JP2006107167A - Scheduling system, scheduling program and scheduling method - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To distribute a load of arithmetic processing, accommodate a flexible change in the number or function of process handling apparatus and enable a change of the function of each process handling apparatus to a desirable function, in scheduling for assigning tasks set in successively generated jobs to a plurality of process handling apparatuses. <P>SOLUTION: A scheduling system comprises a master device 1 and a plurality of slave devices 2 disposed for respective process handling apparatus. The master device 1 determines a job handling order (S1) and sends handling request information about each task in the order to all the slave devices 2 (S2). In response to the request, the slave devices 2 return task handling proposal information (S11). According to the information, the master device 1 determines the process handing apparatus to which the task is assigned (S3) and notifies the determined task assignment to the slave devices 2 (S4). According to the notification, the slave devices 2 update task handling plan information about the corresponding process handling apparatus (S13). <P>COPYRIGHT: (C)2006,JPO&NCIPI

Description

  The present invention relates to a scheduling system, a scheduling program, and a scheduling method for creating a job processing plan when flowing sequentially generated jobs on a production line (processing line, assembly line, chemical process, steel process, etc.).

In a production line (processing line, assembly line, chemical process, steel process, etc.), processing such as processing and assembly is performed sequentially by a processing device (hereinafter referred to as a “process processing device”) that performs each processing step. The product is output. At that time, a job in which one or a plurality of tasks to be processed by the process processing apparatus is sequentially generated by an order, etc., and the tasks set in the job are assigned to the plurality of process processing apparatuses in any order. Scheduling to plan is performed. Here, the job represents a unit of one product or one order content, and the task represents a unit of processing for one batch by the process processing apparatus.
In general, what kind of tasks should be processed in which order for each job is determined in advance (set to the job). Ordinarily, jobs that are generated sequentially are based on a so-called FIFO (First In First Out) method in which they are processed in the order of generation (arrival order). In addition, when a priority is set for a job, and a FIFO method is used as a rule, if a job with a high priority occurs, it is necessary to give priority to processing a job with a higher priority than the first job. In some cases.
With regard to such objects, in the conventional scheduling method, the number and function of the process processing devices (for example, the type of the processing tool installed in the processing device (an example of the process processing device)) is set in advance as a “constraint condition”. However, scheduling is performed on a set of jobs so that each task is assigned to a plurality of process processing apparatuses so that mutual interference (interference scheduled for processing) does not occur.

By the way, in recent process processing devices (processing devices, assembly devices, etc.) for processing tasks, multifunctionalization (generalization) is progressing. For example, in a processing machine, in addition to increasing the number of types of processing tools that can be installed at the same time and enabling automatic switching, the processing function can be changed by changing the processing tools to be installed. Multi-functionalization that enables various types of processing has been attempted.
In the conventional scheduling method, it is necessary to change the setting of the “constraint condition” in order to perform scheduling that effectively uses such a multifunctional process processing apparatus.
Here, the change of the “constraint condition” means that the target problem is to obtain an optimal solution while satisfying the “constraint condition” (minimization of evaluation index such as cost). The problem of scheduling changes to a problem of examining the trade-off between the constraint dissatisfaction amount and the improvement degree of the evaluation index such as the cost after the dissatisfaction of the “constraint condition” is recognized. For this, for example, a method based on the “Lagrange relaxation method” is known. In the “Lagrange relaxation method”, the objective function is first multiplied by a coefficient (Lagrange multiplier) in the constraint equation, and the equation is minimized (or maximized). Furthermore, the deviation (constraint violation amount) in each constraint expression is grasped with respect to the solution obtained by this, and if the deviation is large, the Lagrange multiplier is increased and the expression is minimized (or maximized) again. ) Is repeatedly solved to simultaneously satisfy the constraint equation and minimize (or maximize) the objective function.
Such an optimization method is a so-called “batch type”. That is, this is a method of setting a set of a predetermined amount (predetermined number) of jobs and determining a schedule for a predetermined time ahead for the set of jobs.

On the other hand, a so-called auction method is known as a product sales method. In this case, the product provider indicates the content of the product and solicits the highest bidder, and the user performs a bid for which the amount is designated, and makes a bid for the target product to the highest bidder.
In this regard, Patent Document 1 discloses a transportation means determination method that provides air mail information and truck information via the Internet to a transportation requester who desires transportation of luggage and determines the transportation route of the luggage by an auction method. ing.
JP 2003-223574 A

However, in the conventional general scheduling method, if there is a change in the number or function of process processing devices, the above-mentioned “constraint condition” must be changed and rescheduling must be performed every time, and the operation is complicated. There was a problem.
Further, according to the “batch type” optimization method, such as scheduling based on the “Lagrange relaxation method”, a mathematically optimal (for example, minimum cost) solution (task) for a set of a predetermined amount of jobs. Although a processing schedule and a function (such as a tool attached to the processing apparatus) provided in the process processing apparatus can be obtained, there is a problem that the calculation load is very high. Furthermore, if this is applied to jobs that occur sequentially (arrivals) irregularly, jobs with high priority that occur as interrupts, or targets that cause failures in process equipment, etc., many high-load operations are repeated. As a result, there is a problem that the calculation load becomes higher.
On the other hand, the network-based auction method can distribute the processing load between the server side (master side information processing device) and multiple client sides (slave side information processing devices) without using complicated distributed processing methods. The load on each information processing apparatus can be reduced.
On the other hand, Patent Document 1 shows that the auction method is applied to the problem of determining a route for transporting a package, but there is no specific disclosure about the application to scheduling of a job to be sent to a production line. .
In addition, as described above, in the scheduling problem, it is necessary to interrupt and process a subsequent job having a high priority in the already determined processing plan, and to change the function of each process processing device to a desirable function. However, the patent document 1 or the like does not show specific contents corresponding to such processing.
Accordingly, the present invention has been made in view of the above circumstances, and the object of the present invention is to distribute the load of arithmetic processing in performing scheduling for assigning tasks set in sequentially generated jobs to a plurality of process processing devices. Provided are a scheduling system, a scheduling program, and a scheduling method capable of flexibly responding to changes in the number and functions of process processing apparatuses and capable of processing for changing the functions of the respective process processing apparatuses to desired functions. There is.

In order to achieve the above object, the present invention relates to a job in which one or a plurality of tasks to be processed by a process processing apparatus constituting each process of a production line is set and sequentially generated. A scheduling system comprising: a master device that is assigned to any one of the plurality of slave devices, and a plurality of slave devices that are provided so as to be communicable with the master device for each process processing device and that manage a task processing plan in the corresponding process processing device Because
Processing function storage means for storing processing function information having information on the types of tasks that can be processed by each of the process processing devices;
Processing plan storage means for storing task processing plan information for the task assigned to each of the process processing devices;
Job processing plan storage means for storing job processing plan information in which information related to the process processing device to which processing for each task in each job is assigned and its processing plan is set;
Job processing order determining means for determining a processing order of the jobs that are sequentially generated in the master device;
In accordance with the determined processing order of the job and the predetermined processing order of the task in each job, processing request information in which at least information on the type of the task is sequentially set for each task is transmitted from the master device to all the slaves. Processing request means to be transmitted to the device;
In the slave device, the task requested to be processed can be allocated to the process processing device based on the received processing request information and the processing function information and the task processing plan information related to the corresponding process processing device. Processing plan draft generation / reply means for generating processing plan draft information for the task and returning it to the master device;
In the master device, task assignment means for determining the process processing device to which the task is assigned based on each of the received processing plan proposal information and updating the job processing plan information;
An assignment result notifying means for notifying the determined assignment result of the task from the master device to the slave device;
In the slave device, task processing plan update means for updating the task processing plan information related to the corresponding process processing device based on the notification of the task allocation result;
It is comprised as a scheduling system characterized by comprising.
With such an autonomous decentralized system, in performing scheduling for assigning tasks set to jobs that occur sequentially to a plurality of process processing devices, each of the master device and the plurality of slave devices allows a master-to-slave The load of arithmetic processing can be distributed using communication. In addition, the change in the number of process processing devices can be flexibly handled by simply increasing or decreasing the number of independent slave devices provided for each of the process processing devices without requiring changes in the processing contents of each device. . In addition, it is possible to flexibly cope with a change in function of the process processing apparatus simply by changing the processing function information corresponding to the process processing apparatus.

In the slave device, a first slave-side plan canceling unit that updates the task processing plan information so as to cancel a part or all of the task processing plan already assigned to the corresponding process processing device; ,
Slave-side plan cancellation notification means for notifying the master device of slave-side plan cancellation information related to the task whose processing plan has been canceled by the first slave-side plan cancellation means;
In the master device, the job processing plan information is canceled so as to cancel the task whose processing plan has been canceled on the slave device side based on the notified slave side plan cancellation information and the processing plan of the task linked to the task. Master side plan cancellation means to update,
Cancellation request means for transmitting cancellation request information for requesting cancellation of the processing plans of the chained tasks from the master device to the slave device;
In the slave device, the task processing plan information relating to the corresponding process processing device is updated based on the received cancellation request information, and a job including the task is again set as a target for auction as “task assignment incomplete”. And two slave-side planned canceling means.
With such a configuration, even if it is necessary to cancel a part or all of the task processing plan that has already been assigned due to troubles and inspections such as failures in the process processing apparatus, the rescheduling is automatically performed. Is done.

Further, if a priority is set for each of the jobs, and the job processing order determination means determines the processing order of the jobs based on the generation order of the jobs and the priority, the job In the case where a job with a high priority is generated later, processing that gives priority to the job can be performed.
In this case, the processing request means transmits the processing request information with the priority set from the master device to all the slave devices,
The processing plan draft generation / reply means generates the processing plan draft information for preferentially processing the new task with a high priority, and returns the master plan to the master device.
The first slave-side plan canceling unit cancels the task processing plan with the lower priority when the task processing plan update unit corresponding to the priority processing updates the task processing plan information. It is possible to update the job processing plan information.
As a result, when a job (task) with a high priority is generated later, the processing plan of the task with a low priority necessary for the priority processing is canceled, and the canceled task is chained. The task processing plan to be canceled and the rescheduling associated with the cancellation are automatically performed. Therefore, flexible scheduling corresponding to the generation order and priority of the jobs that are generated sequentially is possible.

The slave device further includes an evaluation index setting unit that sets an evaluation index of the processing plan plan in the processing plan plan information of the task generated by the processing plan plan generation / reply unit, and the task allocation unit However, it is possible to determine the process processing device to which the task is assigned based on the evaluation index set in the processing plan proposal information.
As a result, in determining which one of the plurality of processing plan proposal information returned from each of the slave devices is to be adopted, the master device performs an evaluation of grasping the detailed situation of each of the process processing devices. Compared with the case of extreme concentration, the load of evaluation calculation can be distributed to each of the slave devices.
In this case, the evaluation index setting means is the same type as the task that is the target of the processing plan, the degree of process margin of the task, the number of continuous processes of the task for the same job, in the corresponding process processing device. It is conceivable that the evaluation index is set based on one or more of the past number of processing results of the task and the preference degree according to the preset type of the task.
Here, the higher the evaluation of the task processing, the higher the evaluation, leading to the leveling of the load of each of the process processing devices. In addition, the higher the number of continuous processing of the task for the same job, the higher the evaluation, leading to a reduction in the amount of movement for each job (in fact, an intermediate product corresponding to the job) If continuous processing is performed with the process equipment, it is not necessary to move intermediate products). Further, by increasing the evaluation of the past processing results of the same type of task, the process processing apparatus can be specialized for processing the task of that type. In addition, when there is the type of task that the process processing apparatus mainly wants to process (or does not want to process) by using the preference according to the preset type of the task as an index, Scheduling according to the intention can be performed.

A first result recording means for recording information on the number of past assignments of the task to each of the process processing devices and information on the type of the task in a storage means;
For each of one or more of the process processing devices with a high number of past assignments of the task, specialize in a function that can process only one or more types of the tasks with a high number of past assignments to the process processing device. It is conceivable to include function-specific information output means for outputting function-specific information related to things.
The process processing apparatus having a high number of past task assignments has a sufficiently high operating rate. Therefore, adjusting the operating rate to reduce the operating rate leads to leveling of the load of the entire process processing apparatus.
With the above-described configuration, the function-specific information related to narrowing down (specializing) functions to process only the types of tasks with a high number of past assignments is output for the process processing apparatus with such a high operation rate. . By specializing the function of the process processing apparatus according to the output contents, the load of the entire process processing apparatus can be leveled. Furthermore, since the process processing equipment with a high operation rate can be narrowed down to the functions to be provided (installed tools in processing equipment, raw materials to be held in various process processing equipment, optional equipment, etc.), an efficient configuration with no waste. can do.
As the output of the function specialization information, output of a function specialization message (installation tool removal instruction, etc.) to the user by screen output (display) or printout, etc., or the process processing device automatically removes the installation tool. In the case of a processing device having a function, for example, output of a function-specific command (attached tool removal command) to the process processing device can be considered.

In addition, the processing plan proposal information that allows the task to be assigned from the slave device in response to a single transmission of the processing request information by the processing request means and the number of past assignments of the task to each of the process processing devices. A second result recording means for recording the number of opportunity losses indicating the number of past results for which no reply has been obtained and the type of the task at that time in the storage means;
For each of one or a plurality of the process processing devices with a low number of past assignments of the task, function change information related to changing the task with one or a plurality of types having a high opportunity loss number to a function capable of being processed is output. And a function change information output means.
Since the operation rate of the process processing device having a low number of past assignments of the tasks is low, adjusting the operation rate to increase the operation rate leads to leveling of the load of the entire process processing device.
With the above configuration, the process processing apparatus with such a low operation rate has a high number of opportunity loss cases indicating the past actual number of cases where the reply of the processing plan proposal information to which the task can be assigned was not obtained. The function change information related to changing the task into a processable function (including extension) is output. By changing the function of the process processing device according to the output content, the possibility that the task of the type having the high number of machine losses will be assigned to the process processing device that has had a low operating rate in the past increases. The load on the entire apparatus can be leveled.
The output of the function change information may be similar to the output of the function specialization information.

The present invention can also be understood as a scheduling program executed by each of the information processing device as the master device and the information processing device as the slave device in the scheduling system.
That is, an information processing apparatus that assigns each of the tasks to one of the plurality of process processing devices for a job that is sequentially generated by setting one or a plurality of tasks to be processed by the process processing devices constituting each process of the production line. Executed in each of a master device and a plurality of information processing devices, each of which is a plurality of information processing devices for managing the processing plan of the task in the corresponding process processing device provided communicably with the master device for each process processing device A scheduling program for
Processing function storage means for storing processing function information having information on the types of tasks that can be processed by each of the process processing devices;
Processing plan storage means for storing task processing plan information for the task assigned to each of the process processing devices;
In a situation where there is a job processing plan storage means for storing the process processing device to which processing is assigned to each task in each job and job processing plan information in which information related to the processing plan is set,
A job processing order determining step for determining a processing order of the jobs that are sequentially generated in the master device;
In accordance with the determined processing order of the job and the predetermined processing order of the task in each job, processing request information in which at least information on the type of the task is sequentially set for each task is transmitted from the master device to all the slaves. A processing request step to be sent to the device;
In the slave device, the task requested to be processed can be allocated to the process processing device based on the received processing request information and the processing function information and the task processing plan information related to the corresponding process processing device. A processing plan proposal generating / replying step for generating processing plan proposal information of the task in the case and returning it to the master device;
In the master device, a task assignment step of determining the process processing device to which the task is assigned based on each of the received processing plan proposal information and updating the job processing plan information;
An assignment result notification step of notifying the determined assignment result of the task from the master device to the slave device;
In the slave device, a task processing plan update step for updating the task processing plan information related to the corresponding process processing device based on the notification of the task allocation result;
This is a scheduling program for causing the master device or the slave device to execute the steps.
Similarly, the present invention can also be understood as a scheduling method executed using the master device and the slave device in the scheduling system.
A master device that assigns each of the tasks to one of a plurality of process processing devices for a job that is sequentially generated by setting one or a plurality of tasks to be processed by a process processing device constituting each process of the production line; A plurality of slave devices for managing the processing plan of the task in the corresponding process processing device provided communicably with the master device for each process processing device,
Processing function storage means for storing processing function information having information on the types of tasks that can be processed by each of the process processing devices;
Processing plan storage means for storing task processing plan information for the task assigned to each of the process processing devices;
In a situation where there is a job processing plan storage means for storing the process processing device to which processing is assigned to each task in each job and job processing plan information in which information related to the processing plan is set,
A job processing order determining step for determining a processing order of the jobs that are sequentially generated in the master device;
In accordance with the determined processing order of the job and the predetermined processing order of the task in each job, processing request information in which at least information on the type of the task is sequentially set for each task is transmitted from the master device to all the slaves. A processing request step to be sent to the device;
In the slave device, the task requested to be processed can be allocated to the process processing device based on the received processing request information and the processing function information and the task processing plan information related to the corresponding process processing device. A processing plan proposal generating / replying step for generating processing plan proposal information of the task in the case and returning it to the master device;
In the master device, a task assignment step of determining the process processing device to which the task is assigned based on each of the received processing plan proposal information and updating the job processing plan information;
An assignment result notification step of notifying the determined assignment result of the task from the master device to the slave device;
In the slave device, a task processing plan update step for updating the task processing plan information related to the corresponding process processing device based on the notification of the task allocation result;
A scheduling method comprising the steps of:

According to the present invention, for a job in which one or a plurality of tasks to be processed by a process processing device is set and sequentially generated, the master device that assigns each of the tasks to one of the plurality of process processing devices, and the process processing device A plurality of slave devices for managing the processing plan of the task in the corresponding process processing device provided so as to be communicable with the master device, and determining a processing order of the jobs that are sequentially generated in the master device Then, according to the determined processing order of the job and the predetermined processing order of the task in each job, the processing request information for each task is sequentially transmitted to all the slave devices, and the slave device is responded to the request. The task processing plan draft information is returned from the master device, and the master device Determining the process processing device to which a task is assigned, notifying the slave device of the determined task assignment result, and updating task processing plan information related to the process processing device to which the slave device corresponds based on the notification Thus, load distribution of arithmetic processing can be achieved by each of the master device and the plurality of slave devices. Furthermore, such an autonomous distributed scheduling system can flexibly cope with changes in the number of process processing devices and functional changes.
Further, the slave device cancels part or all of the task processing plan that has already been assigned, notifies the master device of the slave-side plan cancellation information, and the task corresponding to the notification and the task Updating the job processing plan information on the master device side so as to cancel the processing plan of the chained task, and further requesting the slave device side to cancel the processing plan of the chained task to cancel the processing plan. Thus, even if it becomes necessary to cancel a part or all of the task processing plan that has already been assigned due to troubles and inspections such as failures in the process processing apparatus, rescheduling is automatically performed. Therefore, troubles such as setting and changing the constraint conditions do not occur.

In addition, if a priority is set for each of the jobs and the job processing order is determined based on the job generation order and the priority, the job generation order is basically followed. However, when a job with a high priority is generated later, a process for giving priority to the job can be performed.
In this case, the processing request information in which the priority is set is transmitted from the master device to all the slave devices, and the processing plan proposal information for preferentially processing the new task having a higher priority is generated and When the task processing plan information corresponding to the priority processing is updated by replying to the master device, the job processing plan information is updated so as to cancel the task processing plan with the lower priority. For example, canceling the processing plan of the task having a low priority required by preferentially processing the succeeding job (task) having a high priority, canceling the processing plan of the task linked to the canceled task, , Rescheduling associated with the cancellation is automatically performed. Therefore, flexible scheduling corresponding to the generation order and priority of the jobs that are generated sequentially is possible.
In the slave device, an evaluation index of the processing plan is set in the processing plan information of the generated task, and the task is assigned based on the evaluation index set in the processing plan information If a process processing apparatus is determined, in determining which of the plurality of processing plan proposal information is to be adopted, the master apparatus concentrates on the evaluation of grasping the details of each of the process processing apparatuses. Compared to the case, the load of evaluation calculation can be distributed to each of the slave devices.

Further, information related to the number of past task assignments to each of the process processing devices and information about the type of the task is recorded in a storage unit, and each of the one or more process processing devices having a high past number of past task assignments is recorded. If the function-specific information related to specializing in a function that can process only one or a plurality of types of tasks with a high number of past assignments to the process processing device is output, the function-specific information By specializing the function of the process processing apparatus according to the output contents of the above, the load of the entire process processing apparatus can be leveled. Furthermore, since the process processing equipment with a high operation rate can be narrowed down to the functions to be provided (installed tools in processing equipment, raw materials to be held in various process processing equipment, optional equipment, etc.), an efficient configuration with no waste. can do. Furthermore, with this configuration and the following configuration, a system that can simultaneously solve the scheduling problem and the optimization problem of the function of the process processing apparatus can be configured.
In addition, the processing plan proposal information that allows the task to be assigned from the slave device in response to a single transmission of the processing request information by the processing request means and the number of past assignments of the task to each of the process processing devices. One or a plurality of the process processing devices with a low number of past assignments of the tasks recorded in the storage means, and the number of opportunity losses representing the past number of cases for which no reply was obtained and the type of the task at that time If each of them outputs function change information related to changing the task of one or more types having a high opportunity loss number to a function that can be processed, the process processor according to the output contents of the function change information By changing the function, the task processing device that had a low operating rate in the past was assigned the task of the type with the highest number of machine losses. Possibility of Terra is increased, leveling the load of the entire process apparatus can be achieved.

Hereinafter, embodiments of the present invention will be described with reference to the accompanying drawings so that the present invention can be understood. The following embodiment is an example embodying the present invention, and does not limit the technical scope of the present invention.
FIG. 1 is a block diagram showing a schematic configuration of a scheduling system X and a production line Y to which the scheduling system X according to the embodiment of the present invention is applied. FIG. 2 is a flowchart showing a procedure of scheduling processing in the scheduling system X. 3 is a schematic diagram showing how a job processing order is determined in the scheduling system X, FIG. 4 is a schematic diagram showing how task processing plan information is generated in a slave device constituting the scheduling system X, and FIG. 5 is a scheduling system FIG. 6 is a schematic diagram showing a configuration example of job processing plan information in the scheduling system X, and FIG. 7 gives priority to a new task with high priority in the scheduling system X. Task processing plan information when processing FIG. 8 is a flow chart showing the procedure of the function optimization process of the machine tool in the scheduling system X. FIG. 9 is a result of the simulation of the optimization process of the function of the machine tool by the scheduling system X. FIG. 10 is a graph showing the transition of the number of tools attached to each machine tool as a result of the simulation of the optimization process of the function of the machine tool by the scheduling system X. FIG. .

First, the configuration of the scheduling system X according to the embodiment of the present invention and the production line Y to which it is applied will be described with reference to the block diagram of FIG.
The scheduling system X sets each task to a plurality of jobs that are sequentially generated by setting one or a plurality of tasks to be processed by the machine tool 12 (an example of the process processing device) constituting each process of the production line Y. The master device 1 to be assigned to any one of the machine tools 12 and each of the machine tools 12 are provided so as to be communicable with the master device 1 via the network 3 and manage the task processing plan in the corresponding machine tool 12 And a plurality of slave devices 2. The network 3 enables one-to-many communication between the master device 1 and each of the plurality of slave devices 2. Hereinafter, for each of the slave devices 2, the machine tool 12 corresponding to the slave device 2 is referred to as a responsible device 12.
Further, although the configurations of the master device 1 and the slave device 2 are not shown, the CPU that is the arithmetic means, the program executed by the CPU, the storage means such as a hard disk for storing various data, An information processing apparatus such as a personal computer having a network interface that performs communication control via the network 3, display means for displaying information such as a CRT and a liquid crystal panel, and information input means such as a keyboard and a mouse.
Here, the job represents a set of information relating to processing or assembly of products corresponding to orders or production instructions, and is sent to the master device 1 from other information processing devices via the network 3, for example. Alternatively, they are sequentially input from the input means provided in the master device 1 in the order of generation. In the job, a plurality of tasks and their processing order are set in advance.

On the other hand, the job entity (materials, parts, etc.) is accepted into the production line Y and flows through the production line Y. The job entity itself is distinguished from the job (information) in the following. This is called a job.
The production line Y receives the actual job in the receiving yard 11, enters a processing waiting state in the receiving yard 11, and is transported to the machine tool 12 by the transport line 14 according to the processing plan determined by the scheduling system X, When all the scheduled processing (processing by the machine tool 12) is completed, the processing is carried out of the production line Y by the transfer line 14. Each of the machine tools 12 is provided with a buffer 13 for accepting the real job and waiting for processing. It is assumed that each of the machine tools 12 has a restriction that it cannot accept a number of the actual jobs exceeding the capacity of the buffer 13.

Further, the scheduling system X is assigned to processing function information d1 having information on the types of tasks that can be processed by each of the machine tools 12 (process processing devices) and each of the machine tools 12 (process processing devices). A process data storage unit 2a such as a hard disk for storing task processing plan information d2 for the task is provided in each of the slave devices 2 (an example of the processing function storage unit and the processing plan storage unit).
Each of the process data storage units 2a stores only the processing function information d1 and the task processing plan information d2 for the assigned device 12 of the slave device 2 that includes the process data storage unit 2a.
In addition to this, for example, a database relating to the processing function information d1 and the task processing plan information d2 is provided in the storage means of the master device 1 and other information processing devices that can communicate via the network 3 etc. Any other configuration may be used as long as each of the slave devices 2 can access the processing function information d1 and the task processing plan information d2 related to the machine tool 12 managed by the slave device 2.
The processing function information d1 includes task type information representing one or more types of the tasks that can be processed by the responsible device 12, and the task device 12 processing the task for each task type information. And the upper limit number of the real jobs that can be retained in the buffer 13 (hereinafter referred to as the upper limit number of buffers) are set in advance.
The task processing plan information d2 is automatically set as needed by the slave device 2, and for each task scheduled to be processed, the task type information indicating the type and the task to which the task belongs. The job identification information (hereinafter referred to as job ID) and the start time and end time of processing of the task are set.
Further, the scheduling system X stores job processing plan information d3 in which information related to the machine tool 12 (process processing device) to which processing is assigned and its processing plan is set for each task in each job. A job processing data storage unit 1a such as a hard disk is provided in the master device 1 (an example of the job processing plan storage unit).
The job processing data storage unit 1a may be configured to be accessible from the master device 1 in the same manner as the process data storage unit 2a.

(Scheduling process)
Next, the scheduling process procedure of the scheduling system X will be described with reference to the flowchart of FIG. Hereinafter, S1, S2,... Represent identification codes of processing procedures (steps). Each process shown below is implemented by the master device 1 and the slave device 2 executing a scheduling program installed in advance in each of them. Here, the program executed by the master device 1 and the program executed by the slave device 2 are different, but the programs executed by each of the slave devices 2 are the same.

<Step S1: Determine job processing order>
First, in the master device 1, the processing order of the jobs that occur sequentially is determined as a queue for the job (S1, corresponding to the processing of the job processing order determination means).
In this processing, in principle, the processing order is determined in accordance with the job generation order (first-come-first-served basis). However, a priority is set for each of the jobs, and if the priority is set higher than the first job, the job with the higher priority is given priority. , The processing order is changed (an example of processing of the job processing order determination means).
FIG. 3 schematically shows how the job processing order is determined (queue determination) in step S1. Here, the priority d31 indicates that the higher the value, the higher priority should be given.
First, after the job (job 1) with the priority d31 of “1” is generated (arrived) (FIG. 3A), the job with the priority d31 of “1” (job 2) When this occurs, the last-arrived job (job 2) is ordered after the first-arrived job (job 1), and a queue is generated (FIG. 3B).
Next, when the job (job 3) with the priority d31 of “2” is generated, each of the first-arrived jobs (job 1, job 2) is compared with the priority d31, and the priority d31 is high. The job that arrives third (job 3) is placed at the head of the queue, and the processing order of the first two jobs (job 1 and job 2) is lowered (FIG. 3 (c)).

<Step S2: Processing Request>
Next, in accordance with the job processing order determined in step S1 and the task processing order (default processing order) set in advance in each job, a processing request for sequentially requesting each task. Information is transmitted from the master device 1 to all the slave devices 2 via the network 3 by broadcast transmission or the like (S2, an example of processing by the processing request unit).
This process can be thought of as corresponding to “public offering” in an auction-type purchasing system using a network.
Here, the processing request information includes task type information indicating the type of the task requesting processing, the job ID of the job to which the task belongs, the priority of the task, and the earliest of the task. The processing start time T1 is set and transmitted to all the slave devices 2.
Here, the earliest processing start time T1 represents that processing should be started after that time (that processing cannot be started before that time). That is, the task in which a predetermined processing order is set for each job cannot perform the next task processing unless the processing of the previous task is completed. Therefore, when the processing plan for the task to be processed immediately before has already been determined at the earliest processing start time T1, a predetermined setup change time or the like is added to the processing end time or the time. The time is set.
In the present embodiment, the processing request information is first transmitted in a batch for all the tasks set in one job, and thereafter, the processing plan of the task to be processed first is determined. Each time it is done, a processing plan is sent in batches for the remaining tasks that have not yet been determined.
At this time, if no processing plan has been determined for any of the tasks set in the job, the processing request information is added with a predetermined margin time to the current time as the earliest processing start time T1. The time is set. Further, when the processing plans of some of the tasks to be processed first have been determined, the processing end time of the last task whose processing plan has been determined or the time as the earliest processing start time T1 Is set to the time when the predetermined setup change time or the like is added.

<Step S11: Generate / Reply Process Plan>
On the other hand, each of the slave devices 2 is based on the processing request information received from the master device 1, and the processing function information d1 and the task processing plan information d2 related to the responsible device 12 (corresponding process processing device). Thus, it is determined whether or not the task requested to be processed by the responsible device 12 (hereinafter referred to as a requested task) can be assigned. Further, when it is determined that allocation is possible, processing plan (plan) information of the requested task is generated, and this is returned to the master device 1 via the network 3 (S11, processing plan plan generation / reply means) Example).
This process can be thought of as corresponding to “bidding” for “public offering” in an auction-type purchasing system using a network.
The processing plan generation / reply processing in step S11 will be described with reference to the flowchart shown in FIG. 5 and the schematic diagram shown in FIG.
FIG. 4 schematically shows how the slave device 2 generates the processing plan (plan) information.
Here, as shown in FIG. 4A, the task processing plan information d2 in which the start time and the end time are already set for each of the two tasks Ts1 and Ts2 is stored in the process data storage unit 2a. (That is, the tasks Ts1 and Ts2 are assigned). In the example shown in FIG. 4A, the idle time between the tasks Ts1 and Ts2 (time interval from the end time of the task Ts1 to the start time of the task Ts2) is τ.
Further, although not shown in FIG. 4, the task processing plan information d2 includes the priority of each task already assigned (hereinafter referred to as an already assigned task) and the job to which the already assigned task belongs. The job ID and the number of jobs to which the assigned task belongs, that is, the number of substantive jobs that can stay in the buffer 13 (hereinafter referred to as job acceptance number) are set. Therefore, even if there are a plurality of already assigned tasks, if the jobs to which the tasks belong (the job ID) are common, the number of accepted jobs is one. The number of accepted jobs is updated by the slave device 2 every time the task assignment content in the task processing plan information d2 is changed.

The slave device 2 monitors whether or not the processing request information has been received from the master device 1 (S21). By checking whether the task type information is included in the processing mechanism information d1 in order from the task, it is determined whether or not the requested task can be processed by the responsible device 12 ( S22).
Here, if it is determined that the requesting device 12 can process a plurality of request tasks at the head or continuous from the head, the slave device 2 determines whether the request task and each of the already assigned tasks. The priorities are compared (S23), and if the priority of the requested task is equal to or lower than the priorities of all the assigned tasks, the process proceeds to steps S25 to S27, and the assigned tasks are The processing plan (plan) information is generated with priority (while leaving the processing plan of the already assigned task). Hereinafter, a plurality of request tasks at the head or continuous from the head determined to be processable by the responsible device 12 will be referred to as a processable request task set.

(Process plan draft generation: When the priority of the requested task is higher than the priority of the already assigned task)
In this case, the slave device 2 checks whether or not the job acceptance number in the task processing plan information d2 is equal to the buffer upper limit number in the processing function information d1 (S25). It is determined that processing is impossible (processing is impossible), and the process proceeds to step S28.
On the other hand, when the number of accepted jobs is less than the upper limit number of buffers, that is, when it is determined that allocation is possible, the slave device 2 displays the task type information of each of the requested tasks set in the processing request information. The processing function information d1 is searched as a key, and a time tx (hereinafter referred to as a required time tx) required when each of the request tasks included in the processable request task set is processed by the responsible device 12 is acquired (S27). ).
Further, the slave device 2 performs processing plan (plan) information generation processing by the responsible device 12 for one or a plurality of the requested tasks included in the processable request task set (S28).

In this process (S28), at the time after the earliest process start time T1, the process of the required time tx of the first request task in the processable request task set is designated as a vacancy in the task process plan information d2. Check whether it can be inserted in a time zone (a time zone other than the processing time zone of each of the tasks), and the case where it is inserted in an empty time zone that is the earliest possible insertion (close to the earliest processing start time T1) Processing plan (plan) information is generated. On the other hand, if it cannot be inserted, the processing plan (plan) information to which the requested task is allocated is generated immediately after the processing end time of the last allocated task.
More specifically, the process start time of the request task is determined by adding the end time of the process of the task processed immediately before the position where the request task is inserted or just before the end, or a predetermined setup change time, etc. The time obtained by adding the required time tx to the time is taken as a plan for the processing end time of the requested task.
FIG. 4B shows a case where the required time tx of the requested task is shorter than the free time τ between the tasks Ts1 and Ts2 in the task processing plan information d2 shown in FIG. Example).
FIG. 4C shows that the required task tx of the request task is longer than the idle time τ (cannot be inserted), so that the request task is immediately after the end time of the assigned task Ts2 at the end. Represents an example of assigned.
Here, when the first request task in the processable request task set can be inserted, the subsequent request tasks are continuously allocated (inserted) as much as possible. In addition, when assigning the first requested task to the end of the already-assigned task, all of the processable requested task sets are continuously assigned to the end.
As described above, the processing start time plan and processing end time plan for the request task (one or a plurality of request tasks from the top of the processable request task) assigned by the slave device 2 ( Alternatively, when only the processing end time plan) is generated as the processing plan (plan) information, the process proceeds to step S28 described later.
Note that, in the slave device 2, whether or not the responsible device 12 (corresponding process processing device (machine tool 12)) is processing the task according to the task processing plan information d2 is finished. Whether or not the process is being processed and a process that has been completed is excluded from the task process plan information d2, and a plan progress monitoring process is executed. In this process, for example, a signal indicating that the task is being processed and the process has ended is received from the responsible device 12, and the current time is the process start time and process end time of the already assigned task in the task processing plan information. It is conceivable to detect the processing of the task and the end of the processing by passing each of them.
As a result, the processing of the task being processed is not canceled.

(Process plan generation: When the priority of the requested task is higher than the priority of the already assigned task)
On the other hand, when the priority of the requested task is higher than the priority of one or a plurality of already assigned tasks, the processing plan (plan) information giving priority to the requested task (new task) is generated. (An example of processing plan draft generation processing of the processing plan draft generation / reply means).
In this case, the slave device 2 first assumes that the already-assigned task having a lower priority than the requested task in the task processing plan information d2 does not exist (provisionally a task with a lower priority). (S24), the processing of steps S25 to S27 is performed in the same manner as in the above-mentioned “when the priority of the requested task is equal to or lower than the priority of the already assigned task”. Here, as the number of accepted jobs, a value obtained by subtracting the number of jobs corresponding to the already allocated task having a lower priority than the requested task is used. There is no case (Y side of step S25).
As described above, when the processing plan (plan) information of each of the request tasks, that is, the plan information of the processing start time and the processing end time of the request task is generated to the extent possible. Depending on the processing function of the responsible device 12 and the buffer margin status (difference between the upper limit number of buffers and the number of accepted jobs), there is a proposal that the requested task for the same job is assigned to the responsible device 12. Will be generated.
In addition, for each reception of the processing request information, none of the request tasks is assigned (the processing plan (draft) is not generated), or only the top request task or subsequent to it. A plurality of the requested tasks including those are continuously allocated. That is, only the second and subsequent request tasks are not assigned, and the non-sequential request task assignments (for example, only the first and third request tasks are not assigned).

(Calculation of evaluation index)
As described above, when the processing plan (plan) information about the request task that can be allocated is generated, the slave device 2 further adds the processing plan (plan) information to the generated request task. The evaluation index of the processing plan is set (S28, an example of the process of the evaluation index setting means).
This process can be considered as corresponding to the calculation of “bid amount (cost)” in an auction-type purchasing system using a network.
In the processing of this step S28, for example, the margin of the buffer 13 representing the margin of processing of the task in the responsible device 12 (corresponding process processing device), the number of continuous processes of the task for the same job. , The degree of preference according to the number of past processing results of the same type of task as the target of the processing plan and the type of the task (the task type information) preset in the responsible device 12 (of which 1 Alternatively, the evaluation index is set based on a plurality of). As will be described later, in the master device 1, the processing plan (plan) information having the largest evaluation index value (high evaluation) is adopted, and the processing plan is determined.

For example, the margin of the buffer 13 can be obtained from the difference between the upper limit number of buffers and the number of accepted jobs, and it is conceivable to use a function or table in which the evaluation value increases as the value increases. As a result, the requested task is preferentially assigned by the machine tool 12 having a high margin (low operating rate), and the load on each of the machine tools 12 is leveled.
Also, the higher the number of continuous processes of the task for the same job, the higher the evaluation, leading to a reduction in the amount of movement for each of the real jobs (intermediate products, etc.) (continuous processing with one machine tool 12). This eliminates the need to move intermediate products (substance jobs)) and increases production efficiency.
In addition, the higher the evaluation the higher the number of past processing results of the same type of task, the specialized device 12 can be specialized for processing that type of task.
Here, the past number of processing results of the same type of task is calculated every time a signal indicating that the responsible device 12 (machine tool 12) has processed the task is input from the responsible device 12, or the task processing plan information d2 Each time the processing end time of the task set in elapses, the counter variable associated with the task type information of the task is counted up and stored in the process data storage unit 2a.
In addition, by using a preference according to a preset type of the task as an index, the requested task related to the type of task or the job to be mainly processed by the responsible device 12 (each of the machine tools 12). It is possible to increase the evaluation value of the processing plan, and to decrease the evaluation value of the processing plan relating to the type of task that is not desired to be processed. Thereby, scheduling (task allocation) according to the intention of the preference setting person can be performed.
When using these plural indexes, a value obtained by combining the respective index values is used as the evaluation index. For example, taking the sum, average value, weighted average value, etc. can be considered.

As described above, when the processing plan (plan) information (plan of processing start time and processing end time of the request task) for the request task in which the evaluation index is set is generated, The slave device 2 sends a reply to the master device 1 via the network 3 (S29, an example of a reply process of the processing plan draft generation / reply means).
On the other hand, when it is determined in step S22 that the first request task in the processing request information cannot be processed (assigned) by the responsible device 12, or in step S25, the task processing plan information d2 When the number of jobs accepted in is determined to be incapable of processing because it is equal to the upper limit number of buffers in the processing function information d1, a notification that assignment is impossible is returned by the slave device 2 to the master device 1. (S29).
Then, the process proceeds to step S21, where the next processing request information is received and monitored, and the above-described processing is repeated.
The above is an example of the process plan creation / reply process in step S11 (FIG. 2).

<Step S3: Task assignment>
On the other hand, the master device 1 determines the machine tool 12 (process processing device) to which the task (the requested task) is assigned based on each processing plan (plan) information received from each of the slave devices 2. The job processing plan information d3 stored in the job processing data storage unit 1a is updated (S3, an example of processing of the task allocation unit).
This process can be thought of as corresponding to a “successful bid” for “bid” in an auction-type purchasing system using a network.
Here, the one with the highest value of the evaluation index set in each of the processing plan (plan) information from each of the slave devices 2 is adopted, and the adopted processing plan (plan) information (the requested task, its The request task is assigned to the responsible device 12 corresponding to the slave device 2 that is the transmission source in accordance with (plan of processing start time and processing end time).
FIG. 6 is a schematic diagram showing a configuration example of the job processing plan information d3.
As shown in FIG. 6, the job ID d30, the priority d31, and the status d32 are set in the job processing plan information d3 in association with each of the jobs included in the queue. Here, the state d32 is “determined” when the master device 1 has assigned (determined) processing to the machine tool 12 for all the tasks set in the job. "Is set, otherwise" Undecided "is set.
Further, in the job processing plan information d3, the task type information d33, the allocation device d34, the start time d35, and the end time d36 are set for each task.
In the allocation device d34, identification information of the machine tool 12 (process processing device) to which the processing of the task is allocated by the master device 1 is set. In addition, “unallocated” (initial value) is set when the allocation is not yet performed. The identification information of the machine tool 12 is set in the processing plan (plan) information by the slave device 2.
When the processing plan (plan) information to be adopted based on the evaluation index is determined, the identification information of the machine tool 12 set in the processing plan (plan) information, the processing start time, and the processing end time Are set (updated) as decision information in the allocation device d34, the start time d35, and the end time d36, respectively. As a result, the requested task is assigned to the machine tool 12.

<Step S4: Notification of allocation result>
Further, the master device 1 notifies the assignment result of the task determined to be assigned from the master device 1 to the slave device 2 (an example of processing of the assignment result notification means). This notification is sent to all the slave devices 2 by broadcast transmission, and on the slave device 2 side, it is determined whether or not the notification is addressed to itself based on the identification information of the machine tool 12 included in the notification. Alternatively, it may be possible to designate and notify the slave device 2 that is a transmission source of the adopted processing plan (plan) information.

<Step S12: Task processing plan update (first slave-side plan cancellation), S13: Slave-side plan cancellation notification>
On the other hand, the slave device 2 updates the task processing plan information d2 related to the responsible device 12 (corresponding process processing device) based on the notification of the assignment result of the task received from the master device 1 ( S12, an example of processing of the task processing plan update means).
In this case, if the notification of the allocation result of the task relates to the requested task having the priority of the already allocated task or lower, the requested task is stored in the task processing plan information d2 according to the notification of the allocation result. The task identification information, the priority, the start time d35 and the end time d36 are added (inserted or added at the end between the already assigned tasks). Further, if the added request task belongs to the job different from the job to which the already assigned task belongs, the job acceptance number is incremented by one.
On the other hand, when the notification of the allocation result of the task is related to the request task having the higher priority than the already allocated task, the slave device 2 has the higher priority according to the notification of the allocation result. The task processing plan information is updated so as to preferentially process the new requested task (S12, an example of processing of the task processing plan update means). In this case, the task processing plan information d2 is updated so as to cancel the processing plan of the already assigned task (part or all) having a low priority in the responsible device 12 (the corresponding process processing device) (the above-mentioned task processing plan information d2). Example of processing of first slave side plan canceling means).
Furthermore, when a cancellation occurs in the process of step S12, slave side plan cancellation information in which information related to the already allocated task whose processing plan is canceled is set by the slave device 2 to the master device. (S13, an example of processing of the slave-side planned cancellation notification means).

<Step S5: Master side planned cancellation, S6: Cancel request>
On the other hand, in the master device 1, based on the slave-side plan cancellation information notified from the slave device 2 side, the task whose processing plan has been canceled on the slave device 2 side and the tasks linked to this task The job processing plan information (FIG. 6) is updated so as to cancel the processing plan (S5, an example of processing of the master side plan cancellation means).
More specifically, in the job processing plan information shown in FIG. 6, the task notified by the slave side plan cancellation information and the task linked to the task, that is, in the same job as that task. For all the tasks after the processing order (based on the job processing order determined in step S1 and the predetermined task processing order for each job), the setting of the “allocation device” d34 is “unallocated”. (S5).
Furthermore, the master device 1 transmits cancel request information for requesting cancellation of the processing schedule of the chained tasks from the master device 1 to the slave device 2 (S6, an example of processing of the cancel request means). .
In the cancellation request information, information on all the tasks whose plan has been canceled ("unassigned") on the master device 1 side (the job ID and the task type information of the job to which the task belongs) is set. Is done.

<Step S14: Second slave plan cancellation>
On the other hand, in the slave device 2, the assigned device 12 (the corresponding process processing device is arranged so that the designated processing plan for the task is canceled based on the cancellation request information received from the master device 1 side. The task processing plan information d2 related to () is updated (an example of processing of the second slave-side plan canceling means).
As described above, the task whose processing plan is canceled on the master device 1 side is added to the queue again in step S1.
The processing of steps S1 to S5 and steps S11 to S14 described above is repeated until there is no queue for the task, that is, until there is no “unallocated” allocation device d34. At that time, in response to the transmission of the processing request information (S2), the task processing plan (plan) information indicating that the requested task can be allocated from any of the slave devices 2 is not returned. The master device 1 waits for the generation of a new job or the elapse of the earliest end time d36 among the tasks to which a processing plan has already been assigned, and repeats the processing from step S1 again. As a result, scheduling is performed for the jobs that occur sequentially.
In the above-described processing, the job processing plan information d3 is updated so as to cancel the processing plan of the task having the low priority. However, the function of canceling the processing plan of any previously allocated task is provided. It is more preferable if it is added to the slave device 2.
For example, in the slave device 2, an arbitrary means of the assigned task processing plan (the already assigned task) of the assigned device 12 (corresponding process processing device) by the input means (keyboard, mouse, etc.) provided in the device. An input function for canceling part or all of the information is added, and a function for updating the job processing plan information d3 is added so as to cancel the processing plan for the already assigned task specified by the input function (the first processing unit). Example of processing of slave side planned canceling means).
As a result, even if it is necessary to cancel a part or all of the task processing plan that has already been assigned due to trouble or inspection of the machine tool 12 or the like, rescheduling is automatically performed. .

FIG. 7 schematically shows how the task processing plan information is updated to preferentially process a new task having a high priority d31 when the task having a high priority d31 occurs later. It is a representation.
At a certain point, each of the three machine tools 12 (represented as “device 1”, “device 2”, and “device 3” in the figure) has the task processing plan information as shown in FIG. Suppose that it was set. The assigned tasks in “device 1” to “device 3” are “task 1” in “job 1” and “task 2” in “job 2”, “task 2” in “job 1”, and “job 2”, respectively. "Task 1" of "" and "task 3" of "job 1".
In this state, the notification of the allocation result for allocating “task 1” of “job 3” having higher priority d31 than any of the already allocated tasks to “device 2” is made to “device 2”. To do.
In this case, as shown in FIG. 7B, the processing plan of “task 2” of the already assigned task “job 1” is canceled and a new “task 1” of “job 3” is assigned. Then, the information of “task 2” of the canceled “job 1” is notified to the master device 1 as the slave-side planned cancellation information.
In response to this notification, the master device 1 notifies the slave device 2 of the cancel request information for requesting cancellation of “task 3” of “job 1” linked to “task 2” of “job 1”. As shown in FIG. 7C, the processing plan for “task 3” of “job 1” already assigned to “device 3” is canceled. The “task 2” and “task 3” of “job 1” whose processing plan has been canceled in this way are added to the queue again and scheduled.

(Functional optimization of machine tools)
Next, the function optimization process (function specialization process and function change process) of the machine tool 12 executed by the master device 1 will be described.
The machine tool 12 having a high number of past task assignments has a sufficiently high operation rate. Therefore, adjusting the operation rate to reduce the operation rate leads to leveling of the load on the entire machine tool 12. Further, if the function of the machine tool 12 is specialized to a function corresponding to one or a few kinds of the tasks, the number of tools and the like that the machine tool 12 should have can be reduced, which is efficient.
On the other hand, the machine tool 12 with a low number of past task assignments has a low operation rate, so adjusting the operation rate to increase the operation rate leads to leveling of the load on the entire machine tool 12.
Hereinafter, the function optimization process of the machine tool 12 for leveling the load and improving the efficiency of the apparatus will be described using the flowchart shown in FIG. This process is performed by the master device 1 executing a program installed in advance.
Here, in the job processing data storage unit 1a of the master device 1, all of the task type information and necessary function information regarding functions necessary for processing of each task (the tool to be included in the machine tool 12). It is assumed that task / function correspondence information associated with type information etc. is stored.

First, in parallel with the scheduling process shown in FIG. 2, the master device 1 stores past tasks to the machine tool 12 (an example of the process processing device) based on the job processing plan information d3. Record the actual number of assignments and the task type information d33 of each of the tasks, the number of opportunity losses described later, and the task type information at that time in the job processing data storage unit 1a so as to be accumulated for a certain past period. (S31, an example of processing of the first and second performance recording means).
Here, the number of opportunity loss cases refers to the processing plan (draft) that allows the slave device 2 to assign the task to one transmission of the processing request information by the processing request processing (S2 in FIG. 2). ) Represents the number of past records for which no information was returned.
Then, for example, for each predetermined period, the master device 1 may use the machine tool for each of one or a plurality (small number) of the machine tools 12 (an example of the process processing device) with a high number of past assignments of the task. The process of outputting the function specialization information related to specializing in the function that can process only one or a plurality of (small number) types of tasks with a high past allocation record number for 12 is executed (S32 to S34, the function specialization) An example of the process of the conversion information output means).
For example, in this process, first, the master device 1 selects the machine tool 12 with the highest number of actual task assignments (cumulative number) in the past certain period based on the accumulated performance information ( S32).
Next, the master device 1 selects the type of task (the task type information d33) having the highest number of assignments in the past certain period for the selected machine tool 12 based on the accumulated result information. (S33).
Further, the master device 1 refers to the task / function correspondence information by using the task type information d33 of the selected type of task as a key, extracts the corresponding necessary function information, and selects the information and the information in step S32 The function-specific information in association with the information representing the machine tool 12 is output (S34).

Similarly, for example, for each predetermined period, the master device 1 is configured for each of one or a plurality (small number) of the machine tools 12 (an example of the process processing device) with a low number of past assignments of the task. A process of outputting function change information related to changing the task of one or a plurality (small number) of types having a high opportunity loss number to a processable function is executed (S35 to S37, the process of the function change information output means) Example).
For example, in this process, first, the master device 1 selects the machine tool 12 with the lowest number of task assignments (cumulative number) in the past fixed period based on the accumulated record information ( S35).
Next, the master device 1 selects the type of task (the task type information) having the highest opportunity loss number based on the accumulated performance information (S36).
Further, the master device 1 refers to the task / function correspondence information using the task type information d33 of the selected type of task as a key, extracts the necessary function information corresponding thereto, and selects the information and the information in step S35. The information representing the machine tool 12 is output in association with the information (S37).

With the processing shown in FIG. 8, the function-specific information for narrowing down (specializing) the functions of the machine tool 12 having a high operation rate to a function that can process only the types of tasks with a high number of past assignments is output. (S34). By specializing the function (tool or the like) of the machine tool 12 according to the output contents, the load of the entire machine tool 12 can be leveled. Furthermore, since the functions to be provided for the machine tool 12 with a high operating rate can be narrowed down, an efficient configuration without waste can be achieved.
In addition, for the machine tool 12 with a low operation rate, the function change information for changing (including expansion) the function of the type of task that could not be assigned immediately to a function that can be processed is output ( S37). By changing / extending the function of the machine tool 12 according to the output content, the possibility that the task of the type having a high number of machine losses is assigned to the machine tool 12 having a low operation rate is increased. The load on the entire 12 can be leveled.
As the output of the function specialization information or the function change information, a function specialization message (attached tool) to the user by screen output (display) or printout of display means provided in the master device 1 or the slave device 2 is used. It may be possible to output a removal instruction, an installation instruction, etc.). In addition, when the machine tool 12 has an automatic removal / attachment function of a mounted tool, it may be possible to output a mounted tool removal command or a mounting command to the machine tool 12.

Next, a simulation result of the function optimization process of the machine tool 12 will be described.
The simulation condition is that the number p of the machine tools 12 (that is, the number of slave devices 2) is 5, the number of types of tasks is 10, and the number of types of tools (number of types of functions) of each of the machine tools 12 is. It is assumed that a dedicated tool necessary for the processing corresponds to each task on a one-to-one basis. In the initial state, two kinds of tools are mounted on each of the machine tools 12.
Further, the number of types of the job is 5, and a maximum of three tasks are set for each job. Then, scheduling was performed by sequentially generating 200 jobs. However, if there are three or more jobs that stay (the job whose status d32 in the job processing plan information is “undecided”) in the queue, the number of the staying jobs becomes two or less. New jobs are not generated.
Further, it is assumed that an express job with the priority set higher than the normal priority is generated once for the occurrence of the seven jobs, and once for the occurrence of the job for five times. The function optimization process is executed, and the tool to be mounted on the machine tool 12 is added or changed according to the output result.
9 and 10 are graphs showing simulation results of the function optimization process under the above conditions.
Here, FIG. 9 is a (vertical axis) graph showing the change of the task sharing rate of each of the five machine tools (M1 to M5), and the horizontal axis represents the number of occurrences of the job. Here, the share ratio is represented by the ratio of the number of tasks processed by each of the machine tools to the total number of tasks processed by all the machine tools (M1 to M5). Accordingly, since the five machine tools are used, the case where the share of each of the machine tools is 20% (= 100% / 5) is the state where the leveling of processing is most achieved. I can say that. Moreover, it can be said that the said machine tool with the said high share rate has a relatively high operation rate.
As can be seen from the graph of FIG. 9, there was a large variation in the share at the beginning of the process, but the share optimization (equalization of the operation rate) can be achieved rapidly by the function optimization process. I understand that.
On the other hand, FIG. 10 is a graph showing the transition of the number of tools (vertical axis) mounted on each of the five machine tools (M1 to M5), and the horizontal axis represents the number of occurrences of the job.
As shown in FIGS. 9 and 10, it can be seen that by increasing the number of tools in each of the machine tools, the sharing rate is equalized (operating rate is leveled). At that time, the machine tool (M1) has a relatively small number of tools (three) and is specialized in function.
On the other hand, for the machine tool (M2), as the number of jobs stays, that is, as the number of lost opportunities increases, the number of tools must be increased to level the operation rate (5 to 6 pieces). ) After the situation, it can be seen that the number of tools has finally converged to the same number of 5 as other machine tools (M3 to M5).
As described above, according to the present invention, it is possible to optimize the function for leveling the operation rate of each of the machine tools together with the scheduling of the job.

  The present invention is applicable to a process scheduling system in a production line.

The block diagram showing the schematic structure of the scheduling system X which concerns on embodiment of this invention, and the manufacturing line Y to which it is applied. FIG. 4 is a flowchart showing a procedure of scheduling processing in the scheduling system X. FIG. 3 is a schematic diagram showing how a job processing order is determined in a scheduling system X. FIG. 4 is a schematic diagram showing a state of generation of task processing plan proposal information in a slave device constituting the scheduling system X. The flowchart showing the procedure of the process plan production | generation / reply process in the scheduling system X. 3 is a schematic diagram illustrating a configuration example of job processing plan information in the scheduling system X. FIG. The schematic diagram showing the mode of the update of task processing plan information at the time of preferentially processing a new task with high priority in the scheduling system X. The flowchart showing the procedure of the function optimization process of the machine tool in the scheduling system X. The graph showing the transition of the task processing share of each machine tool, which is the result of the simulation of the optimization process of the function of the machine tool by the scheduling system X. The graph showing the transition of the number of the installation tools of each machine tool which is the result of the simulation of the optimization process of the function of the machine tool by the scheduling system X.

Explanation of symbols

X ... scheduling system Y ... production line 1 ... master device 1a ... job processing data storage unit 2 ... slave device 2a ... process data storage unit 3 ... network 11 ... receiving yard 12 ... machine tool (process processing device)
13 ... Buffer 14 ... Conveyance line d1 ... Processing function information d2 ... Task processing plan information d3 ... Job processing plan information S1, S2, ... Processing procedure (step)

Claims (10)

A master device that assigns each of the tasks to one of a plurality of process processing devices for a job in which one or a plurality of tasks to be processed by a process processing device constituting each process of the production line is set and sequentially generated; A scheduling system comprising: a plurality of slave devices for managing a processing plan of the task in the corresponding process processing device provided communicably with the master device for each processing device;
Processing function storage means for storing processing function information having information on the types of tasks that can be processed by each of the process processing devices;
Processing plan storage means for storing task processing plan information for the task assigned to each of the process processing devices;
Job processing plan storage means for storing job processing plan information in which information related to the process processing device to which processing for each task in each job is assigned and its processing plan is set;
Job processing order determining means for determining a processing order of the jobs that are sequentially generated in the master device;
In accordance with the determined processing order of the jobs and the predetermined processing order of the tasks in each of the jobs, processing request information in which at least information on the type of the task is sequentially set for each of the tasks from the master device to all the slaves Processing request means to be transmitted to the device;
The slave device can assign the task requested to be processed to the process processing device based on the received processing request information and the processing function information and the task processing plan information regarding the corresponding process processing device. Processing plan draft generation / reply means for generating processing plan draft information for the task and returning it to the master device;
In the master device, task assignment means for determining the process processing device to which the task is assigned based on each of the received processing plan proposal information and updating the job processing plan information;
An assignment result notifying means for notifying the determined assignment result of the task from the master device to the slave device;
In the slave device, task processing plan update means for updating the task processing plan information related to the corresponding process processing device based on the notification of the task allocation result;
A scheduling system comprising: In the slave device, first slave-side plan cancellation means for updating the task processing plan information so as to cancel a part or all of the processing plan of the task already assigned to the corresponding process processing device;
Slave-side plan cancellation notification means for notifying the master device of slave-side plan cancellation information related to the task whose processing plan has been canceled by the first slave-side plan cancellation means;
In the master device, the job processing plan information is canceled so as to cancel the task whose processing plan has been canceled on the slave device side based on the notified slave side plan cancellation information and the processing plan of the task linked to the task. Master side plan cancellation means to be updated,
Cancellation request means for transmitting cancellation request information for requesting cancellation of the processing plans of the chained tasks from the master device to the slave device;
In the slave device, a second slave-side plan cancellation means for updating the task processing plan information related to the corresponding process processing device based on the received cancellation request information;
The scheduling system according to claim 1, further comprising: A priority is set for each of the jobs,
The scheduling system according to claim 1, wherein the job processing order determination unit determines the processing order of the jobs based on the generation order of the jobs and the priority. The processing request means transmits the processing request information set with the priority from the master device to all the slave devices,
The processing plan draft generation / reply means generates the processing plan draft information for preferentially processing the new task with a high priority, and returns the master plan to the master device.
The first slave-side plan cancellation unit cancels the task processing plan with the lower priority when the task processing plan update unit is updated by the task processing plan update unit corresponding to the priority processing. The scheduling system according to claim 3, wherein the job processing plan information is updated. The slave device includes an evaluation index setting unit that sets an evaluation index of the processing plan plan in the processing plan plan information of the task generated by the processing plan plan generation / reply unit,
The scheduling system according to claim 1, wherein the task allocation unit determines the process processing apparatus to which the task is allocated based on the evaluation index set in the processing plan proposal information.   The evaluation index setting means, in the corresponding process processing device, the processing margin of the task, the number of continuous processing of the task for the same job, the task of the same type as the task to be processed The scheduling system according to claim 5, wherein the evaluation index is set based on one or a plurality of past processing results and a preset preference according to the task type. A first result recording means for recording information on the number of past assignments of the task to each of the process processing devices and the type of the task in a storage means;
For each of one or more of the process processing devices with a high number of past assignments of the task, specialize in a function that can process only one or a plurality of types of tasks with a high number of past assignments to the process processing device. Function-specific information output means for outputting function-specific information related to things;
The scheduling system according to claim 1, further comprising: Reply of the processing plan proposal information enabling the assignment of the task from the slave device in response to a single transmission of the processing request information by the processing request means and the number of past assignments of the task to each of the process processing devices A second result recording means for recording the number of opportunity losses indicating the number of past achievements for which No is obtained and the type of the task at that time in the storage means;
For each of one or a plurality of the process processing devices with a low number of past assignments of the task, function change information related to changing the task with one or a plurality of types having a high opportunity loss number to a function capable of being processed is output. Function change information output means,
The scheduling system according to claim 1, further comprising: A master that is an information processing apparatus that assigns each of the tasks to one of a plurality of process processing devices for a job in which one or a plurality of tasks to be processed by a process processing device constituting each process of the production line is set and sequentially generated To execute each of a device and a slave device that is a plurality of information processing devices that are provided so as to be communicable with the master device for each of the process processing devices and manage a processing plan of the task in the corresponding process processing device Scheduling program
Processing function storage means for storing processing function information having information on the types of tasks that can be processed by each of the process processing devices;
Processing plan storage means for storing task processing plan information for the task assigned to each of the process processing devices;
In a situation where there is a job processing plan storage means for storing job processing plan information in which information about the process processing device to which processing is assigned for each of the tasks in each job and its processing plan is set,
A job processing order determination step for determining a processing order of the jobs that are sequentially generated in the master device;
In accordance with the determined processing order of the jobs and the predetermined processing order of the tasks in each of the jobs, processing request information in which at least information on the type of the task is sequentially set for each of the tasks from the master device to all the slaves A processing request step to be sent to the device;
The slave device can assign the task requested to be processed to the process processing device based on the received processing request information and the processing function information and the task processing plan information regarding the corresponding process processing device. A processing plan draft generating / replying step for generating processing plan draft information of the task in the case and returning it to the master device;
In the master device, a task assignment step for determining the process processing device to which the task is assigned based on each received processing plan proposal information and updating the job processing plan information;
An assignment result notification step of notifying the determined assignment result of the task from the master device to the slave device;
In the slave device, a task processing plan update step for updating the task processing plan information related to the corresponding process processing device based on the notification of the task allocation result;
The scheduling program for making the said master apparatus or the said slave apparatus perform each step of these. A master device that assigns each of the tasks to one of a plurality of process processing devices for a job in which one or a plurality of tasks to be processed by a process processing device constituting each process of the production line is set and sequentially generated; A plurality of slave devices for managing a processing plan of the task in the corresponding process processing device provided communicably with the master device for each processing device, and a scheduling method executed using
Processing function storage means for storing processing function information having information on the types of tasks that can be processed by each of the process processing devices;
Processing plan storage means for storing task processing plan information for the task assigned to each of the process processing devices;
In a situation where there is a job processing plan storage means for storing job processing plan information in which information about the process processing device to which processing is assigned for each of the tasks in each job and its processing plan is set,
A job processing order determination step for determining a processing order of the jobs that are sequentially generated in the master device;
In accordance with the determined processing order of the jobs and the predetermined processing order of the tasks in each of the jobs, processing request information in which at least information on the type of the task is sequentially set for each of the tasks from the master device to all the slaves A processing request step to be sent to the device;
The slave device can assign the task requested to be processed to the process processing device based on the received processing request information and the processing function information and the task processing plan information regarding the corresponding process processing device. A processing plan draft generating / replying step for generating processing plan draft information of the task in the case and returning it to the master device;
In the master device, a task assignment step for determining the process processing device to which the task is assigned based on each received processing plan proposal information and updating the job processing plan information;
An assignment result notification step of notifying the determined assignment result of the task from the master device to the slave device;
In the slave device, a task processing plan update step for updating the task processing plan information related to the corresponding process processing device based on the notification of the task allocation result;
A scheduling method comprising the steps of:

Images