JPH08215994A - Scheduler support system and method therof - Google Patents

Abstract

PURPOSE: To cope with irregulars while reducing stock before beginning to work by deciding a product charge program based on a production plan through a work beginning indication lead time, and setting the work begining indication lead time to be an optional time in the the range in which a net processing time is settled among the work beginning times of respective processes. CONSTITUTION: A jobshop factory introducing a scheduler is controlled so as to automatically improve operation rate while considering actual facility efficiency and facility load, based on a planned work beginning indication master. The production control executed by a jobshop scheduler 2 deals with a schedule planned with a net processing time against the actual facility, a work beginning program against respective facilities is executed in a work beginning indication system 3 by a work beginning indication planned based on a work beginning indication lead time. The work beginning indication lead time is a time that the net lead time summing up net processing times of facilities used for respective works and carrying times is multiplied by a prescribed number.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a scheduler support system and a scheduler support method used as a scheduler optimizing device in a job shop factory or the like in which a scheduler is installed.

[0002]

2. Description of the Related Art Conventionally, in a job shop scheduler used in a large-scale job shop factory, a so-called dispatching method is basically used as a basic control, and equipments (hereinafter referred to as equipments) related to the work of each process.
The products (lots) in progress are efficiently operated.

FIG. 4 is a schematic view showing a state where product lots are processed by each facility in a job shop factory. Here, the process A consists of operations a to e (each of which is also equipment). Or work a to e
Although it is similar even if it is regarded as a process, here, the dispatching will be described assuming that the process A is composed of the works a to e.

In this process A, the work flow is a, b,
The order is c, d, e, b, and in FIG. 4, R1 to R7
Is indicated by an arrow. After completing these six operations, the product lot is transferred to the next process. Process A from the process before Process A or by the first input of raw materials
There is a product lot of 4 lots in the equipment a for performing the first work as a pre-work-in-process inventory, and one lot is currently under work. In the equipment b, 5 lots that have been transported after the work of the equipment a or the equipment e have been carried out are in-process inventory, and the equipment b is a 3 lot batch, and 3 lots are currently in process. . In the equipment c, there is currently no lot in process and no lot in process. The equipment d has two lots of pre-working stock and one lot in process, and the equipment e has one lot of pre-working stock and work-in-progress.

Now, the equipment b has a stock of 5 lots before work in progress, and the equipment b is in full operation. In the equipment b, since the lot arrives not only from the equipment a but also from the equipment e,
Only then will it be known when and how many lots will be in process. Therefore, the schedule of the facility b cannot be decided in advance, and what can be decided is how to deal with the stock lot before work in progress. In other words, the one that is currently in process determines the next lot to be processed from the stock before work, triggered by the fact that the work is completed (unloaded or issued).
How to deal with such deciding rules is a matter of dispatching, and some such dispatching techniques have been shown.

For example, there is one disclosed in Japanese Patent Laid-Open No. 5-61879. In this case, each process on the manufacturing line is provided with a process-waiting product memory that stores each product waiting to be input in the order of input, and the process of the corresponding product detected by the priority change request receiving means. It discloses an interleaving method for changing the order of introduction of the corresponding products in the product memory waiting for the process introduction based on the progress status and the like. For example, in the equipment b of FIG. 4, since it is a three-lot batch processing operation, there is a process of deciding the priority order of the lots in the inventory before work in progress (waiting for process input) and selecting three lots from the five lots. Done.

[0007] On the other hand, in order to solve the above-mentioned problems, there is a system in which a certain margin is provided in the inventory in order to control the securing of the product inventory waiting to be input (inventory before work) in each process. There is a technique described in Japanese Patent Laid-Open No. 5-73574. This is to solve the above-mentioned problems in order to realize the given production plan, in particular, to put in a bottleneck process, calculate the delivery margin for each product, and use a certain algorithm. This is a production control device characterized by including a determining means for determining a dispatching rule on the basis of the dispatching rule.

For example, in FIG. 4, in the equipment c, both the stock before work and the work in progress are currently 0. Therefore, in the equipment c, the work can be immediately started when the lot arrives. However, since the processing capacity of the equipment b is limited, ideally, immediately after the processing of the lot in process is completed, it is desirable that the pre-work inventories for three lots always exist immediately before. In order to do so, the lot payout of the equipment a is made to have a little margin, so that the operating load of the equipment b, which is a bottleneck, is reduced and the efficiency is improved.

Further, on the other hand, in the case of a material arrangement delay, equipment failure, etc., the progress status is grasped, and when a delay occurs, the production plan of the corresponding product is changed according to the progress. Japanese Patent Laid-Open No. 5-128123 is related to the plan management device, and there is a rescheduling technique for an unexpected situation such as equipment failure.

The above-mentioned prior art is basically based on a dispatch method by a scheduler or the like as a production in-process instruction at a job shop factory, but in many cases irregular support is flexible in the scheduler situation.

[0011]

However, in the technique of the above-mentioned Japanese Patent Laid-Open No. 5-61879, it is premised that each product has a product waiting to be put in each process in the manufacturing flow, and If there is an event such as a flow stop due to equipment failure in a factory with a narrowed down inventory, and there are no products waiting to be put into each process, the equipment at that factory is forced to stop operating. However, there is a problem that the delivery date of the product cannot be guaranteed and the product cannot be produced in due course.

The technique disclosed in Japanese Patent Laid-Open No. 5-73574 is intended to solve the problem of unscheduled production by only a simple interleaving method and to guarantee the delivery time of the product.
It is said that the inventory control of products waiting to be put into each process cannot be implemented, and only a margin for delivery is provided to realize the production plan, leading to an increase in in-process inventory from the perspective of the entire factory. There's a problem.

Even in the case of the rescheduling technique, the production plan is reviewed, and if the production plan cannot be changed or the corresponding inventory is held in order to allow the production plan to be changed, inventory control is performed. It's not realized. Also, in the case of making it flexible, the calculation result of the scheduler becomes ambiguous, and it is not the original solution to operate more efficiently. In addition, the irregular problem has the aspect that everything is unpredictable, and that point has not been solved.

In view of the above problems, as a better scheduler in the job shop factory, the processing time for the schedule, the waiting time, etc. are registered more strictly.
It should work, but in order to be strictly defined, irregular measures such as equipment failure must be taken into consideration. Optimized schedules are not possible with flexible data. SUMMARY OF THE INVENTION Therefore, an object of the present invention is to provide a scheduler support system capable of irregularly adapting and realizing the most optimized scheduling while reducing the stock before work.

[0015]

In order to solve the above-mentioned problems, the structure of the present invention is applied to a factory equipped with an automatic production factory management system equipped with a job shop scheduler for executing an automatic production instruction in a job shop factory. The automatic production factory management system has a work instructing system in which a net lead time, which is the total of the processing time for scheduling of each facility in the factory, and a work instructing lead time in which a margin time is added to the net lead time are registered. , The product introduction schedule based on the production plan is determined by the in-process instruction lead time, the net lead time is used in the in-process time calculation for determining the actual equipment start-up by the job shop scheduler, and the in-process instruction lead time Arbitrary time within the range that the net processing time falls within the in-process time of each process It is to be set. Further, a characteristic configuration of the related invention is that the work-in-process system includes an automatic updating means for updating the work-instruction lead time every predetermined period.

The configuration of the method for carrying out the present invention as an invention comprises a job shop factory in which a scheduler is introduced, a job shop scheduler for executing an automatic production instruction by the scheduler, and an automatic production factory management system. At the factory, the in-process instruction system calculates the net lead time, which is the net processing time for scheduling, and the margin time is added to the net lead time to calculate the in-process instruction lead time. Registering in the in-process instruction system, inputting materials according to the in-process instruction lead time, and calculating the in-process time for determining the actual equipment start-up using the net lead time by the job shop scheduler. Alternatively or additionally, the in-process instruction lead time depends on the amount of inventory before in-process of the facility, the idle time of each facility, the operating rate and delivery rate compliance rate of each facility, the failure frequency of each facility, and the quality abnormality of each product. It is characterized in that it is automatically updated by a predetermined procedure according to the measurement of the process stop period.

[0017]

[Operation] In order to avoid obscuring the calculation result of the scheduler itself, the net processing time of each equipment is set as the net lead time for scheduling, and the schedule result calculated by the scheduler is actually processed without being increased by coefficient processing. Build with the net processing time and cycle time of. This includes the transportation time between each equipment and the automatic carrier (AG
V) ・ Strictly defined from the movement of inter-process carriers.
Further, in the present invention, each product or lot has an in-process instruction lead time to be input earlier (that is, earlier in time) than the input time determined by the production plan delivery date. Therefore, the timing of introducing the product into the job shop factory is determined according to the in-process instruction lead time. This in-process instruction lead time is registered as a fixed coefficient multiple or more of the net lead time, which is the total of the processing time for scheduling of the corresponding product (duplicated lead time). During operation, data such as inventory before work in progress, idle time of each equipment, operation rate of each equipment, delivery rate compliance rate, etc., constant failure frequency at each equipment, and each product The process stop time due to quality abnormality is measured, and the in-process instruction lead time is automatically updated by the in-process system according to a predetermined procedure.

Therefore, due to the duplicated lead time, there are two types of net in-process time, that is, the net lead time which is the sum of the machining time for scheduling, and the in-process instruction lead time which is obtained by multiplying the net lead time by a factor. The difference in the lead time of the product will be launched ahead of schedule. On the other hand, in each process (equipment) where the product flows out, the in-process lot is sequentially determined by the net lead time which is the actual processing time by the dispatching method as in the past. Therefore, except when an irregular event occurs,
The amount of spare time allocated by the front-loading is distributed as a product waiting to be supplied (stock before work) in each of the following processes.

[0019]

The adoption of only the simple postal system presupposes the existence of pre-work-in-process inventory of each equipment. Therefore, the effect is realized only when there is the pre-work-in-process stock. However, in the configuration of the present invention, Pre-work inventories can be guaranteed and irregular problems can be absorbed.

In other words, the difference between the two lead times serves as a temporal buffer to ensure the in-process inventory of each process (equipment), thereby creating the in-process inventory. Therefore, if there is an event such as a flow stop due to equipment failure, etc., or if there are no products waiting to be put into each process, the corresponding equipment will be forced to stop operating and the product delivery deadline is guaranteed. It cannot be done, and it is inevitable that it will be production. ”

Further, the automatic updating of the in-process instructing lead time shortens the time of the in-process inventory which is always wasted, so that the input time can be shortened, which also "guarantees the completion delivery time of the product. However, it is impossible to control the inventory of products waiting to be put in each process, and only a margin is provided as a delivery time margin to realize the production plan, which leads to an increase in the in-process inventory of the entire factory. The problem of can be solved.

Further, due to the duplication of the lead time, extension of the in-process instruction lead time of the product resulting from the adjustment of the scheduler and detection of the in-process inventory before each equipment obtained by the process management become a bottleneck in the process flow. As a result, it is possible to clarify and understand the cause of irregularity that does not arrive on time as scheduled, and it is possible to promptly resolve the problem.

Furthermore, according to the present invention, since the schedule is calculated in advance by including the allowance time, and a proper schedule is calculated, it is possible to avoid ambiguity of the scheduler itself such as coefficient processing of the schedule result as in the conventional case. Also, since there is no ambiguity factor, it is possible to independently evaluate the performance of the job shop scheduler itself. In addition, in addition, in the configuration of the present invention, a general-purpose scheduling system can be easily introduced, and it has been difficult to discriminate in the past, such as "whether the schedule algorithm is bad, or is there originally no facility capacity / process capacity?" The difference can be clearly shown, and it will easily support the introduction of the scheduler into the job shop factory.

[0024]

EXAMPLES The present invention will be described below based on specific examples. (Configuration) FIG. 1 is an overall block configuration diagram of a scheduler support system to which the present invention is applied. In FIG. 1, an automatic production factory management system 1 is a higher-level system in which the system of the present invention is incorporated, which is highly automated by being connected to an equipment group by a LAN or the like, and also being connected to an inter-process transport vehicle or an AGV. It is the management system of the job shop factory. The job shop scheduler 2 to be evaluated / supported is composed of a high-speed AI scheduler composed of a high-speed computer having a learning function and a dispatcher (both not shown) (here, the high-speed AI).
Is irrelevant to the configuration of the present invention). As an alternative, this system may be a general-purpose job shop factory scheduler, in which case the general-purpose scheduling system itself can be evaluated by the present invention.

Since process management for inventory management and progress management plays an important role, it is shown as a configuration included in the automatic production factory management system 1, and results such as inventory data are shown in the work instructing system 3. The actual work-in-process instruction is calculated with reference to the above. At that time, the standard of the in-process instruction is the in-process instruction lead time.

Also, the characteristic doubled lead time of the present invention (net lead time, in-process instruction lead time)
And the in-process instruction system 3 that develops the production plan into the product introduction plan by the in-process instruction lead time are incorporated in the automatic production factory management system 1. The in-process instruction system 3, as shown in the lower half of FIG.
This is a block configuration composed of two systems (the configuration of FIG. 1 does not mean that there are three system blocks below the in-process instruction system 3 here;
Indicates that it is composed of two systems). The basic unit registration system 5 uses, as facility basic units, processing time for each recipe (condition), cycle time, batch method, size,
It is for defining detailed processing time for each device such as buffer capacity and load time. It defines the basic unit registration system 51 of the process basic unit that describes the flow of each product, and the duplicate in-process instruction lead time. The basic unit registration system 5 is composed of two systems (the configuration of FIG. 1 means here that the basic unit registration system 5 is composed of two systems).

The production plan data prepared by the job shop scheduler 2 or another system is shown in FIG.
The production plan development system 6 develops the actual product introduction plan in association with the inventory data managed by the process management system No. 3, and the lead time used here is the in-process instruction lead time. Then, according to the finished product introduction plan, the material arrangement system 7 executes the necessary amount calculation and the order calculation for the part numbers required for the material arrangement, and prepares the data so that the material can be automatically supplied to the equipment. The result of the input plan is delivered to the job shop scheduler 2, where the process control of the job shop factory is performed in real time based on the input plan.

The time when the product should be introduced into the factory is determined by tracing the delivery date determined by the production plan. At that time, the net lead time that is the sum of the net operating time of each facility in the factory and the transportation time is determined. If you decide the time to put it in, it will not be possible to deal with some irregularities such as delays and breakdowns, resulting in a delay in delivery. Therefore, it is necessary to make some allowances for product launch timing. In the present invention, the charging time is determined not at an early stage by mischief, but at an appropriately determined time. That is a production plan using the in-process instruction lead time here.

The production plan is first constructed by the job shop scheduler 2. In order to configure the production planning data so that the in-process instruction system 3 can actually instruct the in-process together with the data of the basic unit registration system 5 and the material arrangement system 7 by referring to the inventory data of the process management system 4. Calculated by the production plan development system 6. At this time, the in-process instruction lead time is used. Therefore, the work in process for each facility includes a margin. The production plan development system 6 converts the production plan data provided from the automatic production factory management system 1 or another system into work instructing data while associating it with the inventory data from the process management system 4 in FIG. To do. The specific action is shown in FIG.

(FIG. 2) FIG. 2 is a schematic diagram for explaining standard work-in-process product (lot) delivery date setting, required work-in-process instruction calculation, and the like in general production management.
The production plan master 8 that indicates when, what and how much is needed is data that is drafted by the production planning department and determines the flow of processes throughout the job shop factory.
This is the most basic data. In addition to this, an inventory master 9 for managing how many stocks can be allocated.
To prepare the inventory so that the stock status can be grasped. Further, basic unit data 10 such as lead time, defective rate, and component composition is constructed such as what to standardly arrange, and in-process instruction calculation 11 based on these 8, 9, and 10 data. Is carried out to form a work-in-process instruction master 12. This also includes the delivery date when completed. Then, factories without a scheduler carry out work in progress as planned based on this result.

At the job shop factory where the scheduler is installed, management is performed so as to automatically improve the operating rate while observing the actual equipment operating rate and equipment load based on the planned work-in-process instruction master 12. In the present invention, the production management executed by the job shop scheduler 2 is
A schedule scheduled for the net processing time for the actual equipment is handled, and the in-process schedule for each equipment is executed by the in-process instruction planned by the in-process instruction system 3 based on the in-process instruction lead time.

As will be described later with reference to FIG. 3, this in-process instruction lead time is a time obtained by multiplying the net lead time, which is the total of the net processing time and the transportation time of the equipment used in each work, by a predetermined multiple, and the breakdown thereof. For each case, divide each process (or work) by its own ratio. Then, depending on the division, the work-in-process time for each facility is determined.

(FIG. 3) In FIG. 3, when it is assumed that a series of manufacturing steps are constituted by steps A to F, lead time
On the side of (1), the net lead time showing the total of the actual machining time of each net process calculated for scheduling (total scheduling machining time) is defined as the lead time.
In (2), the in-process instructing lead time is a time that is a predetermined multiple of the lead time (1) for in-process instructing.
Here, this lead time is the lead time.
The ratio of the processing time of each process in (1) is doubled as it is. Here, the process A represents the process to be set next to the current lot, the process to be set, or the current process if it is in process, and the leading process in the product (lot) that has not been introduced. The lead time (2) is an in-process instruction lead time for a product that has not been input to the factory, and is the time from the current time to the completion of delivery of the relevant product lot for products in any process in the factory.

In this way, the lead time (1) of the actual machining time is approximately doubled, so that in each period of the lead time (2), the waiting time is other than the actual machining time. Become. Before processing, if the preceding process is delayed for some reason and the pre-work-in-process inventory does not occur, it will be time to wait for arrival. If the processing ends early, there will be a margin. For example, the process of B in FIG.
In (2), the period is stipulated as B ', but the actual lot processing is performed at the time such as when processing is started immediately, and the remaining time until C'is before the in-process of process C. Will be in stock. Even if it is processed at a time such as when the arrival from the process A is delayed, there is still room to absorb the irregularity of the previous process and the production can be continued according to the scheduled schedule. However, if it is too late, it will affect the subsequent process, so it is desirable to be able to process as soon as possible. However, even if an irregularity that occurs later than this time occurs, it is possible to absorb it in the margin of the lead time (2) after the process C.

In the high-speed AI scheduler included in the job shop scheduler 2, the lead time (2) corresponding to the remaining machining time is divided as described above, the corresponding processes are allocated within the range, and according to the scheduling, For each facility, the dispatcher starts the products (lots) with the highest priority in the inventory before work in each process. Scheduling including how to determine the priority order depends on the in-process instruction lead time. Then, if there is too much free time or the like in each step of the lead time (2), it is automatically set again by statistical processing.

The judgment of shortening is to determine the amount of inventory before work of equipment,
Free time of each facility, operating rate of each facility and delivery rate compliance rate,
Measure the failure frequency at each facility and the process stoppage period due to quality abnormalities for each product, and carry out based on the results of several times according to the procedure of AI. This review is executed, for example, in a predetermined period such as once a week. In some cases, the review may prolong the period of each step of the lead time (2). This is because that is the condition under which it operates efficiently, and it is by no means redundant or obscure.

Even if a general-purpose scheduler is introduced into the configuration of the present invention as a scheduling system instead of the high-speed AI scheduler, the result of irregular factors can be clearly shown, and due to the dual lead time configuration, The general-purpose scheduler itself can be evaluated by measuring and managing the lots that are in process during the spare time by the in-process instruction system that uses the net lead time.

[Brief description of drawings]

FIG. 1 is a conceptual diagram of the configuration of a scheduler support system of the present invention.

FIG. 2 is a basic conceptual diagram of production management.

FIG. 3 is a conceptual diagram showing the configuration of a duplicated lead time.

FIG. 4 is an explanatory diagram of dispatching at a job shop factory.

[Explanation of symbols]

 1 Automated production factory management system 2 Job shop scheduler 3 Work in progress system 4 Process management system 5 Basic unit registration system 6 Production plan development system 7 Material arrangement system 8 Production plan master 9 Inventory master 10 Basic unit data such as lead time and parts composition 11 In-process instruction calculation 12 In-process instruction master 51 Process basic unit registration system 52 In-process instruction basic unit registration system

Claims (4)

[Claims] 1. A net lead time, which is the sum of processing time for scheduling of each equipment of a factory, in a factory having an automatic production factory management system equipped with a job shop scheduler for executing an automatic production instruction at the job shop factory. And the in-process instruction lead time in which a margin time is added to the net lead time is registered in the automatic production factory management system, and the product launch schedule based on the production plan is determined by the in-process instruction lead time. The net lead time is used in the in-process time calculation for determining the actual equipment start-up by the job shop scheduler, within a range in which the net processing time falls within the in-process time of each step of the in-process instruction lead time. A scheduler support system characterized by being set at an arbitrary time. 2. The scheduler support system according to claim 1, wherein the work-in-progress system has an automatic updating means for updating the work-in-process instruction lead time every predetermined period. 3. A job shop factory in which a scheduler is installed, equipped with a job shop scheduler for executing automatic production instruction by the scheduler, and a factory equipped with an automatic production factory management system. The net lead time, which is the net processing time, is calculated, the margin lead time is added to the net lead time to calculate the in-process instruction lead time, and these duplicated lead times are registered in the in-process instruction system to input the material. Is performed by the in-process instruction lead time, and the in-process time for determining the actual equipment activation is calculated by the job shop scheduler using the net lead time. 4. The in-process instruction lead time depends on the amount of inventory before in-process of the facility, the idle time of each facility, the operating rate of each facility and the delivery rate compliance rate, the failure frequency of each facility, and the quality abnormality of each product. The automatic update according to a predetermined procedure according to the measurement of the process stop period.
Described scheduler support method.