JP5119949B2 - Production plan creation device - Google Patents

Description

The present invention relates to a raw production planning system to develop a plan for producing production lot of a plurality of steps in a plurality plants.

  Conventionally, regardless of the product, in the production plan, a skilled person creates a plan through trial and error while taking into consideration the production cost, production constraints, and the delivery date of the production lot. However, with recent improvements in computer performance, a method has been proposed that automatically formulates a production plan and achieves optimality.

  In the above Patent Document 1, a continuous casting set (casting by connecting a plurality of converter charges) of a production lot that satisfies the constraints related to the manufacturing order is extracted, and the continuous casting set and a continuous casting set that has already been determined; From the state of the empty (unoperated) position in the converter, the continuous casting set is sought as a candidate. The above processing is repeated to determine the manufacturing order for one day, and the time information of the continuous casting set is determined based on the time-related constraints, thereby creating a manufacturing schedule. When the manufacturing schedule satisfies the evaluation standard (primary evaluation), it is set as a production plan candidate. A method has been proposed in which a predetermined number of candidates are created and an optimum one is selected based on an evaluation criterion (secondary evaluation).

  In Patent Document 2, when the latest planning information at the start of planning is fetched from each database, a production planning model for the planning period is created based on the fetched planning information for the order to be planned. Is building. The production plan model is composed of an order delivery date and a manufacturing flow, a processing time in each manufacturing process, an operation schedule of each manufacturing process, and an evaluation function for judging whether the production plan is good or bad. Using the production planning model, the processing timing and the processing amount in each manufacturing process for each order so that the evaluation function is minimized (best) has been proposed.

JP 2000-315109 A JP 2006-260462 A

  As in the above-mentioned Patent Document 1, in production plan creation for producing multi-process production lots for each factory, the production plan considering production costs, production constraints, and delivery dates of production lots is made for each factory. Therefore, production plans can be formulated for each factory. However, in the case of a method of preparing a production plan for each factory after determining the quantity of production lots to be manufactured in advance for each factory, the optimal distribution such as the equipment load of each factory and the delivery date of the production lot is impaired. there is a possibility. On the other hand, in the production plan of manufacturing lots of multiple processes targeting multiple factories at the same time, the scale of the plan becomes large by considering even the combination between multiple factories, and even with the current computer in a realistic time There was a problem that there were cases where it could not be solved.

  Further, in the above-mentioned Patent Document 2, since the target period of the planning is divided and the solution is obtained based on the delivery deadline and the priority order of the lots, the problems of the above-mentioned Patent Document 1 are partially solved. However, since the interference of the passing process of the production lot is not considered, there is a problem that the leveling of the load of each passing process is not considered.

The present invention has been made to solve the above-mentioned problems. First, it enables leveling of the load of each passing process, and secondly, when the problem is large and small. Correspondingly, prepared as in solving a batch process and when solving division processing problems, and aims to provide a raw production planning system to develop a plan for producing production lot of a plurality of steps in a plurality plant To do.

The production plan creation device according to the present invention is a production plan creation device for creating a plan for producing a production lot of a plurality of processes in a plurality of factories. A storage means for storing order information including a process and a requested delivery date; and a calculation means for creating a manufacturing schedule for each factory based on the information stored in the storage means, wherein the calculation means is stored in the storage means. Based on the stored manufacturing lot information of each manufacturing lot and the manufacturing lot passing process included in the order information, the first calculation processing for factory distribution of each manufacturing lot, and the first calculation processing After that, a second calculation for performing distribution within each factory based on the requested delivery date for each order of the production lot included in the order information stored in the storage means Management and, after said second arithmetic processing, the third arithmetic processing and, each production lot, which is stored in the storage means to create a production schedule for each plant based on the information of the production lot distributed to each plant Based on the information of factory that can be manufactured and order information, each manufacturing lot handles the manufacturing schedule for each factory that satisfies the leveling of the number of facilities that each factory manufactures and the requested delivery date. The fourth calculation process for creating the manufacturing schedule is performed, and the first, second and third calculation processes and the fourth calculation process are selectively executed.

Next, the present invention will be described based on specific examples.
When five kinds of production lots as shown in Table 1 are produced in two factories (F1 / F2), the lot numbers F1T1 and F2T1 are designated as factories, and the remaining FFT1, FFT2, FFT3 Will describe an example in which both F1 and F2 can be manufactured. Since F1T1 and F2T1 are designated to be manufactured at factories F1 and F2, the production lots are assigned to FFT1, FFT2, and FFT3, respectively. At this time, the calculation is performed so that the factories including the number of the process lots A, B, and C occupied by the passing lots of FFT1, FFT2, and FFT3 and the manufacturing lot designating the factories are equalized.

The amount occupied by each passing process of the production lot manufactured at each factory is calculated. Equations (1) to (3) show the relationship in which the production lots that can be assigned are assigned to F1 and F2.
Here, n3_F1 and n3_F2 indicate values when the production lot FFT1 is allocated to the F1 factory and the F1 factory. Similarly, n4_F1 and n4_F2 indicate values when the production lot FFT2 is allocated to the F1 factory and the F1 factory, and n5_F1 and n5_F2 indicate values when the production lot FFT3 is allocated to the F1 factory and the F1 factory. F1_A, F1_B, and F1_C are the numbers occupied by the passage processes A, B, and C in the F1 factory, and F2_A, F2_B, and F2_C are the numbers occupied by the passage processes A, B, and C in the F2 factory.

  Here, the distribution of the FFT1, FFT2, and FFT3 production lots is calculated so that the passing processes A, B, and C of the F1 factory and the F2 factory are made as uniform as possible while satisfying the expressions (1) to (3). (First step or first arithmetic processing). At that time, n3_F1, n3_F2, n4_F1, n4_F2, n5_F1, and n5_F2 are determined such that the differences between F1_A and F2_A, F1_B and F2_B, and F1_C and F2_C become as small as possible.

  In the steps so far, the number of distribution of the production lots of the FFT1, FFT2, and FFT3 to the F1 factory and the F2 factory has been determined, and it is determined which delivery date is allocated according to each distribution number. As shown in Table 1, the delivery date has a normal range. In order to classify these delivery dates into several categories, the average delivery date range is used, and those with similar average values are treated and classified as the same delivery date. Here, it is assumed that the delivery date of each production lot is classified into three types D1, D2 and D3.

  The following equation (4) shows a relational expression in the case where the respective delivery dates in each production lot are allocated to the F1 factory and the F2 factory. Equation (5) shows the relationship between the number of delivery steps A, B, and C to the F1 and F2 factories (n3_F1, n3_F2, n4_F1, n4_F2, n5_F1, and n5_F2) and the delivery time as evenly distributed as possible. Equation (6) represents the number of delivery dates that are classified into three delivery dates (D1, D2, D3) in the F1 factory and the F2 factory.

  Here, the classification of each delivery date in each factory, the number of D1_F1, D2_F1, D3_F1 and the number of D1_F2, D2_F2, D3_F2 are equally distributed in each factory n31_F1 to n33_F1, n41_F1 to n43_F1, n51_F1 to n53_F1 and n31_F2 to n33_F2, n41_F2 to n43_F2, n51_F2 to n53_F2 are determined (second step or second arithmetic processing).

  By these calculations, the production lot manufactured at each factory and the data on which delivery date within the production lot is manufactured at the F1 factory or the F2 factory are prepared. And the manufacturing schedule for every factory is created based on the data (3rd process or 3rd arithmetic processing).

  That is, in the present invention, when the first to third steps (first arithmetic processing to third arithmetic processing) are processed over time (time-division processing), the first step There are cases where the third process is processed in a lump (collective process), and any one of them is appropriately selected and executed.

  According to the present invention, it is possible to create a production plan in which the load of each passing process is leveled, and it is also possible to create a production plan that takes into account production lots and delivery dates of a plurality of processes in a plurality of factories. ing.

  FIG. 1 is a block diagram showing the configuration of an embodiment of the present invention. The factory production management computer is composed of a host computer, and stores delivery date based on the production lot order and passing process data necessary for production (functions as a storage means of the present invention). These data are taken into the planning computer (functioning as the computing means of the present invention), and the plan is created and returned to the host computer, thereby being reflected in the factory operation plan. Production lot data determined from the production plan database determined based on the order and the operating state of the factory is taken into the planning computer, and processing based on the flowchart shown in FIG. 2 is performed. Hereinafter, with reference to FIG. 2, a case where the problem is solved by the division process and a case where the problem is collectively processed will be described.

(1) When solving a problem by division processing [STEP 1]
In STEP1, it is distributed to a plurality of factories based on the production lot passing process.
Here, as an example, distribution of the production lots of FFT1, FFT2, and FFT3 is calculated so that the passage processes A, B, and C of the F1 factory and the F2 factory are made as uniform as possible. N3_F1 and n3_F2, n4_F1, and n4_F2, n5_F1, and n5_F2 are determined so that differences between F1_A and F2_A, F1_B and F2_B, and F1_C and F2_C are as small as possible. As an example, the method can be solved as a linear programming method (mixed integer programming problem) using the equations (1) to (3) as constraints and equalizing the number of distribution to each factory as an evaluation function.

[STEP2]
With the processing in STEP 1 above, the number of allocation of the production lots of FFT1, FFT2, and FFT3 to the F1 factory and the F2 factory has been determined. In STEP2, it is determined which delivery date is allocated according to each allocation number. . As shown in Table 1, the delivery date has a range. In order to classify these delivery dates into several categories, the average delivery date range is used, and those with close average values are treated as the same delivery date. Here, the delivery date of each production lot is classified into three types D1, D2 and D3 (note that the number of classifications is set as appropriate). Here, equations (4) to (6) are used as constraint conditions, and the equalization of the number of lots for delivery date is used as an evaluation function, which can be solved as a linear programming method (mixed integer programming problem).

[STEP3]
From the processing in STEP 1 and STEP 2 described above, data on manufacturing lots manufactured in each factory and which delivery date in the manufacturing lots are manufactured in the F1 factory or the F2 factory is prepared. As shown in FIG. 3, based on the number of distribution to the F1 factory and the number of distribution to the F2 factory, the number of production lots for each delivery date is determined for each day from 1 to 5 days. To do. The determination method is as follows.

  Applying the linear programming method (mixed integer programming problem) to each factory, using the number of lots assigned to each factory obtained in STEP 1 and STEP 2 as a constraint, and using the equalization of the number of daily passing processes and delivery dates as an evaluation function Can be solved. The F1 factory schedule plan and the F2 factory schedule plan according to this embodiment are shown in the right column of FIG. In these plan results, the production lots are arranged within the delivery date (in the figure), and the daily total number is also equal.

(2) When solving a problem by batch processing The processing from the allocation of manufacturing lots to multiple factories to the final determination of the number of daily manufacturing lots is determined in a batch.
Using the above formulas (1) to (6) as constraints, linear programming (mixed integer programming problem) with an evaluation function that equalizes the number of distribution to each factory, the number of lots for delivery, and the number of daily passing processes ).
The F1 factory schedule plan and the F2 factory schedule plan according to this embodiment are shown in the right column of FIG.
In FIG. 4, since the problems are processed in a lump, the number of passing steps every day is equalized compared to the plan result of FIG. 3 in which the processing is divided.

  However, when solving by batch processing, the number of combinations increases, and when the scale of the problem is large, it may not always be possible to solve in a practical time. The scale of the problem depends on the number of lots, delivery date conditions, planning period, and the like. Therefore, as an implementation method, first, a method of solving by batch processing is applied, and if it takes time until a solution is obtained, it is solved by division processing. Alternatively, there is a method of solving by dividing processing first and applying a batch processing solution if time is available. In addition, when the contents of the production lot are frequently changed and the re-planning is repeated, there is a method using only the division processing.

It is the block diagram which showed one Embodiment of this invention. It is a flowchart which shows the process sequence of this invention. It is the figure which showed the example of plan preparation by this invention. It is the figure which showed the example of plan preparation by this invention.

Claims (1)

In a production plan creation device that creates a plan for producing production lots of multiple processes at multiple plants,
Storage means for storing factory information of each production lot and order information including a passing process and a desired delivery date of each production lot;
Computation means for creating a manufacturing schedule for each factory based on the information stored in the storage means,
The computing means is
Based on the manufacturing lot passing process included in the information of the factory that can be manufactured and the order information of each manufacturing lot stored in the storage means, a first calculation process that performs factory allocation of each manufacturing lot;
After the first calculation process, a second calculation process for performing distribution within each factory based on the requested delivery date of the order of the production lot included in the order information stored in the storage means;
After the second calculation process, a third calculation process for creating a manufacturing schedule for each factory based on information on the manufacturing lot distributed to each factory ;
Based on the factory information and order information of each manufacturing lot stored in the storage means, each manufacturing lot equalizes the number of occupying facilities manufactured in each factory and manufactures each factory satisfying the requested delivery date. A fourth calculation process that creates a manufacturing schedule by handling a plurality of factories at once.
And
A production plan creation device that selectively executes the first, second, and third arithmetic processes and the fourth arithmetic process .

Images