JPH05346926A - Method and device for production planning - Google Patents

Abstract

PURPOSE:To obtain an optimum schedule result by making the arrangement of objects of planned arrangement meet inhibitory restriction conditions and recommendatory restriction conditions by using simple constitution and planning the arrangement of the objects of planned arrangement according to priority corresponding to the importance of the recommendatory restriction conditions. CONSTITUTION:This production planning device is provided with a file 2 which contains the inhibitory restriction conditions and recommendatory restriction conditions, an inference means 10 for an inhibition rule which finds the allocatable period of the objects of the planned arrangement according to the inhibitory restriction conditions and arranges the objects of planned arrangement in the increasing order of allocation tolerance obtained from the allocatable period. and an inference means 20 for a recommendation rule which sets priority for the recommendatory restriction conditions and corrects the arrangement on the basis of the arrangement of the objects of planned arrangement obtained by an inference means for inhibitory restrictions according to the priority of the recommendatory restriction conditions so that the conditions are satisfied.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a production plan creating method and its creating apparatus used when creating a production plan, a job shop scheduling, etc. using a planning type expert system, and particularly to a constraint condition rule. The present invention relates to a production plan creating method and an creating apparatus for obtaining an optimum schedule result while effectively utilizing the constraint contents.

[0002]

2. Description of the Related Art Conventionally, when scheduling an object to be planned and arranged, that is, an arrangement for making a production plan for a lot using a planning-type expert system, a startable date and delivery date of each lot, constraint conditions between preceding and following processes, The smallest allocation allowance (ratio between the allocatable period and the allotment required period) after deciding the allocatable period of each lot, while taking into account forbidden constraints such as constraints with other related lots. Starting with, the allocation plan is set up sequentially within the assignable range.

Further, although there is an area in the lot that can be placed within the planning period, there are cases where the lot cannot be placed due to the relationship between the forbidden constraint condition and the placement order. Once the other planned lots are placed in the unassigned state, the allocatable period is calculated again, and the lots placed in the unassigned state are placed again in order, and scheduling is performed so that all lots fall within the planned period. Is going.

In the following, a concrete example will be explained. When arranging parts 1, parts 2, parts 3, and parts 4 on two machines, 1M / C and 2M / C, and processing them, "the same within the plan period If there is a constraint condition that "the same component should not be assigned at the same time", if the component placement plan as shown in Fig. 3A is made while considering the constraint condition, the same component will not overlap at the same time. However, it is possible to sufficiently satisfy the above constraint conditions.

[0005]

Therefore, the above-mentioned production planning method has the main purpose of ensuring that each lot is delivered within the planning period. Therefore, the constraint condition may satisfy the prohibited constraint condition exclusively. Although it is possible, there are various requirements in the actual production plan. For example, in addition to the constraint condition that "the same parts should not be assigned at the same time within the planning period," In some cases, it is desirable to arrange the color and the color in a continuous manner as much as possible. "

However, in the conventional production planning method, the prohibitive constraint conditions are exclusively satisfied, and the above-mentioned request "the same mold or color is arranged continuously among the parts as much as possible". It was difficult to satisfy the condition that "it is desirable to do", and as a result, it was not possible to obtain a satisfactory scheduled result.

The present invention has been made in view of the above circumstances, and is a method for producing a production plan, in which planned placement objects are placed so as to satisfy the recommended constraint conditions in addition to the forbidden constraint conditions and obtain an optimum schedule result. The purpose is to provide.

Another object of the present invention is to satisfy the forbidden constraint condition and the recommended constraint condition by using a simple configuration for the arrangement of the planned layout object, and the importance level of the recommended constraint condition. The object of the present invention is to provide a production plan creation device that obtains an optimum schedule result by arranging the objects to be planned and arranged according to the priority according to.

[0009]

In order to solve the above-mentioned problems, the invention according to claim 1 divides a constraint condition rule for defining a production plan into a prohibition constraint condition and a recommended constraint condition, and a plan arrangement target. In the production plan creating method, the objects are arranged so as to satisfy the prohibition constraint condition, and then the arrangement of the planned arrangement object is corrected so as to satisfy the recommended constraint condition based on the arrangement result.

According to the second aspect of the invention, a rule database file for storing a constraint condition rule defining a production plan by dividing it into forbidden constraint conditions and recommended constraint conditions, and a plan arrangement based on the forbidden constraint conditions. In addition to determining the allocatable period of the target object, the inference means for the prohibition rule that sequentially arranges the planned allocation target objects in the order of the allocation margin obtained using this allocatable period and the recommended constraint condition are given priority. Degree is set, and based on the placement of the planned placement object obtained by the prohibition rule inference means, the placement of the planned placement object is corrected so as to be satisfied in accordance with the priority order of the recommended constraint conditions. This is a production plan creation device provided with inference means for recommended rules.

[0011]

Therefore, in the invention corresponding to claim 1, by taking the above-mentioned means, it is possible to appropriately arrange the objects within the planning period by satisfying the forbidden constraint condition for the objects to be planned, but at the same time, the work can be performed simultaneously. If the recommended constraints are satisfied while considering the efficiency and other conditions, a desirable placement plan can be constructed at the same time, and an optimal schedule can be established.

[0012] Next, the invention according to claim 2 is, for a period during which the allocation object of the planned arrangement can be allocated under a certain forbidden constraint condition,
Further, after the allocation allowance is calculated from the allocatable period, the allocation targets are arranged in the ascending order of the allocation allowance, so that the planned arrangement target objects can be properly arranged within the planning period. Moreover,
Based on the placement of the planned placement objects that have been obtained so as to satisfy the forbidden constraint conditions, the layout modification is performed so as to satisfy the priority order of the recommended constraint conditions as much as possible, so that the user's request is sufficiently satisfied. Can make a satisfactory placement plan.

[0013]

EXAMPLE An example of the method of the present invention will be described below.

Usually, when allocating each lot using the planned type export system, allocation is performed while applying the constraint condition rule to each lot. For planning problems that have recommended constraints, divide them into forbidden constraints and recommended constraints, execute inference that satisfies the forbidden constraints, and determine lot placement.
Based on this lot allocation, the allocation is modified while considering the recommended constraint conditions. Here, the prohibitive constraint conditions are, for example, the startable date and delivery date of each lot, the constraint conditions with the preceding and following processes, the constraint conditions with other related lots, and various other rules to be prohibited, and each lot is mainly prohibited. It means a rule to arrange so that it enters within the planning period.

On the other hand, the recommended constraint condition rule is a rule determined when "this is preferable" from the viewpoint of work efficiency, work environment, etc., and is arranged for each lot while fully utilizing the user's intention. Means the rule to do. Therefore, in the case of this recommended constraint condition, it is possible to set the priority by applying the user's request and apply it. At this time, if the recommended constraints are given priority, lots will be assigned according to the order of priority and the inference is made.

Therefore, according to the method of the above embodiment, after satisfying the prohibition constraint condition for each lot,
Since it is arranged while executing inference in the order of priority that should satisfy the recommended constraint conditions, it is possible to satisfy both constraint conditions at the same time, and it is possible to make a production plan while incorporating various user requests. is there. Next, an embodiment of the production plan creating apparatus according to the present invention will be described with reference to the drawings.

FIG. 1 is a functional block diagram showing an embodiment of the device of the present invention. This device has an input data file 1 in which each lot information necessary for making a schedule of a production plan, such as an available start date, a delivery date, and a lot number, is stored, and a prohibition constraint for applying an inference method of an expert system. The rule database file 2 in which the condition rules and the recommended constraint condition rules are stored, and the rules in the rule database file 2 are applied to the lot information stored in the input data file 1 to perform inference to arrange each lot. It consists of an inference section that creates a plan.

The inference unit executes the inference based on the prohibition constraint condition among the constraint conditions, and the inference means 10 for the prohibition rule.
And recommended rule inference means 20 for executing inference under recommended constraint conditions.

The prohibition rule inference means 10 is
Data reading means 11 for reading all lot information necessary for production planning and at least prohibitive constraint condition rules from each of the files 1 and 2 and storing them in a working memory (not shown). From the lot information, lot start date for each placement target lot,
Allocatable period determining means 12 for deciding the allocatable period for each lot while taking into account forbidden constraint conditions such as delivery date, constraint conditions relating to lots before and after, and constraint conditions with other related lots, allocatable period and allocatable request Allocation order determination means 13 is provided for determining the allocation allowance from the ratio with the period and for determining the temporary allocation order of the lots in consideration of the allocation allowance and the priority of each lot and the other priorities.

Whether the lots provisionally allocated in the allocation order by the allocation order determining means 13 can be actually allocated,
That is, the reallocation determining means 14 for determining whether or not the reallocation is necessary, and when the reallocation determining means 14 determines that the reallocation is not necessary, the allocation candidate is selected, that is, the temporarily allocated lot information is selected and the time series is selected. When it is determined that reassignment is necessary, the allocation candidate selection means 15 that is stored in the buffer memory, selects lot information to be reassigned from the lot information that has already been provisionally assigned based on the reassigned lot information. Reassignment candidate selection means 16 and these selection means 1
The real allocation is performed using the lot information selected as the allocation candidates selected in 5 and 16, and if the allocation is impossible here, the allocation is again sent to the allocation possible period determining means 12, and when the allocation is completed, the lot allocation file Allocation means 18 for allocating to 17 is provided. If assignment is not possible,
The temporary allocation order is determined again by changing the priority by the allocation order determination means 13.

On the other hand, in the recommended rule inference means 20, the data reading means 21 for reading the placement result of the lot placement file 17 and the recommended constraint condition rule, and the assignable lot which is the placement result of the lot placement file 17 has priority. Recommended rule placement means 22 for sequentially performing inference from highly recommended recommendation rule rules to determine placement, and assignment means 23 for assigning to the lot placement file 17 according to the placement determined by this placement means 22 are provided. There is. Next, the operation of the apparatus configured as described above will be described with reference to FIG.

First, the data reading means 11 in the prohibition rule inference means 10 fetches data such as all lot information and prohibition constraint rule from each of the files 1 and 2, but a working memory on a computer, though not shown. (ST1). Thereafter, the interruptable period determination means 12 prohibits constraint conditions such as the possible date to start the placement target lot from the lot information, the delivery date, the constraint conditions for the preceding and subsequent lots, and the constraint conditions related to other lots. The applicable period for each lot is determined while applying the rule (ST2). In other words, the allocatable period can be individually determined uniquely from the date when the lot can be started and the delivery date, but here, the allot information is comprehensively determined.

After that, the allocation order determining means 13 determines the temporary allocation order of which lot is to be arranged first. That is,
The provisional lot placement order is determined in consideration of the ratio (assignment allowance) of the assignment request period (lead time) and the assignable period for each lot and the placement priority of each lot (ST3).

If the tentative lot placement order is determined as described above, the relocation determining means 14 can appropriately assign the determined tentative lot placement order in the actual assignment. It is determined whether or not (ST4). When it is determined that the allocation is possible, the allocation candidate selection means 15 selects the temporarily allocated lot information and temporally stores it in the buffer memory (ST5). On the contrary, when it is determined that allocation is impossible, the re-allocation candidate selection means 16 selects lot information (replacement lot information or interrupt lot information) to be re-allocated from the lot information that has already been provisionally allocated based on the re-allocation lot information. Is selected (ST6).

Here, the allocating means 18 comprises the selecting means 15,
The real allocation is performed using the lot information of the allocation candidate selected in 16, and if the allocation is not possible, the allocation is again sent to the allocatable period determining means 12. On the other hand, when the allocation is completed, the lot allocation file 17 Are allocated to the arrangement area (ST7). Then, it is determined whether or not the arrangement of all lots is completed (ST8), and if not completed, the process returns to ST2 and the same process is repeatedly executed.

Therefore, by using the prohibition rule inference means 10 as described above, the parts 1 to 4 are connected to two machines 1M /
When arranging C and 2M / C, the rule that "the same component cannot be assigned to 1M / C and 2M / C at the same time" is applied as a prohibitive constraint condition.
The lots will be allocated in a lot arrangement as shown in FIG.
What happens is that the same parts are not assigned to each machine at the same time, and step 2 is assigned to each part in succession after the end of step 1, which satisfies the prohibitive constraint conditions sufficiently. This is because

However, in the case of such a component arrangement, each of the components 1 to 4 has product information (model, color). It is desirable to arrange the elements as consecutively as possible. ”Therefore, the prohibitive constraint condition alone cannot be satisfied. This is because the molds are not continuous.

Therefore, the device of the present invention is provided with the inference means 20 for the recommended rule and performs the processing after step ST9 in FIG. 2 so as to satisfy the recommended constraint condition as much as possible.

That is, after reading the lot arrangement data of the lot arrangement file 17 and the recommended constraint condition rule of the rule database file 2 in step ST9, the assignable lot is selected according to the priority of this recommended constraint condition rule. Perform (ST10). Then, the selected allocatable lot is provisionally placed in the lot placement file 17 and other placement areas of the memory (ST11).

Thereafter, in step ST12, it is determined whether or not the allocatable lot temporarily allocated is satisfied with the previous recommended rule of higher priority, and if it is satisfied, the temporary allocation is regarded as the correct arrangement. It is determined and the temporary placement is moved to the main placement area (ST13). Here, if the previous recommended rule is not satisfied, this placement process is skipped.
Further, in the next step ST14, it is determined whether or not the placement according to the recommended rule has been performed for all lots (ST
14) If there are still lots left, step ST
The procedure returns to 10 and the same processing is repeatedly executed. On the other hand, when the placement determination of the recommended rules for all lots is performed, the processing ends. As a result, the final schedule result is stored in the main arrangement area at this point.

Therefore, in the recommended rule inference means 20,
After obtaining the placement result that satisfies the prohibition constraint condition, the rule is applied to this placement result according to the priority of the recommended rule to select the assignable lots and satisfy the recommended constraint condition. As a recommended constraint condition, for example, a rule that “each part has the same product information as consecutive as possible” is used, the first priority is “type A” (part 1-1, part 3-1), and “type B”. "(Part 4-
1, the component 2-1) is made continuous, and as the next priority, “color red” (component 3-1 and component 4-1) is made continuous so that the arrangement configuration as shown in FIG. Can be It should be noted that the “color white” cannot be made continuous because other conditions (type or color) are destroyed when it is made continuous and the recommended constraint condition having a high priority cannot be satisfied.

Therefore, according to the configuration of the above-described embodiment, the constraint condition rule is divided into the prohibited constraint condition and the recommended constraint condition and stored in the file, while the inference unit is used as the forbidden rule inference means 10. It is divided into the recommended rule inference means 20. Then, the prohibition rule inference means 10 calculates the allocatable period in lot units while taking into consideration various forbidden constraint conditions, and obtains the allocation allowance. Then, each lot is allocated while considering the allocation allowance and the allocation priority on the lot. In this state, there is a possibility that the temporal plan will be satisfied, but it is difficult to satisfy the recommended constraint conditions. ..

Therefore, by providing a recommended constraint condition having a priority and allocating in a desirable arrangement in the order of the priority, it is possible to sufficiently satisfy the user's request, and, by extension, work efficiency and work environment are sufficient. It is possible to make a production plan that takes into consideration. Further, although the recommended constraint conditions are applied based on the placement result of the forbidden constraint conditions, each inference means 10 and 20 individually perform inference using rules, so that a very simple program can be used to achieve good results. The placement result can be obtained.

In the above embodiment, in the production planning of each lot, the prohibition rule inference means 10 considers the lot arrangement by applying the prohibition constraint condition so that the inferencing means 10 mainly falls within the plan period. In the inference means 20, the lot allocation is performed by applying the recommended constraint conditions in the preferred priority order as much as possible in consideration of the above allocation result, but the same idea is effective for other various production plans. It can be used for. Further, the recommended constraint condition may be applied without setting the priority. Besides, the present invention can be variously modified and implemented without departing from the scope of the invention.

[0035]

As described above, according to the present invention, the following various effects are exhibited.

First, according to the invention of claim 1, the planned placement object can be placed so as to satisfy the recommended constraint condition in addition to the prohibited constraint condition, and the flexible and optimum schedule result can be obtained.

According to the second aspect of the present invention, it is possible to sufficiently satisfy both the forbidden constraint condition and the recommended constraint condition by using a simple structure for the layout of the planned layout object, and the recommended constraint condition is satisfied. An optimal schedule result can be realized by making a placement plan of planned placement objects according to the priority according to the importance.

[Brief description of drawings]

FIG. 1 is a functional block diagram showing an embodiment of a production plan creation device according to the present invention.

FIG. 2 is a flowchart illustrating the operation of the apparatus shown in FIG.

FIG. 3 is a diagram showing an example of lot arrangement when the device of the present invention is used.

[Explanation of symbols]

DESCRIPTION OF SYMBOLS 1 ... Input data file, 2 ... Rule database file, 10 ... Prohibition rule inference means, 12 ... Allocation period determination means, 13 ... Allocation order determination means, 14 ... Relocation determination means, 15 ... Allocation candidate selection means, 16 ... Reassignment candidate selection means, 17 ... Lot placement file, 20 ... Recommended rule inference means, 22 ... Recommended rule placement means, 23 ... Allocation means.

Claims (2)

[Claims] 1. A constraint condition rule that defines a production plan is divided into a prohibition constraint condition and a recommended constraint condition, and the objects to be planned are arranged so as to satisfy the prohibition constraint condition. In addition, the production plan creation method is characterized in that the layout of the planned layout object is corrected so as to satisfy the recommended constraint condition. 2. A rule database file for storing constraint conditions rules for defining a production plan, which are divided into forbidden constraint conditions and recommended constraint conditions, and an allocatable period of a plan placement target object based on the prohibited constraint conditions. The inference means for the prohibition rule that sequentially obtains the planned placement objects in the ascending order of allocation allowance obtained using this allocatable period, and sets the priority for the recommended constraint condition A recommended rule inference means for modifying the layout of the planned placement objects so as to be satisfied in accordance with the priority order of the recommended constraint conditions based on the placement of the planned placement objects obtained by the constraint reasoning means. A production planning device characterized by being provided.