaluation function setting device 14, the steelmaking sequence and blowing method represented by each linear logistics model 12a, under the furnace number selection, The optimal process start and end times for each process from the converter to the CC are calculated using a linear optimization technique such as LP. This calculation is performed every time a charge steeling event occurs in the simulation as described above (every plan decision period).

  The optimum steelmaking sequence and the steeling time calculated by the optimization calculation device 13 are sent to the discrete event simulator 10. Based on the result of the optimization calculation given from the optimization calculation device 13, the discrete event simulator 10 determines the time of the next steeling event (optimization of the charge that is to be output next by the optimization calculation). Advance the simulation time until (steeling time). Then, the simulation is performed again with the time advanced in this way, and the current physical distribution state such as the casting end time is calculated and supplied to the logical physical distribution model construction apparatus 11 and the linear physical distribution model construction apparatus 12. Hereinafter, in the same manner, the calculation of the discrete event simulator 10, the logical logistics model construction device 11, the linear logistics model construction device 12, and the optimization computing device 13 is repeatedly performed for each steeling event (every plan decision period). Create an entire schedule.

Here, the formulation of the above-described logical logistics model will be described in more detail. The above-described task has a process start time and a process end time as variables, and tasks having a predecessor relationship can express a constraint on the predecessor relationship by the following expression. This example shows the constraint that task B must be executed after task A, variable S is the task start time represented as a subscript, and variable F is the task end time represented as a subscript Represents.
S _B ≧ F _A (1)

In addition, since tasks that require the same resource cannot be executed simultaneously, the capacity constraint of such a resource can be expressed by a combination of the following line format and logical expression. Here, task C and task D require the same resource.
S _C ≧ F _D OR S _D ≧ F _C (2)

  If the required resources differ depending on the conditions, the constraint can be expressed by the following expression. In the following example, task E is resource 1 or resource 2, task F is resource 1, task G is resource 2, and variable M is a variable indicating task E resource selection. It is assumed that it is an integer variable that takes a value of 2.

(M = 1 AND (S _E ≧ F _F OR S _F ≧ F _E ))
OR
(M = 2 AND (S _E ≥ F _G OR S _G ≥ F _E )) (3)

  The complicated logistics constraints of the target process can be expressed as a logical logistics model by the constraint expressions as in the above formulas (1), (2), and (3). Here, the above-described evaluation function will be described in detail. As the evaluation function, for example, the following equation (4) is considered. This evaluation function is expressed as a weighted sum of three indicators: (b) whether the charge is ORP blowing, (b) waiting time on the turret, and (c) casting time.

Evaluation function z = Σ [f (i) (M _j = 2) w ₁ + g (i) w ₂ + h (i) w ₃ ] (4)
i ∈ all charges where M _j is a variable that is 1 if charge i is normal blowing and 2 if ORP blowing.
f (i) = 0 (M _j = 1), 1 (M _j = 2)
g (i) is the waiting time on the turret of charge i,
h (i) is the charging time of charge i,
w ₁ , w ₂ , w ₃ are weight parameters,
Respectively.

  The result of the discrete event simulator 10 of the above embodiment is output to the actual plant 16 shown in FIG. 1, and the actual plant 16 directly controls the operation of the actual plant 16. One charge obtained as the optimum steel production order can be produced or cast at the optimum steel production time obtained by the apparatus 13.

  Thus, when one charge is made, the current distribution state in the actual plant changes, so that information is taken out and supplied to the logical distribution model construction apparatus 11 and the linear distribution model construction apparatus 12. The The logical logistics model construction device 11 and the linear logistics model construction device 12 construct a logistics model in a line format based on the given current logistics state and logistics constraints, and supply them to the optimization computing device 13. The optimization calculating device 13 calculates the optimal steeling sequence and time of charging from the converter to the CC by evaluating the given distribution model formula with an evaluation function. Thereafter, the operation control is executed by repeating the same processing.

  In the above embodiment, the Petri net model is given as an example of the model of the discrete event simulator 10. However, the present invention is not limited to this, and the current distribution state is determined for each discrete event. Any simulator of a discrete event model that can output information can be applied.

  The discrete event simulator 10, the logical logistics model construction device 11, the linear logistics model construction device 12, the optimization calculation device 13, and the evaluation function setting device 14 are, for example, a CPU (Central Processing Unit), a RAM (Random Access). The microcomputer is composed of a memory), a ROM (read only memory), and the like, and can be realized by a computer such as a personal computer.

(Another embodiment according to the present invention)
Each means constituting the processing plan creation device and the operation control device, and each step of the processing plan creation method and the operation control method in the embodiment of the present invention described above is executed by a program stored in a RAM or ROM of a computer. It can be realized by doing. This program and a computer-readable storage medium storing the program are included in the present invention.

  Specifically, the program is recorded on a recording medium such as a CD-ROM, or provided to a computer via various transmission media. As a recording medium for recording the program, a flexible disk, a hard disk, a magnetic tape, a magneto-optical disk, a nonvolatile memory card, and the like can be used in addition to the CD-ROM. On the other hand, as the transmission medium of the program, a communication medium (wired line such as an optical fiber, etc.) in a computer network (LAN, WAN such as the Internet, wireless communication network, etc.) system for propagating and supplying program information as a carrier wave A wireless line or the like.

  Moreover, not only the functions in the above-described embodiments are realized by executing a program supplied by the computer, but also an OS (operating system) or other application software running on the computer. When each function of the above-described embodiment is realized jointly, or all or part of the processing of the supplied program is performed by a function expansion board or a function expansion unit of the computer, each of the above-described embodiment Such a program is also included in the present invention even when the function is realized.

It is a block diagram which shows the principal part structure of the processing plan preparation apparatus in the manufacturing process which is one embodiment of this invention. It is a figure for demonstrating the operation | movement (processing plan preparation method by this Embodiment) performed by the processing plan preparation apparatus by this Embodiment. It is a figure which shows the position in the production plan system of the processing plan preparation apparatus by this Embodiment. It is a figure for demonstrating formulation of the physical distribution model by this Embodiment.

Explanation of symbols

DESCRIPTION OF SYMBOLS 10 Discrete event simulator 11 Logical logistics model construction device 12 Linear logistics model construction device 13 Optimization calculation device 14 Evaluation function setting device 15 Linear logistics model holding device 16 Actual plant 17 Logical logistics model holding device

Claims (13)

In a processing plan creation device for creating a processing plan in a processing process for manufacturing multiple products with different multiple process paths,
A discrete event model that represents the processing flow and processing constraints of the above processing process with a discrete event model, and detects the processing state and processing constraints when an event occurs,
Logical distribution model that combines the line form obtained from the processing status and processing constraints of the processing process with logical operators for the target period (plan creation period) set in advance from the planning start date and time of the processing plan A logical logistics model holding device for holding
A linear logistics model holding device that selects and holds a linear logistics model that satisfies a predetermined condition among the linear logistics models included in the logical logistics model;
An evaluation function setting device for setting an evaluation function for evaluating the linear logistics model;
An optimization calculation device that performs an optimization calculation process using the linear logistics model and the evaluation function to calculate a processing instruction for the discrete event simulator,
A processing instruction is calculated for a period (instruction calculation period) preset from the present time by the optimization calculation device, and is given to the discrete event simulator to execute simulation, and only for a preset period (plan decision period). A series of simulation results obtained by repeating the process of finalizing the processing plan, setting the date and time immediately after the determined period as a new planning start date and planning the processing plan, as the processing plan in the processing process A processing plan creation device characterized by deciding.   The linear logistics model holding device analyzes the relationship between the linear formats of the logical logistics model and obtains conditions for satisfying each linear format, and as a result, the linear logistics model contained in the logical logistics model 2. The processing plan creation apparatus according to claim 1, wherein an infeasible linear logistics model that does not satisfy processing constraints is deleted, and only an executable linear logistics model is constructed and held.   The evaluation function is represented by a weighted sum of two indexes of an amount representing a desired processing method and apparatus selection status and a time difference from an occurrence time to an end time of each event. 2. The processing plan creation device according to 2.   The processing plan creation device according to claim 1, wherein the discrete event model is a Petri net model. An operation control device for creating an operation instruction in a processing process for manufacturing a plurality of products with different plurality of process paths,
A discrete event model that represents the processing flow and processing constraints of the above processing process with a discrete event model, and detects the processing state and processing constraints when an event occurs,
Logical distribution model that combines the line form obtained from the processing status and processing constraints of the processing process with logical operators for the target period (plan creation period) set in advance from the planning start date and time of the processing plan A logical logistics model holding device for holding
A linear logistics model holding device that selects and holds a linear logistics model that satisfies a predetermined condition among the linear logistics models included in the logical logistics model;
An evaluation function setting device for setting an evaluation function for evaluating the linear logistics model;
An optimization calculation device that performs an optimization calculation process using the linear logistics model and the evaluation function to calculate a processing instruction for the discrete event simulator,
A processing instruction is calculated for a period (instruction calculation period) preset from the present time by the optimization calculation device, and is given to the discrete event simulator to execute simulation, and only for a preset period (plan decision period). A series of simulation results obtained by repeating the process of finalizing the processing plan, setting the date and time immediately after the determined period as the new planning start date and planning the processing plan, as operation instructions in the processing process An operation control device characterized by deciding. In a processing plan creation method for creating a processing plan in a processing process for manufacturing a plurality of products with different multi-step paths,
A discrete event model that represents the processing flow and processing constraints of the above processing process with a discrete event model, and detects the processing state and processing constraints when an event occurs,
Logical distribution model that combines the line form obtained from the processing status and processing constraints of the processing process with logical operators for the target period (plan creation period) set in advance from the planning start date and time of the processing plan A logical logistics model holding device for holding
A linear logistics model holding device that selects and holds a linear logistics model that satisfies a predetermined condition among the linear logistics models included in the logical logistics model;
An evaluation function setting device for setting an evaluation function for evaluating the linear logistics model;
Using an optimization calculation device that performs an optimization calculation process using the linear logistics model and the evaluation function and calculates a processing instruction for the discrete event system simulator,
A processing instruction is calculated for a period (instruction calculation period) preset from the present time by the optimization calculation device, and is given to the discrete event simulator to execute simulation, and only for a preset period (plan decision period). A series of simulation results obtained by repeating the process of finalizing the processing plan, setting the date and time immediately after the determined period as a new planning start date and planning the processing plan, as the processing plan in the processing process A processing plan creation method characterized by deciding.   The linear logistics model holding device analyzes the relationship between the linear formats of the logical logistics model and obtains conditions for satisfying each linear format, and as a result, the linear logistics model contained in the logical logistics model 7. The processing plan creation method according to claim 6, wherein an infeasible linear logistics model that does not satisfy processing constraints is deleted, and only an executable linear logistics model is constructed and held.   The evaluation function is represented by a weighted sum of two indexes of an amount representing a desired processing method and apparatus selection status and a time difference from an occurrence time to an end time of each event. The processing plan creation method according to 7.   The method of creating a processing plan according to any one of claims 6 to 8, wherein the discrete event model is a Petri net model. An operation control method for creating an operation instruction in a processing process for manufacturing a plurality of products with different plurality of process paths,
A discrete event model that represents the processing flow and processing constraints of the above processing process with a discrete event model, and detects the processing state and processing constraints when an event occurs,
Logical distribution model that combines the line form obtained from the processing status and processing constraints of the processing process with logical operators for the target period (plan creation period) set in advance from the planning start date and time of the processing plan A logical logistics model holding device for holding
A linear logistics model holding device that selects and holds a linear logistics model that satisfies a predetermined condition among the linear logistics models included in the logical logistics model;
An evaluation function setting device for setting an evaluation function for evaluating the linear logistics model;
Using an optimization calculation device that performs an optimization calculation process using the linear logistics model and the evaluation function and calculates a processing instruction for the discrete event system simulator,
A processing instruction is calculated for a period (instruction calculation period) preset from the present time by the optimization calculation device, and is given to the discrete event simulator to execute simulation, and only for a preset period (plan decision period). A series of simulation results obtained by repeating the process of finalizing the processing plan, setting the date and time immediately after the determined period as the new planning start date and planning the processing plan, as operation instructions in the processing process An operation control method characterized by deciding. In a program for causing a computer to execute a processing plan creation method for creating a processing plan in a processing process for manufacturing a plurality of products with different plurality of process paths,
A discrete event model that represents the processing flow and processing constraints of the above processing process with a discrete event model, and detects the processing state and processing constraints when an event occurs,
Logical distribution model that combines the line form obtained from the processing status and processing constraints of the processing process with logical operators for the target period (plan creation period) set in advance from the planning start date and time of the processing plan A logical logistics model holding device for holding
A linear logistics model holding device that selects and holds a linear logistics model that satisfies a predetermined condition among the linear logistics models included in the logical logistics model;
An evaluation function setting device for setting an evaluation function for evaluating the linear logistics model;
Using an optimization calculation device that performs an optimization calculation process using the linear logistics model and the evaluation function and calculates a processing instruction for the discrete event system simulator,
A processing instruction is calculated for a period (instruction calculation period) preset from the present time by the optimization calculation device, and is given to the discrete event simulator to execute simulation, and only for a preset period (plan decision period). A series of simulation results obtained by repeating the process of finalizing the processing plan, setting the date and time immediately after the determined period as a new planning start date and planning the processing plan, as the processing plan in the processing process A computer program for causing a computer to execute a process for determining. A program for causing a computer to execute an operation control method for creating an operation instruction in a processing process for manufacturing a plurality of products through different multi-step paths,
A discrete event model that represents the processing flow and processing constraints of the above processing process with a discrete event model, and detects the processing state and processing constraints when an event occurs,
Logical distribution model that combines the line form obtained from the processing status and processing constraints of the processing process with logical operators for the target period (plan creation period) set in advance from the planning start date and time of the processing plan A logical logistics model holding device for holding
A linear logistics model holding device that selects and holds a linear logistics model that satisfies a predetermined condition among the linear logistics models included in the logical logistics model;
An evaluation function setting device for setting an evaluation function for evaluating the linear logistics model;
Using an optimization calculation device that performs an optimization calculation process using the linear logistics model and the evaluation function and calculates a processing instruction for the discrete event system simulator,
A processing instruction is calculated for a period (instruction calculation period) preset from the present time by the optimization calculation device, and is given to the discrete event simulator to execute simulation, and only for a preset period (plan decision period). A series of simulation results obtained by repeating the process of finalizing the processing plan, setting the date and time immediately after the determined period as the new planning start date and planning the processing plan, as operation instructions in the processing process A computer program for causing a computer to execute a process for determining.   13. A computer-readable recording medium in which the computer program according to claim 11 or 12 is recorded.

Images