JPH0535747A - Scheduling method - Google Patents

Abstract

PURPOSE:To make all evaluating indexes even and satisfactory even in the case of the presence of plural evaluating indexes of the scheduling result by assigning the job data to the resources based on the assigning priority calculated from the preferent points set against those evaluating indexes. CONSTITUTION:In a step 21 the priority is calculates when a job is assigned to the resources. Then a combination of the largest value is detected out of the calculated assigning priority in a step 22. In a step 23 when working time is calculated for the detected combination. This combination is assigned to a Gantt chart in a step 24. In a step 25 a fact whether any unassigned data are recognized or not. If so, the steps 21-24 are repeated for the unassigned data. If not, a step 26 is carried out so as to output the final result assigned repetitively as the result of assignment.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a scheduling system for allocating jobs to resources in various planning work.

[0002]

2. Description of the Related Art Scheduling services for allocating jobs to resources have been widely performed in various fields. For example, in process scheduling, order
(Job) is assigned to a device (resource), and in a vehicle delivery plan, work (job) is assigned to a worker (resource). Scheduling specialists have been performing these scheduling tasks manually for a long time, relying on their experience and intuition, but they depended heavily on the specialists, and the quality of the results was largely left to them. In recent years, the movement to computerize these scheduling tasks has become very popular. When computerizing the scheduling work, it is necessary to describe the allocation process procedurally. In this case, the allocation job is to take one job out of multiple jobs and allocate it to allocatable resources. The method of repeating until it disappeared was common. The evaluation of the scheduling result is important for computerization of scheduling work. If there is only one evaluation index, the value of the evaluation index can be made relatively good. For example, order
In the process scheduling that allocates (jobs) to devices (resources), if there is an evaluation index that wants to minimize the setup change time of the device, retrieve one from the order data and search for the resource that will reduce the setup change time. Allocate it. Note that, as a system related to this type, for example, the technique disclosed in JP-A-63-214874 can be cited.

[0003]

However, when there are a plurality of evaluation indexes of the scheduling result, it is difficult to set any evaluation index to a good value with only one scheduling result. Even in the above process scheduling, the allocation is biased toward a specific device in order to reduce the setup change time, and it is not always seen from the evaluation index that the operation rate of each device is uniform or the load leveling of the device is equalized. It is not always a good value. In such a case, some scheduling results that make the value of each evaluation index favorable are created, and one of them is selected according to the situation at that time. Therefore, it was not very satisfactory even when compared with the result when the expert manually performed the scheduling. The present invention has been made in view of the above circumstances, and an object of the present invention is to solve all the above-mentioned problems in the related art and to make all the evaluation indexes even when the scheduling result has a plurality of evaluation indexes. An object of the present invention is to provide a scheduling method capable of uniformly setting the value of p to a good value.

[0004]

The above object of the present invention is to create a plan having means for inputting a plurality of job data, means for allocating the input job data to one resource, and means for outputting the allocation result. In the apparatus, the scheduling method is characterized by allocating the job data to the resource based on the allocation priority calculated based on the priority point set corresponding to each of the plurality of evaluation indexes.

[0005]

In the scheduling system according to the present invention, the priority point table, which is provided corresponding to each of the plurality of evaluation indexes, can be used to freely set the priority point which is the basis of the allocation priority. Therefore, the value of the allocation priority can be set to a value in consideration of the balance of the values of the evaluation indexes, and the values of all the evaluation indexes can be made good with only one scheduling result.

[0006]

Embodiments of the present invention will now be described in detail with reference to the drawings. In the embodiments described below, the case where the present invention is applied to a general process scheduling operation in a manufacturing industry or the like will be described as an example. The process scheduling work shown in the present embodiment aims at two points: leveling the load on the apparatus and reducing the number of setup changes. Therefore, the evaluation index is the variation degree of the device load and the total time for setup change. FIG. 2 is an input / output related diagram of the plan creation device 1 according to an embodiment of the present invention. The plan creation device 1 inputs the order data 2 and the device data 3, allocates the order to the devices, and finally outputs the result as the allocation result 4. The interim allocation results are developed and managed in the main storage device 5 of the plan making device 1, and the interim allocation results are sequentially displayed on the Gantt chart of the display 6. The Gantt chart is a diagram for displaying a scheduling result in which the horizontal axis is time and the vertical axis is a device, and is generally used in planning work. FIG. 3 shows a Gantt chart in a state where nothing is assigned yet. In this example,
The start time is 8:00, the end time is 20:00 including overtime, and no lunch break is considered. FIG. 4 shows the manufacturing process of the painted plastic plate, which is the target of the scheduling work of this embodiment. The colorless plastic plate 11 is produced as a painted plastic plate 13 through a painting step 12. The coating process 12 has three coating devices 14, 15 and 16 respectively called a device (A), a device (B) and a device (C), regardless of the type of plastic plate to be produced. Coloring is possible with either device.

The contents of the order data shown in FIG. 2 are shown in FIG. In FIG. 6, the product name represents the product name of the plastic plate to be produced, the product color represents the coating color corresponding to the product, and the production amount represents the number of products to be produced. In addition, FIG.
FIG. 7 shows the contents of the device data of the above. The device name represents the name of the coating device 14, 15, 16 in FIG. 4, and the device capability represents the number of plastic plates that can be coated per hour. Since the device capacity is the same for all devices, the production amount and the painting time are proportional. In this example,
In order to give priority to the products having a long painting time, the data are sorted in descending order of production as shown in FIG.
Next, a procedure for allocating the order data shown in FIG. 6 on the Gantt chart shown in FIG. 3 will be described with reference to the flowchart shown in FIG. In step 21, the allocation priority when the job is allocated to the resource is calculated. Normally, calculation is performed for all combinations of allocatable jobs and resources that satisfy the constraint conditions. However, in this embodiment, since all products can be produced by any device, the constraint condition check is omitted. The following is used as the allocation priority calculation formula. Allocation priority = priority point X + priority point Y (1) In addition, both the priority point X and the priority point Y in the equation (1) are values referred to from the priority point table, and the priority point X is The priority regarding the load leveling of the apparatus is represented, and the priority point Y indicates the priority regarding the reduction of the setup change time. 9 shows a priority point table X referring to the priority point X, and FIG. 10 shows a priority point table Y referring to the priority point Y.

In the priority point table X of FIG. 9, the priority point X is determined according to the total allocated time of the device. For example, if the allocated time of the device is 6 hours, the priority point X
Is 80 points. In the priority point table Y of FIG. 10, the priority point Y is determined according to the relationship between the color of the product to be allocated and the color of the product previously allocated on the allocation device. The priority point Y corresponds to the setup change time of the setup time table shown in FIG. 11, and the combination of colors that require setup change has a low priority and the combination of colors that does not require setup has a high priority. Has been done. For example, if a yellow product is assigned to the device and then a blue product is assigned to the device, the priority points are 5 points. The formula shown above
The value of the allocation priority calculated by (1) is shown in FIG.
Shown in. The first allocation indicates that the priority is 110 no matter which product is allocated to which device. Next, in step 22 of FIG. 1, a combination of the maximum values is searched for from the calculated allocation priorities and taken out.
If the products have the same priority, select the product with the longest painting time among the products with the same priority, and if the products with the same priority have the same priority, the device with the lowest device number To choose. Therefore, in the example of FIG. 12, the combination of the product 5 and the device 1 is taken out.

Next, in step 23, the coating color working time for the combination taken out in step 22 is calculated. The formula for calculating the working time is shown below. Painting color work time = production amount / equipment capacity (2) When the product 5 and the equipment 1 are combined, the equipment capacity is 10 pieces per hour and the production quantity is 60 pieces, so the painting work time is 6 hours. .. Next, in step 24, the combination taken out in step 22 is assigned to the Gantt chart shown in FIG. When assigning an order to a device, assign it to the earliest available time of the assigned device. Therefore, the work of painting the product 5 on the device 1 is performed from 8:00 to 14:00.
Allocate time. Next, in step 25, it is checked whether or not there is order data that has not been allocated yet. If yes, steps 21 to 24 are repeated again, and if not, the process proceeds to step 26. Here, since there are eight order data that have not been assigned yet, step 21 to step 24
Will be repeated again. In this case, when the allocation priority is calculated in step 21, it becomes as shown in FIG. Since the product 5 is allocated to the device 1 in the first allocation, the allocation priority of the device 1 is low. Step 22
Then, the combination of the product 1 and the device 2 is taken out as the combination with the highest allocation priority. In step 23, the painting color work time is calculated to be 5 hours. In step 24, the work of painting the product 1 on the device 2 is performed from 8:00 to 13:00.
Allocate time.

At step 25, since there are seven order data that have not been assigned yet, steps 21 to 24 are performed.
Will be repeated again. In this case, when the allocation priority is calculated in step 21, it becomes as shown in FIG. Although the product 1 was allocated to the device 2 in the second allocation, the allocation time of the device 2 is only affected by the setup change because the allocated time is only 5 hours. In step 22, the combination of product 8 and device 3 is retrieved as the combination with the highest allocation priority. In step 23, the painting time is calculated to be 4 hours.
In step 24, the work of painting the product 8 on the device 3 is allotted for 4 hours from 8:00 to 12:00. At step 25, since there are six order data that have not been assigned yet, steps 21 to 24 are repeated. In this case, if the allocation priority is calculated in step 21,
It becomes like 5. Although the product 8 is allocated to the device 3 in the third allocation, the allocation time of the device 3 is only 4 hours, so the allocation priority of the device 3 is only affected by the setup change. In step 22, the combination of product 9 and device 3 is retrieved as the combination with the highest allocation priority. In step 23, the painting time is calculated to be 4 hours. In step 24, the work of painting the product 9 on the device 3 is allocated for 4 hours from 12:00 to 16:00. At step 25, since there are five order data that have not been assigned yet, steps 21 to 24 are repeated. In this case, when the allocation priority is calculated in step 21, it becomes as shown in FIG.

Since the product 9 is allocated to the device 3 in the fourth allocation, the allocated time of the device 3 is 8
Since it is time, the allocation priority of the device 3 is low. In step 22, the combination of product 2 and device 2 is retrieved as the combination with the highest allocation priority.
In step 23, the painting time is calculated to be 3 hours. In step 24, the work of coating product 2 on device 2
Allocate for 3 hours from 3:00 to 16:00. In step 25, since there are four order data that have not been assigned yet, steps 21 to 24 are repeated. In this case, when the allocation priority is calculated in step 21, it becomes as shown in FIG. Since the product 2 is allocated to the device 2 in the fifth allocation, the allocated time of the device 2 becomes 8 hours, so that the allocation priority of the device 2 is low. In step 22, the combination of product 6 and device 1 is retrieved as the combination with the highest allocation priority. In step 23, calculate the painting time
It will be time. In step 24, the work of painting the product 6 on the apparatus 1 is allocated for 3 hours from 14:00 to 17:00. In Step 25, the order data that has not been assigned is 3
Since there are some cases, Step 21 to Step 24 will be repeated again. In this case, when the allocation priority is calculated in step 21, it becomes as shown in FIG. Since the product 6 is allocated to the device 1 in the sixth allocation, the allocated time of the device 1 becomes 9 hours, so that the allocation priority of the device 1 is low. In step 22, the combination of product 3 and device 2 is retrieved as the combination with the highest allocation priority. In step 23, the painting time is calculated to be 2 hours. In step 24, the work of painting the product 3 on the device 2 is allocated for 2 hours from 16:00 to 18:00.

At step 25, since there are two order data that have not been assigned yet, step 21 to step 24
Will be repeated again. In this case, when the allocation priority is calculated in step 21, it becomes as shown in FIG. Since the product 3 is allocated to the device 2 in the seventh allocation, the allocated time of the device 2 becomes 10 hours, so that the allocation priority of the device 2 is low. In step 22, the combination of product 7 and device 1 is retrieved as the combination with the highest allocation priority. In step 23,
Calculating the painting color work time is 1 hour. Step 24
Then, the work of painting the product 7 on the apparatus 1 is allotted for 1 hour from 17:00 to 18:00. In step 25, there is one order data that has not been assigned yet, so step 21
~ Step 24 will be repeated again. in this case,
When the allocation priority is calculated in step 21, it becomes as shown in FIG. Since the product 7 is allocated to the device 1 in the eighth allocation, the allocated time of the device 1 becomes 10 hours, so that the allocation priority of the device 1 is low. In step 22, the combination of product 4 and device 3 is retrieved as the combination with the highest allocation priority. At this time, since the color of the product 9 previously allocated in the device 3 is blue and the color of the product 4 taken out is red, a setup change has occurred. In step 23, the painting color work time is calculated to be 2 hours. In step 24, the setup change time when switching from the blue product to the red product in FIG. 11 is 60 minutes, so the work of painting the product 4 on the device 3 is 60 minutes apart and is allocated for 2 hours from 17:00 to 19:00. .. In step 25, there is no order data that has not been allocated yet, so the process proceeds to step 26. Step 21 of Figure 1 ~
FIG. 5 shows a Gantt chart assigned by repeating steps up to step 25.

At step 26, steps 21 to 25 are carried out.
The final result assigned by repeating the above is output as the assignment result 4 in FIG. The output plan result,
The evaluation is performed using two evaluation indexes, namely, load leveling of the device and reduction of the number of setup changes. As shown in FIG. 5, the allocated time of each of the devices 1 to 3 is 10 hours,
All the devices are uniformly allocated, and the load leveling of the devices is satisfied. In addition, in FIG. 5, the case where the setup change occurs is from 16:00 in the device 3 to 17:00.
Only once until the time, and the minimum number of times required,
We are also satisfied with the reduction in the number of setup changes. As described above, according to the scheduling method shown in the present embodiment, the two evaluation indexes, namely the load leveling of the device and the reduction of the number of setup changes, are good for the values of the two evaluation indexes with only one scheduling result. Can be
In the above-described embodiment, the example in which the process scheduling for one day is performed has been described, but the number of days may be collectively scheduled. In this case, the process scheduling work generally aims at the following three points: compliance with the delivery date, leveling the load on the equipment, and reducing the number of setup changes.
It is the degree of variation in equipment load and the number of setup changes. After adding the delivery date data to the order data shown in FIG. 6 and sorting from the earliest delivery date, the order data is assigned to the Gantt chart according to the flowchart of FIG. 1 to comply with the delivery date and level the load on the device. With respect to the three evaluation indexes of reducing the number of setup changes, it is possible to improve the values of all the evaluation indexes with only one scheduling result.

The manufacturing process of the painted plastic plate in the above-described embodiment was only the painting process, but it may be composed of a plurality of processes such as a finishing process. When there are a plurality of processes, it becomes necessary to consider the work in process.
Work-in-process is a waiting state after processing in one process until processing in the next process. In this case, the process scheduling business generally aims at three points of minimizing the waiting time between the processes of the product, leveling the load on the device, and reducing the number of setup changes, and the evaluation indexes are the waiting time between the product processes and the device. It is the degree of load variation and the number of setup changes. In order to reduce the waiting time between steps of a product, it is necessary to schedule the processing so that after the processing in one step, the processing can be started immediately in the next step. For this purpose, in step 21 of FIG. 1, the allocation priority when the job is allocated to the resource may be calculated by the following formula. Allocation priority = priority point X + priority point Y + priority point Z (3) The priority point X and the priority point Y in the equation (3) are respectively referred to from the priority point tables of FIGS. 9 and 10 described above. It is a value.
The priority point Z represents the priority regarding the waiting time between product processes, and refers to the value from the priority point table in which the priority point changes according to the time difference between the start time of the job to be allocated and the end time of the previous process. Is. As a result, one of the three evaluation indexes of minimizing the waiting time between product processes, equalizing the load on the equipment, and reducing the number of setup changes
It becomes possible to improve the values of all the evaluation indexes only by the scheduling result of the number of times. Further, in the above embodiment, the priority points are calculated from the priority point table in the calculation of the allocation priority, but it is also possible to use the output value of the function that inputs the value related to the evaluation index as the priority point. Is. Here, the value related to the evaluation index corresponds to the allocated time in the priority point table for load leveling of the device in FIG. 9, for example. By converting the priority table into a function, it becomes possible to calculate the priority points at high speed, and the allocation processing can be speeded up as compared with the case of the priority point table.

[0015]

As described above in detail, according to the present invention, even when there are a plurality of evaluation indexes of the scheduling result, it is possible to make all the evaluation index values uniformly good values. It has a remarkable effect that a different scheduling method can be realized.

[0016]

[Brief description of drawings]

FIG. 1 is a flowchart showing an operation of a plan creating device according to an embodiment of the present invention.

FIG. 2 is an input / output related diagram of the plan creating device according to the embodiment.

FIG. 3 is a diagram showing a configuration of a Gantt chart in an unallocated state.

FIG. 4 is a diagram showing a manufacturing process of a painted plastic plate which is a target in the scheduling work of the embodiment.

FIG. 5 is a diagram showing a Gantt chart displaying a planning result.

FIG. 6 is a diagram showing an example of order data.

FIG. 7 is a diagram showing an example of device data.

FIG. 8 is a diagram showing order data sorted in descending order of production volume.

FIG. 9 is a diagram showing a table X that refers to a priority point X.

10 is a diagram showing a table Y that refers to a priority point Y. FIG.

FIG. 11 is a diagram of a setup time table showing setup change times.

FIG. 12 is a diagram showing an allocation priority calculation result (first time).

FIG. 13 is a diagram showing an allocation priority calculation result (second time).

FIG. 14 is a diagram showing an allocation priority calculation result (third time).

FIG. 15 is a diagram showing an allocation priority calculation result (fourth time).

FIG. 16 is a diagram showing an allocation priority calculation result (fifth time).

FIG. 17 is a diagram showing an allocation priority calculation result (sixth time).

FIG. 18 is a diagram showing an allocation priority calculation result (seventh time).

FIG. 19 is a diagram showing an allocation priority calculation result (8th time).

FIG. 20 is a diagram showing an allocation priority calculation result (9th time).

[Explanation of symbols]

1: Planning device, 2: Order data, 3: Device data,
4: Allocation result, 5: main memory, 6: display.

Claims (1)

Claims: 1. A means for inputting a plurality of job data,
Allocation priority calculated on the basis of priority points set corresponding to each of a plurality of evaluation indexes in the plan creation device having means for allocating the input job data to one resource and means for outputting the allocation result. A scheduling method characterized by allocating job data to resources based on