JP5666377B2 - Production schedule creation device and production schedule creation method - Google Patents

Description

  The present invention relates to a production scheduling technique when manufacturing a variety of products, and more particularly to a technique for performing better scheduling at high speed.

  When producing multiple types of products, it is necessary to plan a production schedule for each facility that satisfies the evaluation items such as adherence to delivery date, maximization of productivity, and timely supply to downstream processes. .

  However, even when planning a single facility for a single day, scheduling targets range from tens to hundreds, so it is difficult to formulate an optimal schedule, and even if it can be planned, it takes time. It becomes. In addition, when there are a plurality of facilities that should be planned, such as a production schedule for the whole factory, it is very difficult to achieve both high quality planning and high-speed planning.

  In addition, for example, even if a schedule is set for each process that can be optimally scheduled in a relatively short time instead of the schedule for the entire factory, if the connection with other processes is poor, the result is that the entire factory Productivity will deteriorate.

  In view of this, a method has been proposed in which a schedule for each process is planned in consideration of the schedule for the entire process (see Patent Document 1, etc.). This method creates a schedule for all processes by using a lot in which products are grouped as a processing unit. When creating a schedule for a certain process, the process schedules before and after the process are referred to. . This lot is determined from the upper and lower limit values of the lot size, which are parameters set in advance by the user, order information, and the like.

JP 2010-97506 A

  However, in such a scheduling method, the calculation time increases as the number of lots to be scheduled in each process increases. In order to shorten the calculation time, it is conceivable to reduce the number of scheduling targets. However, in order to reduce the number of scheduling targets, it is necessary for the user to reset parameters such as the upper and lower limit values of the lot size, and to perform scheduling again, which takes time and effort.

  Therefore, an object of the present invention is to create a schedule for producing a plurality of types of products at a higher speed than in the past and to create a more optimal schedule.

In one aspect according to the present invention, a production schedule creation device for creating a schedule of equipment for sequentially processing a plurality of types of products from an initial schedule created in advance , wherein a first group consisting of a plurality of products is created. A creation unit that creates a candidate schedule as a processing unit; a first unit that groups the products to be processed in the initial schedule using a first condition to generate the first group, and causes the creation unit to create a candidate schedule; and processing the first group were a processing unit when said producing means candidate schedules further grouped by using the second condition to generate a second group, the said second group the as a group, and control means for performing a second process of repeating an thereby creating a candidate schedule to said creation means, the initial scan And evaluating means for evaluating each module and candidate schedules the producing means, out of the initial schedule and the candidate schedule said evaluation means to evaluate, and determining means for determining a highest evaluation schedule as the best schedule It is characterized by providing.

And the production schedule creation method which concerns on the other one aspect | mode of this invention is the production schedule creation used with the production schedule creation apparatus which creates the schedule of the equipment which processes several types of products in order from the initial schedule created beforehand. A method of creating a candidate schedule using a first group of a plurality of products as a processing unit, and grouping products to be processed in the initial schedule using a first condition. A first process for generating a group and creating a candidate schedule in the creation step and a first group as a processing unit when the candidate schedule is created in the creation step are further grouped using a second condition. generating a second group, the second group as the first group, the candidate schedule in the creation step And a control step of performing a second process of repeating the to create a le, the initial scheduling and the evaluation step of evaluating each candidate schedule created by the creating step, the initial schedule and the candidate was evaluated in the evaluation step A determining step of determining a schedule with the highest evaluation as an optimal schedule among the schedules.

  According to the production schedule creation device and production schedule creation method configured as described above, the scheduling targets are repeatedly grouped, so that the scheduling targets are gradually reduced, that is, the search range when searching for a plan is gradually narrowed. The possibility of creating an optimal schedule efficiently and in a short time increases. The multiple types of products refer to multiple products that include at least different types of products.

In the production schedule creation device described above, the control unit performs the second process until a predetermined end condition is satisfied, and the predetermined end condition is an evaluation of the candidate schedule created by the creation unit, it is preferable number of consecutive times not exceeding the highest evaluation among the evaluation of the evaluation of candidate schedules created before that and the initial schedule is to exceed the predetermined number of times.

In the above production schedule creation device, when the predetermined end condition is satisfied, the control unit uses a new predetermined condition as the second condition instead of the second condition currently used. It is preferable that a second group is generated by grouping, and the creation unit is made to create a candidate schedule with the second group as the first group.
According to this configuration, the conditions for grouping can be changed, so that there is a high possibility that a better production schedule can be created.
Further , according to this configuration, when a better plan cannot be searched with the current grouping condition, the grouping condition can be changed and the grouping can be performed again to continue the search. The possibility that it can be created is increased.

Further, in the above production schedule creation device, in order to process the product, it is necessary to change the setup of the equipment, and the first and second conditions are when the products belonging to the same group are processed continuously. When it is necessary to perform the setup change, it is preferable to include a condition that these products are not in the same group.

  According to this configuration, since a group in which no setup change occurs is scheduled as a processing unit, there is a high possibility that a schedule with a reduced number of setup changes can be created.

In the production schedule creation apparatus described above, it is preferable that the evaluation unit increases the evaluation of the initial schedule or the candidate schedule as the number of times of setup change included in the initial schedule or the candidate schedule is smaller.

  According to this configuration, since a high evaluation is made on a schedule with a small number of setup changes, there is a high possibility that a highly productive schedule is determined as an optimal schedule.

In the production schedule creation device described above, the production schedule creation device further includes storage means for storing a large schedule that takes into account equipment other than the equipment, and the creation means is based on the large schedule. a candidate schedules, from the time of performing the product of each process in the atmospheric scheduled dates, it is preferable to create a candidate schedules for performing processing within a predetermined time period.

  According to this configuration, scheduling is performed based on a large schedule that takes into account other equipment. For example, if the large schedule is a schedule that takes into consideration the productivity and delivery date of the entire factory, the productivity and delivery date of the entire factory It is possible to create a schedule that takes into account the productivity of individual facilities while satisfying the above. In addition, since the search range of the schedule is narrowed on condition that the process is performed within a predetermined period from the process in the large schedule, the schedule can be created at a higher speed.

  The production schedule production apparatus according to the present invention can create a schedule for producing a plurality of types of products at a higher speed and can create a more optimal schedule.

It is a block diagram which shows the functional structure of the production schedule preparation apparatus in embodiment. It is a flowchart for demonstrating the schedule planning process of the production schedule preparation apparatus in embodiment. It is a figure which shows the example of a structure and content of the order information table in the production schedule planning apparatus shown in FIG. It is a figure which shows the example of a big schedule plan in the production schedule preparation apparatus shown in FIG. It is a figure which shows the example of the initial plan in the production schedule planning apparatus shown in FIG. It is a figure which shows the example of the candidate plan in the production schedule planning apparatus shown in FIG. It is a figure which shows the example of the candidate plan in the production schedule planning apparatus shown in FIG.

<Embodiment>
<Overview>
The production schedule of the whole factory that manufactures multi-product and multi-process products on multiple production lines is expected to maximize the productivity of the whole factory as well as complying with the delivery date. It is desirable that such a schedule can be created quickly. As such a production line, for example, a production line of a material factory such as steel, copper plate, aluminum plate or the like can be cited.

  The production schedule of the whole factory that manufactures such multi-product and multi-process products on a plurality of production lines can be said to be a collection of production schedules for facilities that manufacture multi-product products on a single production line. Therefore, increasing the productivity of this single production line while considering the overall schedule leads to an increase in the productivity of the entire factory. The production line performs setup changes to produce multiple types of products. For example, replacement of a rolling roll in the rolling process. Reducing the number of times of setup change is one major factor that increases productivity.

  Further, in order to create a schedule at high speed, it is necessary to reduce the amount of calculation. To that end, it is conceivable to reduce the number of scheduling targets.

  The production schedule creation apparatus according to the embodiment groups products in consideration of setup change, creates a schedule with the group as a processing unit, and further repeats the creation of a schedule by grouping the group. The optimal schedule is searched while gradually reducing the number of subjects. Therefore, it is possible to create a schedule in which the productivity of the entire factory is as high as possible in a short time.

  Hereinafter, an embodiment according to the present invention will be described with reference to the drawings.

<Configuration>
FIG. 1 is a diagram showing a functional block configuration of the production schedule creation apparatus 1000.

  The production schedule creation apparatus 1000 includes a production schedule creation control unit 1001, an input unit 1002, and an output unit 1003.

  The input unit 1002 is a device that inputs various commands such as a command for starting a program for creating a production schedule and various data necessary for creating a schedule to the production schedule creating apparatus 1000. For example, a keyboard or a mouse.

  The output unit 1003 is a device that outputs (presents) the command and data input from the input unit 1002 and the schedule created by the production schedule creation device 1000. For example, a display device such as a CRT display, LCD, organic EL display, and plasma display, or a printing device such as a printer.

  The production schedule creation control unit 1001 includes, for example, a microprocessor and its peripheral circuits. Functionally, the production schedule creation control unit 1001 is functionally configured as an initial plan creation unit 1100, a plan search unit 1200, an optimum plan determination unit 1300, a plan evaluation unit 1400, a large size. A schedule plan storage unit 1500, an operation condition storage unit 1600, a grouping condition storage unit 1700, and an optimum plan storage unit 1800 are provided, and the input unit 1002 and the output unit 1003 are controlled according to the function according to the control program. . An arrow in the production schedule planning processing unit 1001 represents a data flow between functional blocks.

  The initial plan creation unit 1100 has a function of reading a large schedule plan that is a production plan of several days to several months stored in the large schedule plan storage unit 1500 and creating an initial plan that is an initial production schedule. . The created initial plan is passed to the plan search unit 1200 and the optimum plan determination unit 1300.

  The initial plan creation unit 1100 reads various information stored in the operation condition storage unit 1600 when creating a schedule, and creates an initial plan based on the read information. The production method is any one of the conventional methods. However, since the initial plan is preferably consistent with the read large schedule plan and can be planned at high speed, the production schedule creation device 1000 For each order, a rule is adopted such that the orders are arranged without changing the date scheduled for processing in the large schedule and without taking into account the time required for the setup change.

  The plan search unit 1200 controls the search process based on the initial plan received from the initial plan creation unit 1100 and creates a candidate plan. Specifically, products are grouped, and a candidate plan is created using the group as a processing unit. The created candidate plan is passed to the optimum plan determination unit 1300. At the time of grouping, the plan search unit 1200 performs grouping according to the conditions read from the grouping condition storage unit 1700. This grouping method will be described later in <Grouping>. In addition, the method for creating the plan is created by any of the conventional methods. However, scheduling will be performed with the aim of observing the delivery date and maximizing the productivity of the entire factory.

  The optimal plan determination unit 1300 has a function of determining a plan with the highest evaluation based on the evaluation of the plan calculated by the plan evaluation unit 1400, which will be described later, and storing it in the optimal plan storage unit 1800. Specifically, the plan evaluation unit 1400 is requested to evaluate the plan received from the initial plan creation unit 110 or the plan search unit 1200, and the evaluation received from the plan evaluation unit 1400 is stored in the optimum plan storage unit 1800. If it is higher than the evaluation, the plan stored in the optimum plan determination unit 1300 is updated with the received evaluation plan.

  The plan evaluation unit 1400 has a function of evaluating the plan received from the optimal plan determination unit 1300 and notifying the optimal plan determination unit 1300 of the evaluation result. This evaluation method will be described later in <Evaluation Method>.

  The large schedule plan storage unit 1500 has a function of previously storing a large schedule plan that is a production plan of several days to several months. This large schedule is a schedule scheduled in consideration of productivity, delivery date, etc. of the whole factory.

  The operation condition storage unit 1600 has a function of storing in advance information necessary for creating an initial plan in the initial plan creation unit 1100 and creating a candidate plan in the plan search unit 1200. For example, order information, equipment capacity information, production line information, and the like. Also, the conditions for creating (searching) the plan are stored. In the embodiment, a condition is set such that the schedule can be searched within a range that does not greatly deviate from the rough processing time determined by the large schedule. For example, the user sets and stores a constraint that each order does not deviate more than one day (this value can be arbitrarily set) from the scheduled processing date in the large schedule. By including this restriction, the search range is narrowed, and the schedule can be searched more efficiently. Furthermore, it is possible to create a schedule that takes into account the productivity of individual equipment while satisfying the things considered when creating a large schedule (such as productivity of the entire factory and delivery date).

  The grouping condition storage unit 1700 has a function of storing grouping conditions when products are grouped.

  The optimum plan storage unit 1800 has a function of storing an optimum plan.

  These storage units such as the large-scale schedule storage unit 1500 are, for example, volatile storage elements such as a RAM (Random Access Memory) serving as a so-called working memory of the production schedule creation device 1000, ROM (Read It is composed of a non-volatile storage element such as an only memory (EEPROM) or rewritable EEPROM (electrically erasable programmable read only memory), a hard disk, or the like.

  Further, the production schedule creation apparatus 1000 displays the contents of each storage unit such as the optimum plan storage unit 1800 on the output unit 1003 or displays the input unit 1002 in response to a command input from the user via the input unit 1002. It also has a function of creating, modifying, etc. the contents of the storage unit such as the grouping condition storage unit 1700.

  The production schedule creation apparatus 1000 having such a configuration can be configured using a computer such as a personal computer, for example, and is described above by executing software programmed with a plan search method stored in a storage unit such as a hard disk. The plan search unit 1200 is functionally configured in a computer. Note that one or more of the other functional units shown in FIG. 1 may be configured in the computer.

  When a program for programming a schedule creation method or the like is not stored in a storage unit such as a hard disk of the production schedule creation device 1000, the storage in the production schedule creation device 1000 is recorded from the recording medium on which these are recorded via an external storage unit. It may be configured such that each program is downloaded from a server (not shown) that manages these programs via a network and a communication interface unit. Further, the data stored in the large schedule plan storage unit 1500 or the like may be configured to be input to the production schedule creation apparatus 1000 via the external storage unit by a recording medium storing this data, You may comprise so that it may input into the production schedule production apparatus 1000 via a network and a communication interface part from a user.

<Target of production schedule>
Here, an example of an object for which the production schedule creation apparatus 1000 creates a schedule will be described. Hereinafter, a case where a schedule is created for this target example will be described.

<Equipment>
The target is one facility.

  This equipment can only process one order at a time, and if different types of orders are processed in succession, a setup change may occur. In the embodiment, for convenience of explanation, it is assumed that there are A and B as setups to be generated, and the setup change takes 1/5 day (1/5 day), that is, 4.8 hours. . Note that the time for this setup change is the same as the processing time for one order to be described later.

Further, as the equipment capacity, the processing capacity may be different for each order if it is normal, but in this embodiment, the processing capacity is set to 5 orders / day. Of course, the processing capacity may be changed for each order. For example, for order A, one order is processed in one hour, for order B, one order is processed in two hours, and so on.

  In the embodiment, for convenience of explanation, one facility is targeted, but a plurality of facilities may be targeted.

<Order>
FIG. 3 shows the status of the received order.

  FIG. 3 is a diagram illustrating an example of the configuration and contents of the order information table 1610.

  The order information table 1610 is stored in advance in the operation condition storage unit 1600 and is referred to by the initial plan creation unit 1100 and the like.

  The order information table 1610 includes an order name 1611, setup conditions 1612, an order quantity 1613, and a delivery date 1614.

  The order name 1611 indicates a name for identifying and distinguishing a certain order. In the embodiment, it is assumed that the order name 1611 is “order A-1”, “order A-2” to “order F-2”. Hereinafter, “order with order name 1611 of“ order A-1 ”” or the like is simply referred to as “order“ A-1 ”” or the like.

  Also, it is assumed that orders with the same alphabet in the order name are of the same type. For example, the product is the same type. For example, the order “A-1” and the order “A-2” are different orders but are orders of the same type of product.

  The setup condition 1612 indicates a setup condition necessary when processing the order indicated by the order name 1611. In the embodiment, “A” or “B”. Further, when orders with the same setup conditions are continuously processed, setup change is not necessary. For example, when processing of the order “A-2” is to be started, it is assumed that the order processed before that is the order “A-1”. In this case, a record in which the setup condition set as the setup condition 1612 is “A” in the record in which the order name 1611 is “order A-1”, and the record in which the order name 1611 is “order A-2”. Since the setup condition set as the setup condition 1612 is “A”, the setup is not changed. If the order processed before that is the order “B-1”, the setup condition set as the setup condition 1612 in the record whose order name 1611 is “order B-1” is Since “B” is different from the setup condition “A” of the order “A-2”, setup change is necessary. Hereinafter, the setup condition “A” ”set as the setup condition 1612 in the record whose“ order name 1611 is “order A-1” is simply referred to as the setup condition “order“ A-1 ””. A "" etc.

  The order quantity 1613 indicates a necessary quantity in the order indicated by the order name 1611. Hereinafter, “necessary amount set as an order quantity 1613 in a record whose order name 1611 is“ order A-1 ”” or the like is simply referred to as “order quantity of order“ A-1 ”” or the like.

  In the embodiment, scheduling is performed with the quantity “1” as a minimum unit. For example, since the order quantity of the order “A-1” is “3”, an order number is assigned to the order “A-1”, and “A-1-01”, “A-1-02”, “A-1” Scheduling is performed as three orders “−1-03”. Assuming that this “A-1-01” etc. is one order, and as explained in the section <Equipment> above, when processing this one order by equipment, 1/5 day (1/5 day) ), That is, 4.8 hours.

  The delivery date 1614 indicates the delivery date in the facility that processes the order indicated by the order name 1611. Here, the delivery date is not the date but the number of days from a certain date. Hereinafter, “the delivery date set as the delivery date 1614 for the record whose order name 1611 is“ order A-1 ”” is simply referred to as “the delivery date of the order“ A-1 ””.

<Large schedule plan>
The contents of the large schedule are shown in FIG. The production schedule creation apparatus 1000 creates an optimal schedule based on this large schedule plan.

  FIG. 4 is a diagram illustrating an example of the configuration and contents of the large schedule plan 1510.

  The large schedule plan 1510 is stored in advance in the large schedule plan storage unit 1500 and is referred to by the initial plan creation unit 1100 or the like.

  The large schedule plan 1510 includes an order name & serial number 1511 and a processing date 1512.

  The order name & serial number 1511 indicates an identifier (hereinafter referred to as “order number”) for identifying and distinguishing one order.

  The processing date 1512 indicates the date on which the order indicated by the order name & serial number 1511 is processed by the equipment. In the embodiment, instead of the date, the first day and the second day in the plan are represented. Further, the processing date 1512 indicates only the processing date, and does not indicate the processing order.

  As described above in the section <Order>, the order indicated by the order number is an order of the minimum unit of scheduling and is represented by the order name and its serial number. Since one order takes 4.8 hours, the large schedule plan 1510 processes 5 orders per day and schedules 54 orders to ensure delivery.

<Search method for optimal plan (optimal solution)>
In the present embodiment, a local search method that is a known technique is used for searching for an optimal solution, that is, for creating a candidate plan and updating a provisional solution (optimal solution). This method has a problem that it falls into a local optimal solution although the calculation time is fast. In order to solve this problem, the production schedule creation apparatus 1000 can escape from the local optimum solution by changing the group to be scheduled for each search, that is, each time a candidate plan is created.

  Furthermore, in this embodiment, even if an optimal solution is found under a certain grouping condition, a more optimal plan is made by performing a search while changing the grouping condition.

<Evaluation method>
A method for evaluating the initial plan created by the initial plan creation unit 1100 and the candidate plan created by the plan search unit 1200 will be described. This process is a process performed by the plan evaluation unit 1400.

  In the evaluation of the plan in the production schedule creation device 1000, the delay in delivery date and productivity are evaluated. And the evaluation value of a plan is calculated based on these evaluation values.

  First, regarding the delivery delay amount, the total value of the delivery delay days for each order is used as the evaluation value of the delivery delay amount. The priority may be changed for each order, and the relationship between the number of days delayed and the amount of delay may be expressed exponentially, for example, when the number of days delayed in delivery exceeds a certain value, the amount of delay in delivery is increased.

  Next, regarding the productivity, the following two items are evaluated. These are the number of setup changes at the facility and the number of changes in the type of order being processed. The number of times the type of the order being processed is changed is, for example, the case where the order “D-1” is processed after the order “A-1” (see FIG. 3). Since the setup condition of the order “A-1” is “A” and the setup condition of the order “D-1” is “A”, no setup change occurs. However, since the types of orders are “A” and “D”, the number of changes in the type of order being processed is counted. The number of times used for productivity evaluation is better when processing the same type of orders consecutively than when processing different types of orders, even if there is no setup change. From this point of view. Note that the evaluation value may be calculated not based on the number of times of setup change but based on the time required for the setup change. This is particularly effective when the time required varies depending on the type of setup change.

  Here, an example of calculating the evaluation value will be described using an initial plan.

  FIG. 5 shows an initial plan 1810 created by the initial plan creation unit 1100. The initial plan 1810 includes an order 1811, an order name & serial number 1812, a setup condition 1813, a process start time 1814, a process end time 1815, and a group name 1816. However, the initial plan creation unit 1100 creates the sequence 1811 to the processing end time 1815. The group name 1816 is created by the plan search unit 1200 described later.

  The order 1811 indicates the order of orders processed by the equipment. The order of setup times is not set because orders are not processed.

  The order name & serial number 1812 indicates an order to be processed in the order indicated by the order 1811. In this order name & serial number 1812, 54 order numbers set in the order name & serial number 1511 of the large schedule plan 1510 are set. This setting order is a result of scheduling performed by the initial plan creation unit 1100 using one order indicated by the order name & serial number 1511 as a processing unit.

  The setup condition 1813 indicates a setup condition necessary for processing an order with the order number indicated by the order name & serial number 1812. In this setup condition, the setup condition set as the setup condition 1612 in the record of the order information table 1610 in which the order name of the order number is set as the order name 1611 is set. For example, the order name & serial number 1812 is set as the setup condition 1813 for the record “order A-1-01”, and the order name 1611 is set as the setup condition 1612 for the record “order A-1”. Is set. When the setup set in the setup condition 1813 changes, a setup change time is provided. For example, the order 1811 is between “11” and “12”.

  The processing start time 1814 indicates the date and time when processing of the order with the order number indicated by the order name & serial number 1812 starts, and the processing end time 1815 indicates the date and time when processing ends. The time from the processing start time 1814 to the processing end time 1815 is 4.8 hours, that is, 4 hours 48 minutes.

  The group name 1816 indicates the name of the group to which the order indicated by the order name & serial number 1812 belongs. This grouping is performed by the plan search unit 1200. Grouping will be described in the section <Grouping>.

  The evaluation value of the initial plan 1810 is calculated as follows using the evaluation values of the three items described above, that is, the amount of delay in delivery, the number of times of setup change, and the number of times the order type has changed.

  Since no delivery delay has occurred, the delivery delay amount is “0”, the number of setup changes is “9”, and the number of orders changed (see arrows 1 to 12) is “12”.

  Next, a weight is provided for each evaluation item, and the sum of the evaluation values of each item is used as the evaluation value. For example, if the weight of the delivery delay amount is “10”, the weight of the number of setup changes is “1”, and the weight of the number of times the order is changed is “0.1”, the evaluation value of the initial plan 1810 is 10 × 0 + 1. X9 + 0.1x12 = 10.2. In the case of this evaluation method, the smaller the evaluation value, the higher the evaluation of the plan.

  In addition to using weights, priorities may be provided. For example, the priority is set in the order of the delivery delay amount, the number of setup changes, and the number of orders changed, and the delivery delay amount has the highest priority. In this case, even if the number of setup changes is reduced from 9 to 0, if the delivery date is delayed, the evaluation of the plan is lowered.

<Grouping>
Next, conditions for grouping orders will be described. The plan search unit 1200 performs grouping according to the grouping condition. The grouping condition is stored in the grouping condition storage unit 1700. In addition, it is assumed that a plurality of the grouping conditions are stored, and the plan search unit 1200 selects and searches the conditions in a predetermined order when the search is started with a new condition.

  This grouping is divided into two cases: grouping orders first, i.e. grouping initial plans, and grouping grouped groups further, i.e. grouping candidate plans. is there.

  Hereinafter, a description will be given taking a grouping condition as an example.

  First, grouping conditions for grouping initial plans will be described.

  The grouping condition is “same order executed on the same day as one group, but not including setup change”. For example, since the setup condition of the order “D-1” in the order information table 1610 is “A” and the setup condition of the order “D-2” is “B”, the order “D-1” is ordered. Even if “D-2” is processed consecutively on the same day, the order change is performed after the order “D-1” is processed, so the order “D-1” and the order “D-2” are processed. Are not in the same group.

  Taking the initial plan 1810 (see FIG. 5) as an example, grouping as shown by a group name 1816 is performed. The group name is indicated by an alphabet indicating the order type and a serial number. The alphabet indicating the type of order is the same as the alphabet of the order name. For example, an order with the order 1811 of “1” to “11” has the order type “A”, but the date of the process start time 1814 is different, so the order 1811 of the day with the process start time 1814 is “ Orders “1” to “5” form one group as the group name 1816 “Group A-1”.

  When further searching for an optimal solution from the initial plan 1810, normally, the order indicated by the order name & serial number 1812, that is, 54 orders, are targeted for scheduling. However, the production schedule creation apparatus 1000 searches for an optimal solution with 21 groups as scheduling targets. Accordingly, the amount of calculation can be greatly reduced, and an optimal solution can be obtained in a shorter time.

  Next, grouping conditions for grouping candidate plans will be described.

  The grouping condition in this case is “the same kind of orders is one group, but no setup change is included”. Compared to the grouping condition for grouping the initial plan, the condition “Implemented on the same day” is removed.

  Here, an example of grouping candidate plans will be described with reference to FIG.

  FIG. 6 shows a candidate plan 1820 created based on the initial plan 1810 by the plan search unit 1200 using the group indicated by the group name 1816 of the initial plan 1810 as a processing unit. Here, the number of times of setup change is reduced from 9 to 5, and the scheduling is such that productivity is increased.

  Candidate plan 1820 includes order 1821, group name 1822, setup condition 1823, processing start time 1824, processing end time 1825, corresponding order 1 (1826), corresponding order 2 (1827), corresponding order 3 (1828), corresponding order. 4 (1829), corresponding order 5 (1830), and new group name 1831. However, the plan search unit 1200 creates the order 1821 to the corresponding order 5 (1830) as candidate plans, and creates a new group name 1831 before searching for candidate plans.

  The order 1821 is the same as the order 1811 of the initial plan 1810, and shows the order of orders processed by the equipment. The order of setup times is not set because orders are not processed.

  The group name 1822 indicates the group name of the group to be processed in the order indicated by the order 1821. In the group name 1822, 21 group names set in the group name 1816 of the initial plan 1810 are set. This setting order is a result of scheduling performed by the plan search unit 1200 using the group indicated by the group name 1816 as a processing unit.

  The setup condition 1823 indicates a setup condition necessary for processing the order of the group indicated by the group name 1822. That is, a setup condition for an order belonging to the group indicated by the group name 1822 is set. For example, for the group whose group name 1822 is “Group A-1”, the order “A-1” in which the group name 1816 of the initial plan 1810 is set as the order name & serial number 1812 in the record of “Group A-1”. −01 ”to“ A-2-02 ”belong. Accordingly, the setup condition “A” for the order “A-1” in the order information table 1610 is set. Since the grouping condition includes orders of the same type and no setup change occurs, the setup conditions of orders belonging to the same group are the same.

  The process start time 1824 indicates the date and time when processing of an order belonging to the group indicated by the group name 1822 is started, and the process end time 1825 indicates the date and time when the process ends. The time from the processing start time 1824 to the processing end time 1825 varies depending on the number of orders belonging to the group.

  Corresponding order 1 (1826) to corresponding order 5 (1830) indicate the order numbers of orders belonging to the group indicated by the group name 1822. Specifically, the order number set as the order name & serial number 1812 in the record set as the group name 1816 of the previous plan, that is, the initial plan 1810, is the group name indicated by the group name 1822. Set. In this example, the maximum number of orders belonging to each group is five.

  The new group name 1831 indicates the name of the group to which the group indicated by the group name 1822 belongs.

  Taking the candidate plan 1820 (see FIG. 6) as an example, grouping as shown by the new group name 1831 is performed. This grouping is performed before the plan search unit 1200 further searches for an optimal solution.

  The group name is indicated by an alphabet indicating the type of order and a serial number, and the number of times of grouping is indicated before the alphabet. In FIG. 6, since it is grouping with respect to a group obtained by grouping the initial plan 1810 (see the group name 1816 of the initial plan 1810), “2” is added before the alphabet.

  For example, orders with the order 1821 of “1” to “3” have different processing start time 1824 days, but the order type is “A”, so the new group name 1831 “Group 2A-1” is used. It becomes one group.

  When further searching for an optimal solution from the candidate plan 1820, further grouping is performed to search for an optimal solution with 10 groups as scheduling targets.

  Here, FIG. 7 shows a candidate plan 1840 that is a result of scheduling performed by the plan search unit 1200 with 10 groups indicated by the new group name 1831 of the candidate plan 1820 of FIG. 6 as scheduling targets.

  In the candidate plan 1840, the number of setup changes is reduced from five to three in the candidate plan 1820, and the scheduling is such that productivity is increased.

  The candidate plan 1840 includes an order 1841, a group name 1842, a setup condition 1843, a process start time 1844, a process end time 1845, a corresponding group 1 (1846), a corresponding group 2 (1847), and a corresponding group 3 (1848). The However, before further searching for candidate plans, a column corresponding to the new group name 1831 of the candidate plan 1820 is created and added.

  The order 1841, group name 1842, setup condition 1843, process start time 1844, and process end time 1845 have the same contents as the order 1821, group name 1822, setup condition 1823, and process end time 1825 of the candidate plan 1820, respectively. Show. However, in the group name 1842, the 10 group names set in the new group name 1831 of the candidate plan 1820 are set, and the order of the groups indicated by the new group name 1831 is the processing unit. Is the result of scheduling.

  Corresponding group 1 (1846), corresponding group 2 (1847), and corresponding group 3 (1848) indicate group names belonging to the group indicated by group name 1842. Specifically, the group name indicated by the group name 1842 is set to the record set as the previous group, that is, the new group name 1831 of the candidate plan 1820, and the group name set as the group name 1822 is set. To do. In this example, the maximum number of groups belonging to each group is three.

  Candidate plans created after candidate plan 1840 have the same configuration as candidate plan 1840.

  Here, the evaluation values of the candidate plan 1820 and the candidate plan 1840 are calculated.

  The evaluation value of the candidate plan 1820 indicates that the delivery date delay amount is “0”, the number of setup change times is “5”, and the order type change number is “9”. If the weight of the delay amount is “10”, the weight of the number of setup changes is “1”, and the weight of the number of times the order is changed is “0.1”, the evaluation value of the candidate plan 1820 is 10 × 0 + 1 × 5 + 0. 1 × 9 = 5.9.

  The candidate plan 1840 has an evaluation value of “0” for the delivery delay amount, “3” for the number of setup transitions, and “8” for the number of times the order type has changed. The evaluation value of the plan 1840 is 10 × 0 + 1 × 3 + 0.1 × 8 = 3.8.

  In this way, the number of scheduling objects is reduced by repeating grouping, so that the amount of calculation can be greatly reduced, and an optimal solution can be obtained in a shorter time. Further, in the previous search, it is possible to escape from the local optimal solution that has fallen, and it is possible to plan a better schedule. That is, every time the initial plan 1810, the candidate plan 1820, and the candidate plan 1840 are searched, the evaluation increases to “10.2”, “5.9”, “3.8”, and the better schedule It can be confirmed that the planning has been made.

  Here, a certain grouping condition, that is, a grouping condition in the case of grouping the initial plan is “one order of the same kind executed on the same day as one group. However, it is assumed that no setup change is included”. In the case of grouping candidate plans, the grouping condition is described as “the same kind of orders are set as one group, but no setup change is included”. As other grouping conditions, for example, a condition of “upper limit of the number of orders belonging to one group” or “upper limit of time that can be processed by the setup condition” is added.

  In the embodiment, a new grouping condition is used as a new search condition. However, other search conditions may be added, deleted, changed, or the like.

<Operation>
Hereinafter, the operation of the production schedule creation apparatus 1000 will be described with reference to FIGS. 2 and 5 to 7. When appropriate, the order information table 1610 in FIG. 3 and the large schedule plan 1510 in FIG. 4 are referred to.

  FIG. 2 is a flowchart for explaining the optimum plan creation process of the production schedule creation apparatus 1000.

  First, a user reads and creates information necessary for creating a schedule from a master file (not shown). Specifically, the large schedule plan 1510 to be scheduled is read out and stored in the large schedule plan storage unit 1500. Further, information necessary for scheduling is read out, and an order information table 1610 and the like are created and stored in the operation condition storage unit 1600 as necessary. Also, a grouping condition is created and stored in the grouping condition storage unit 1700. It is assumed that the master file stores various information necessary for creating a production schedule in the production schedule creating apparatus 1000.

  Next, the user inputs a command for instructing creation of a schedule via the input unit 1002.

  When the production schedule creation control unit 1001 receives notification that a command for creating a schedule has been input, the production schedule creation control unit 1001 instructs the initial plan creation unit 1100 to create an initial plan.

  Upon receiving the instruction, the initial plan creation unit 1100 reads out the order information table 1610 and various information necessary for the production schedule planning process from the operation condition storage unit 1600, and reads the large schedule plan 1510 from the large schedule plan storage unit 1500. Read (step S10).

  The initial plan creation unit 1100 that has read the large schedule plan 1510 and the like creates an initial plan 1810 based on the large schedule plan 1510 (step S11), and sends the created initial plan 1810 to the optimum plan determination unit 1300 and the plan search unit 1200. hand over.

  Receiving the initial plan 1810, the optimum plan determination unit 1300 passes the received initial plan 1810 (see FIG. 5) to the plan evaluation unit 1400 and requests an evaluation.

  Upon receiving the request, the plan evaluation unit 1400 performs evaluation as described in the section <Method of evaluation> above (step S12), and returns an evaluation value to the optimal plan determination unit 1300. The evaluation value of the initial plan 1810 is “10.2”.

  The optimum plan determination unit 1300 passed the evaluation value stores the initial plan 1810 in the optimum plan storage unit 1800 in association with the evaluation value (step S13).

  In addition, the plan search unit 1200 that has received the initial plan 1810 from the initial plan creation unit 1100 reads out the grouping condition from the grouping condition storage unit 1700 (step S14), as described in the section <Grouping> above. Then, the initial plan 1810 is grouped according to the read grouping condition (step S15). Specifically, grouping is performed as indicated by a group name 1816.

  The plan search unit 1200 that has performed grouping creates a candidate plan 1820 (see FIG. 6) using the group as a processing unit (step S16), and passes the created candidate plan 1820 to the optimum plan determination unit 1300.

  The optimal plan determination unit 1300 that has received the candidate plan 1820 passes the received candidate plan 1820 to the plan evaluation unit 1400 and requests an evaluation.

  Upon receiving the request, the plan evaluation unit 1400 performs evaluation as described in the section <Method of evaluation> above (step S17), and returns an evaluation value to the optimal plan determination unit 1300. Specifically, the evaluation value “5.9” is returned.

  The optimum plan determination unit 1300 passed the evaluation value uses the evaluation value stored in association with the initial plan 1810 stored in the optimum plan storage unit 1800 and the evaluation value passed from the plan evaluation unit 1400. In comparison, when the evaluation value passed from the plan evaluation unit 1400 has a higher evaluation (step 18: Yes), the candidate plan 1820 is stored in the optimum plan storage unit 1800 in association with the evaluation value (step S19). . At this time, the initial plan 1810 and its evaluation value are deleted. Note that the plan having the highest evaluation (optimum solution) may be stored without being erased.

  Specifically, the evaluation value of the initial plan 1810 is “10.2”, the evaluation value of the candidate plan 1820 is “5.9”, and the evaluation of the candidate plan 1820 is higher. Candidate plan 1820 is stored in optimum plan storage section 1800 in association with the evaluation value.

  On the other hand, if the evaluation value passed from the plan evaluation unit 1400 has a lower evaluation (step 18: No), the plan stored in the optimal plan storage unit 1800 is not rewritten, and therefore the optimal plan storage unit 1800 contains The plan (optimum solution) with the highest evaluation up to that point is stored.

  Next, the plan search unit 1200 determines whether or not to end the search under the current grouping condition (search condition) (step S20). Specifically, when a plan (optimum solution) stored in the optimum plan storage unit 1800 is not updated for a certain number of times even though a candidate plan is created by performing a search, or the time for the search process Is determined to end when a certain time has passed. The fixed number of times or the fixed time is set by the user according to the allowable calculation time and the performance of the computer to be used.

  When it is determined that the search is not terminated (step S20: No), the plan search unit 1200 performs the process from step S15.

  Specifically, the plan search unit 1200 groups candidate plans 1820 (see step S15, new group name 1831), and creates a candidate plan 1840 (see FIG. 7) with these ten groups as scheduling targets (step 7). S16). Then, the created candidate plan 1820 is passed to the optimum plan determination unit 1300. The optimal plan determination unit 1300 passes the received candidate plan 1840 to the plan evaluation unit 1400 to request an evaluation, and the plan evaluation unit 1400 that has received the request, as described in the section <Method of evaluation> above. The evaluation is performed (step S17), and the evaluation value “3.8” is returned to the optimum plan determination unit 1300. Since the evaluation value of the candidate plan 1820 stored in the optimal plan storage unit 1800 is “5.9” (step 18: Yes), the optimal plan determination unit 1300 to which the evaluation value has been passed determines the candidate plan 1840 as the candidate plan 1840. It is stored in the optimum plan storage unit 1800 in association with the evaluation value “3.8” (step S19). At this time, the candidate plan 1820 is deleted.

  On the other hand, when it is determined in step S20 that the search is to be ended (step S20: Yes), the plan search unit 1200 determines whether to end the search for the plan (step S21). Specifically, a search (steps S15 to S20) in which different grouping conditions (search conditions) are set is performed a predetermined number of times, or a search under a new grouping condition is performed once in the optimum plan storage unit 1800. When the stored optimal solution is not updated, it is determined that the process is finished.

  When it is determined that the plan search is to be ended (step S21: Yes), the plan search unit 1200 causes the output unit 1003 to display the plan stored in the optimum plan storage unit 1800 (step S22). For example, grouping is performed based on the candidate plan 1840 (see FIG. 7), and the plan is searched. If the optimal solution stored in the optimal plan storage unit 1800 is not updated, the candidate plan 1840 is Displayed as the optimal solution.

  On the other hand, when it is determined that the plan search is not finished (step S21: No), the plan search unit 1200 starts the process from step S14. Specifically, among the grouping conditions stored in the grouping condition storage unit 1700, the group (initial stage) that was the scheduling target of the plan stored in the optimal plan storage unit 1800 using the new grouping condition is used. If plan 1810 is stored, group orders) and continue the search. Note that the initial plan 1810 may be stored in the memory, and the scheduling target of the initial plan 1810 may be grouped. Further, instead of the plan stored in the optimum plan storage unit 1800, the scheduling target of the newest plan created by the plan search unit 1200 may be grouped.

  As described above, it is possible to take into consideration the productivity and delivery date of the entire factory by considering a large schedule at high speed, and to make a highly productive schedule in each process.

  In order to express the present invention, the present invention has been properly and fully described through the embodiments with reference to the drawings. However, those skilled in the art can easily change and / or improve the above-described embodiments. It should be recognized that this is possible. Therefore, unless the modifications or improvements implemented by those skilled in the art are at a level that departs from the scope of the claims recited in the claims, the modifications or improvements are not covered by the claims. To be construed as inclusive.

1000 Production schedule creation device 1001 Production schedule creation control unit 1002 Input unit 1003 Output unit 1100 Initial plan creation unit 1200 Plan search unit 1300 Optimal plan determination unit 1400 Plan evaluation unit 1500 Large schedule plan storage unit 1600 Operation condition storage unit 1700 Grouping condition Storage unit 1800 Optimal plan storage unit 1510 Large schedule plan 1610 Order information table 1810 Initial plan 1820 1840 Candidate plan

Claims (7)

A production schedule creation device for creating a schedule of equipment for sequentially processing a plurality of types of products from an initial schedule created in advance ,
Creating means for creating a candidate schedule using a first group of a plurality of products as a processing unit;
Wherein the product to be processed in the initial schedule groups using a first condition to generate the first group, a first process for creating a candidate schedule to said creation means, when said producing means a candidate schedule the first group was treated units, and further grouped by using the second condition to generate a second group, the second group as the first group, to create a candidate schedule to said creating means Control means for performing a second process of repeating the process ;
Evaluation means for evaluating each of the initial schedule and the candidate schedule created by the creation means;
A production schedule creation device comprising: a determination unit that determines a schedule with the highest evaluation among the initial schedule and the candidate schedule evaluated by the evaluation unit as an optimal schedule. The control means performs the second process until a predetermined end condition is satisfied,
The predetermined end condition is that the number of consecutive times that the evaluation of the candidate schedule created by the creation unit does not exceed the highest evaluation among the evaluation of the candidate schedule created earlier and the evaluation of the initial schedule exceeds the predetermined number of times. production schedule creation device according to claim 1, characterized in that it has. When the predetermined end condition is satisfied, the control unit generates a second group by grouping using a new predetermined condition as the second condition instead of the second condition currently used. , the second group as the first group, the production schedule creation device according to claim 2, characterized in that repeated thereby creating a candidate schedule to the creating means. In order to process the product, it is necessary to change the equipment.
The first and second conditions include a condition that, when products belonging to the same group are successively processed, the products need not be in the same group when the setup change is required. The production schedule creation device according to any one of claims 1 to 3. 5. The production schedule creation according to claim 4, wherein the evaluation unit increases the evaluation of the initial schedule or the candidate schedule as the number of times of the setup change included in the initial schedule or the candidate schedule decreases. apparatus. The production schedule creation device further includes storage means for storing a large schedule that takes into account equipment other than the equipment,
The creation means is a candidate schedule based on the large schedule, and creates a candidate schedule that performs processing within a predetermined period from when each of the products in the large schedule is processed. The production schedule creation device according to claim 1. A production schedule creation method used in a production schedule creation device that creates a schedule of equipment that sequentially processes a plurality of types of products from an initial schedule created in advance ,
A creation process for creating a candidate schedule using a first group of a plurality of products as a processing unit;
When the products to be processed in the initial schedule are grouped using a first condition to generate the first group, and the candidate schedule is created in the creation step, and the candidate schedule is created in the creation step the first group was treated units, and further grouped by using the second condition to generate a second group, the second group as the first group, to create a candidate schedules the generation step and a control step of performing a second process of repeating the you,
An evaluation step for evaluating each of the candidate schedules created in the initial schedule and the creation step;
A determination process for determining a schedule with the highest evaluation among the initial schedule and the candidate schedule evaluated in the evaluation process as an optimal schedule.

Images