KR20240164215A - Multi-constraint greedy job scheduling system and method thereof - Google Patents

Abstract

The present invention relates to a system and a method for scheduling an automatic task of a mold processing smart factory and, more specifically, to a system and a method for scheduling a multi-constraint gridy task, which assigns a task for producing a plurality of molds by applying a multi-constraint gridy algorithm to a plurality of machines performing a single task among one or more tasks of a task for generating a mold to automatically schedule a mold task.

Description

{Multi-constraint greedy job scheduling system and method thereof}

The present invention relates to an automatic work scheduling system and method for a smart factory for mold processing, and more particularly, to a multi-constraint greedy work scheduling system and method for automatically allocating multiple operations of tasks for an arbitrary mold by applying a multi-constraint greedy algorithm to multiple machines performing a single operation among one or more operations for creating a mold, thereby automatically performing scheduling for a mold operation.

Typically, a mold processing plant is equipped with multiple machines that handle the process operations for processing molds, and performs work scheduling to efficiently allocate multiple process operations for each of the multiple molds to the machines in order to efficiently produce multiple molds.

Traditionally, in mold processing plants, most of these work schedulings have been done by the labor of the relevant managers. However, there is a problem that work scheduling by people takes a lot of time and it is difficult to efficiently assign work to machines.

In addition, following the recent boom in artificial intelligence (AI), smart factory technology is spreading rapidly, causing social changes. In response to these social changes, companies are trying to utilize advanced automatic technologies to help with work scheduling and processing, and interest in automatic work scheduling technology that automatically allocates process tasks in smart factories is continuously increasing.

In particular, mold processing plants are trying to automate the scheduling of process operations for mold processing, but automatically scheduling operations for mold processing in mold processing plants causes many problems (commonly referred to as "Job Shop Scheduling Problem (JSSP)") in the production environment of a discontinuous production system (job shop) that produces a variety of products in small or medium quantities.

To solve the JSSP problem, it is necessary to pre-designate a machine for each processing operation of an arbitrary mold, and to be able to assign the order of the processing molds on that machine to reduce the processing time.

However, the machines of the conventional automatic scheduling system have the problem that they do not know the work to be actually processed, cannot assign the order of processing molds, cannot set a processing deadline when contracting for production, and cannot perform scheduling according to priority according to customer requests.

Republic of Korea Patent No. 10-2342028 (announced on December 22, 2021)

Gon¸calves, J. F., de Magalhaes Mendes, J. J., & Resende, M. G. (2005). A hybrid genetic algorithm for the job shop scheduling problem. European journal of operational research, 167, 77-95. Amjad, M. K., Butt, S. I., Kousar, R., Ahmad, R., Agha, M. H., Faping, Z., Anjum, N., & Asgher, U. (2018). Recent research trends in genetic algorithm based feasible job shop scheduling problems. Mathematical Problems in Engineering, 2018.

Accordingly, the purpose of the present invention is to provide a multi-constraint greedy job scheduling system and method capable of automatically allocating multiple tasks for an arbitrary mold by applying a multi-constraint greedy algorithm to multiple machines performing a single task among one or more tasks for creating a mold, thereby automatically performing mold task scheduling, thereby enabling setting mold priorities according to customer requests, reflecting processing deadlines, and performing task scheduling that takes machine efficiency into account.

In order to achieve the above object, the present invention provides a multi-constraint greedy job scheduling system, which comprises: a plurality of job processing units configured to be installed in a machine for performing a job for one of a plurality of processes for producing a product, and which process corresponding process jobs for materials loaded into the machine according to process scheduling information corresponding to the machine; and a scheduling unit configured to store and manage tasks for at least one product ordered for production, at least one process for producing the product of each task, basic work times for each process, and the efficiency of the machine, determine a processing order for each product, and apply the determined product processing order, idle time, and efficiency of the machine to a greedy algorithm with multiple constraints, thereby allocating the process jobs for each product to the corresponding machine to generate scheduling information including process scheduling information for each machine, and outputting the scheduling information so that the corresponding job processing unit processes the job according to the corresponding process scheduling information of the scheduling information.

The above product is a mold, and the scheduling unit comprises: a storage unit including a mold DB storing task information on tasks of at least one mold for which a production order has been placed; a work DB storing mold-specific process information including at least one process and basic work time for each process required to manufacture the mold for each mold; a machine DB storing machine status information including the efficiency of each machine, the work time of each machine, and the idle time of each machine; and a scheduling DB storing scheduling information including process scheduling information of each machine; a task sequencing unit determining a processing order of each mold by sequencing tasks of the task information registered in the mold DB according to a preset sequencing condition, and storing the sequenced processing order information in the mold DB; a work scheduling unit applying the determined mold processing order, the idle time and the efficiency of the machine to a greedy algorithm as multiple constraints, and allocating the processing work of each product to the corresponding machine, thereby generating machine-specific process scheduling information according to the process scheduling of each machine, and storing the same in the work DB; a task scheduling unit generating scheduling information including process scheduling information of each process and storing it in the scheduling DB; And it is characterized by including a scheduling output unit that outputs the scheduling information so that the corresponding work processing unit processes the work according to the corresponding process scheduling information of the scheduling information.

The above-mentioned task scheduling unit is characterized by including: an idle time confirmation unit that checks the idle time of each machine of the machine DB when scheduling a process for a task of an arbitrary mold, outputs the checked idle time, and after allocating a process task for the task to an arbitrary machine, updates the idle time of the machine to which the process task of the machine DB is allocated to the time after the actual work time of the process task according to the allocation of the process task; a task process work time calculation unit that calculates and outputs the actual work time for the basic work time of the process task assigned to the corresponding machine by reflecting the efficiency of each machine of the machine DB; and a process work allocation unit that applies the mold processing order, the actual work time, and the idle time of each machine as multiple constraints to a greedy algorithm to allocate the process task to the machine, generates machine-specific process scheduling information according to the process scheduling of each machine, and stores it in the task DB.

The above system: further includes a progress confirmation unit that confirms the progress of the process work being performed according to the work time allocated to each machine through the work processing unit and provides the progress work time and progress information to the scheduling unit, wherein the scheduling unit further includes a process progress confirmation unit that receives the progress work time and progress information from the progress confirmation unit, calculates the efficiency of the corresponding machine according to the progress work time and progress information, and updates the efficiency of the corresponding machine in the machine DB.

The above task process work time calculation unit is characterized by calculating the actual work time using the following mathematical formula.

[Mathematical formula]

Here, t' _j,A is the time added according to efficiency, p' _bh,j is the standard working time for j operation of h (mold) assigned to machine A, and e _j,A is the efficiency of machine A.

The above task sequencing unit is characterized by sequencing according to the order of the mold tasks.

The above task sequencing unit is characterized by sequencing according to the ordering order of the mold tasks and the importance applied to at least one of the ordering work price, mold task delivery speed, and delivery period.

The above system is characterized by further including an information acquisition unit that acquires machine status information including task information for an ordered mold, process information required for mold processing of the ordered task, basic work time for each process, and operation start time and operation end time for each machine, and stores the information in the mold DB, work DB, and machine DB of the storage unit.

In order to achieve the above object, the present invention provides a multi-constraint greedy job scheduling method, characterized in that: a scheduling process in which a scheduling unit stores and manages tasks for at least one mold for which a manufacturing order has been placed, at least one process for producing a product of each task, a basic working time for each process, and the efficiency of the machine, determines a processing order for each product, and applies the determined processing order, the idle time and the efficiency of the machine to a greedy algorithm as multiple constraints, thereby allocating the processing work for each product to a corresponding machine and generating scheduling information including process scheduling information for each machine; and a process processing execution process in which a task processing unit configured in a machine that performs a task for one process among a plurality of processes for producing a product performs a corresponding public task according to the corresponding process scheduling information of the scheduling information.

The above scheduling process is characterized in that it includes a task sequencing step in which the scheduling unit determines a processing order for each mold by sequencing tasks of task information registered in the mold DB according to a pre-set sequencing condition through a task sequencing unit, and stores the sequenced processing order information in the mold DB; a task scheduling step in which the scheduling unit applies the mold processing order determined through the work scheduling unit, the idle time and efficiency of the machine to a greedy algorithm as multiple constraints, and allocates the product-specific process work to the corresponding machine to generate machine-specific process scheduling information according to the machine-specific process scheduling, and stores the generated process scheduling information in the work DB; a task scheduling step in which the scheduling unit generates scheduling information including process-specific process scheduling information through the task scheduling unit, and stores the generated process scheduling information in the scheduling DB; and a scheduling output step in which the scheduling unit outputs the scheduling information through a scheduling output unit so that the corresponding work processing unit processes the work by the corresponding process scheduling information of the scheduling information.

The above work scheduling step includes: an idle time checking step in which the scheduling unit checks the idle time of each machine in the machine DB when scheduling a process for a task of an arbitrary mold through the idle time checking step of the work scheduling unit, and outputs the checked idle time; an actual working time calculation step in which the scheduling unit calculates and outputs the actual working time for the basic working time of the process task assigned to the corresponding machine by reflecting the efficiency of each machine in the machine DB through the task process working time calculation step of the work scheduling unit; a process working allocation step in which the scheduling unit applies the mold processing order, the actual working time, and the idle time of each machine as multiple constraints to a greedy algorithm through the process working allocation unit of the work scheduling unit to allocate the process tasks to the machines, thereby generating machine-specific process scheduling information according to the process scheduling of each machine, and storing the information in the work DB. And the scheduling unit includes an idle time updating step for updating the idle time of a machine to which the process task is assigned in the machine DB to a time after the actual working time of the process task according to the assignment of the process task after the process task is assigned to an arbitrary machine through the idle time checking unit.

The above method further includes a progress confirmation process in which the progress confirmation unit confirms the progress of the ongoing process work according to the work time allocated to each machine through the work processing unit and provides the ongoing work time and progress information to the scheduling unit, wherein the scheduling process further includes a process progress confirmation step in which the scheduling unit receives the ongoing work time and progress information from the progress confirmation unit, calculates the efficiency of the corresponding machine according to the ongoing work time and progress information, and updates the efficiency of the corresponding machine in the machine DB.

The above task process work time calculation unit is characterized in that it calculates the actual work time by the following mathematical formula.

[Mathematical formula]

Here, t' _j,A is the time added according to efficiency, p' _bh,j is the standard working time for j operation of h (mold) assigned to machine A, and e _j,A is the efficiency of machine A.

The above task sequencing unit is characterized by sequencing according to the ordering order of the mold tasks and the importance applied to at least one of the ordering work price, mold task delivery speed, and number of delivery days.

The present invention has the effect of resolving JSSP by sequentially processing ordered molds so that the processing order of the molds is fixed in advance.

In addition, the present invention has the effect of improving customer satisfaction by sequencing ordered molds while reflecting a number of conditions, such as the mold delivery period.

The present invention performs scheduling by considering single machines with different efficiencies and working times, so that it can solve JSSP, maximize the efficiency of the machines, and improve work efficiency and work speed.

FIG. 1 is a diagram showing the configuration of a multi-constraint greedy job scheduling system according to the present invention.
FIG. 2 is a drawing showing the detailed configuration of the scheduling part of the multi-constraint greedy job scheduling system according to the present invention.
FIG. 3 is a flowchart illustrating a multi-constraint greedy work scheduling method according to a first embodiment of the present invention.
FIG. 4 is a flowchart illustrating a multi-constraint greedy work scheduling method according to a second embodiment of the present invention.

Hereinafter, with reference to the attached drawings, the configuration and operation of a multi-constraint greedy job scheduling system according to the present invention will be described in detail, and a job scheduling method of the system will be described.

FIG. 1 is a diagram showing the configuration of a multi-constraint greedy job scheduling system according to the present invention.

The multi-constraint greedy job scheduling system according to the present invention includes a plurality of job processing units (10) and a scheduling unit (20), and further includes a progress checking unit (30) according to an embodiment.

The work processing unit (10) is equipped in a machine that performs a task for one of a number of processes (types) for producing a specific product, and processes the corresponding process task for materials loaded in the machine according to process scheduling information corresponding to the machine. The product is preferably a mold, but is not limited to a mold. In the following description, a case where the product is a mold will be described as an example. The work scheduling system according to the present invention is equipped with at least one machine corresponding to each process. That is, two or more machines that perform tasks of the same process type may be equipped.

The work processing unit (10) may be controlled by a worker who has confirmed the process scheduling information assigned to the machine, but it is preferable to receive and store the process scheduling information from the scheduling unit (20) and configure the tasks of the stored process scheduling information to be automatically performed according to the scheduling of the process scheduling information.

The work processing unit (10) may be configured to count the work progress time from the start time of the work assigned to it according to an embodiment, check the progress of the work (defined as a percentage), and output it to the progress check unit (30).

The scheduling unit (20) stores and manages task information for at least one mold ordered by customers, process information including at least one process for producing the mold for each task and basic working time for each process, and machine status information for each machine.

The above task information may include a 3D model for the ordered mold, the number of molds requested to be manufactured, the delivery period, etc.

If the above product is a mold, the above process (type: Class) may be a grinding electrode, a high-speed processing, a high-speed electrode, grinding, a wire, a discharge electrode, etc. In this case, the machine types may be a grinding electrode machine, a high-speed processing machine, a grinding machine, a wire machine, a discharge electrode machine, etc.

The above machine status information may include information such as unique identification information for each machine (hereinafter referred to as “machine identification information”, the machine identification information may be identification information of the work processing unit (10), efficiency for each machine (defined as a percentage), and work time including the work start time and work end time for each machine.

When multiple mold orders are generated, the scheduling unit (20) determines the processing order for the ordered molds, applies the determined mold processing order, machine idle time, machine efficiency, etc. to a greedy algorithm as multiple constraints, and allocates the processing work for each mold to the corresponding machine, thereby generating scheduling information for all machines including process scheduling information for each machine, and outputs the scheduling information so that the corresponding task processing unit (10) can process the work according to the corresponding process scheduling information of the scheduling information.

The detailed configuration of the above scheduling unit (20) is explained with reference to the following Figure 2.

The progress confirmation unit (30) confirms the progress (defined as a percentage) according to the work progress time of the process work being performed and allocated to each machine through the work processing unit (10) and provides the work progress time and progress information to the scheduling unit (20).

FIG. 2 is a drawing showing the detailed configuration of the scheduling part of the multi-constraint greedy job scheduling system according to the present invention.

Referring to FIG. 2, the scheduling unit (20) includes a storage unit (110), a task sequencing unit (130), a work scheduling unit (140), a task scheduling unit (150), and a scheduling output unit (160), but may further include an information acquisition unit (120) and a process progress confirmation unit (170) depending on the embodiment.

The storage unit (110) includes a mold DB (111), a work DB (112), a machine DB (113), and a scheduling DB (114).

The above mold DB (111) stores task information for at least one mold-specific task ordered for production.

The above work DB (112) stores process information including at least one process operation required to manufacture the mold for each mold, a basic work time for each process, and an actual work time for each machine for the process. The basic work time is the time taken to perform the process operation for the machine assuming that the efficiency of the machine to be assigned is 100%. The actual work time is the time taken to complete the process operation for the machine according to the actual efficiency of the machine. The actual work time is the same as the basic work time when the actual efficiency of the machine is 100%, and the actual work time is longer than the basic work time when it is less than 100%, and shorter than the basic work time when it is 100% or more. However, since the efficiency of the machine is very unlikely to improve from the initial level, the case where the efficiency is 100% or more is not considered in the following description.

The machine DB (113) stores machine status information for each machine configured in a smart factory for processing various types of molds according to the present invention. The machine status information may include machine identification information for each machine, efficiency for each machine, work start time and work end time for each machine, etc.

The scheduling DB (114) stores scheduling information including process scheduling information for each machine.

The information acquisition unit (120) is configured according to an embodiment, and acquires task information for a mold ordered by a customer, process information for processing the mold of the task, and machine status information, and stores them in the mold DB (111), work DB (112), and machine DB (113) of the storage unit (110).

The above information acquisition unit (120) may acquire the task information and process information from an administrator, or may automatically acquire them from a 3D model input for an ordered mold.

In addition, the information acquisition unit (120) will acquire machine status information from the manager. The efficiency of the machine status information may be periodically updated by the manager through the information management unit (120), or may be configured to be automatically updated through the process progress confirmation unit (170) described later.

The task sequencing unit (130) determines the processing order for each mold by sequencing the tasks for each mold of the task information registered in the mold DB (111) according to preset sequencing conditions, and stores the sequenced processing order information in the mold DB (111).

The above ordering condition may be a pre-order condition based on the order receiving date, or may calculate the importance by comprehensively applying one or more of the order order, order work price, mold task delivery speed, and delivery period, and may be ordered according to the calculated importance. The above importance may be calculated by assigning a weight to each of the requirements (order order, order work price, mold task delivery speed, delivery period), calculating a score according to the value range of each requirement, and then multiplying the score for each requirement by the respective weight. For example, the value category and score of the requirement may be determined as 1-10, 11-20, etc. when the requirement is the order order, and the scores may be assigned as 100 points for 1-10, 80 points for 11-20, etc. As another example, if the requirement is mold task delivery speed, the value ranges could be Ultra Fast, Urgent, Normal, etc., and the points could be assigned as 100 for Ultra Fast, 80 for Urgent, 60 for Normal, etc.

The work scheduling unit (140) includes an idle time checking unit (141), a task process work time calculation unit (142), and a process work allocation unit (143), and applies the determined mold processing order, the idle time of the machine, and the efficiency of the machine as multiple constraints to a greedy algorithm, thereby allocating the process work for each mold to the corresponding machine, thereby generating process scheduling information for each machine according to the process scheduling for each machine, and storing it in the work DB (112).

Specifically, when scheduling a process for a task of an arbitrary mold, the idle time confirmation unit (141) confirms the idle time of each machine corresponding to the process of the task of the machine DB (113) and outputs the confirmed idle time to the process work allocation unit (143).

In addition, the idle time verification unit (141) updates the idle time of the machine to which the process task is assigned in the machine DB (113) after the process task for the task is assigned to an arbitrary machine, to the time after the actual work time of the process task according to the assignment of the process task.

The task process work time calculation unit (142) calculates the actual work time for the basic work time of the process work assigned to the corresponding machine by reflecting the efficiency of each machine in the machine DB (113) and outputs it to the process work allocation unit (143). In particular, when two or more machines exist for the same type of process, the task process work time calculation unit (142) calculates the actual work time of each machine reflecting the efficiency of each of the two or more machines of the same process type and stores it as process information in the task DB (112) or outputs it to the process work allocation unit (143).

The process work allocation unit (143) applies the mold processing sequence, basic work time, and idle time per machine as multiple constraints to a greedy algorithm according to the first embodiment to allocate process work to the machine, thereby generating process scheduling information per machine according to process scheduling per machine and storing it in the work DB (112).

In addition, the process work allocation unit (143) applies the greedy algorithm with the mold processing order, actual work time according to the efficiency of each machine, and idle time of each machine as multiple constraints according to the second embodiment to allocate process work to the machines, thereby generating process scheduling information for each machine according to process scheduling for each machine and storing it in the work DB (112).

A smart factory has four types of processes: grinding electrode (j = 1, where j is the process type index), high-speed machining (j = 2), wire (j = 3), and grinding (j = 4); and machines include machine 1 (m _j,k = m _1,1 , where k = machine index, k = 1,2,3...), machine 2 (m _1,2 ) that process grinding electrode processes, machine 3 (m _2,3 ) that process high-speed machining processes, machine 4 (m _3,4 ) that process wire processes, machine 5 (m _4,5 ) and machine 6 (m _4,6 ) that process grinding processes; and there are mold 1, mold 2, and mold 3 that are ordered and sequenced. Mold 1 requires grinding electrode, wire, and high-speed machining processes, mold 2 requires grinding electrode, wire, and grinding processes, and mold 3 requires high-speed machining and wire processes.

In addition, the status information of the above machines is as shown in Table 1 below, and the process information of the mold is explained as an example in Table 2 below.

In the case of the example described above, the task process work time calculation unit (142) calculates the actual work time using the following mathematical expression 1 and records it in the actual work time field as shown in Table 2.

Here, t' _j,A is the delay time according to the efficiency of machine A, and e _j,A is the efficiency of machine A.

The process work allocation unit (143) checks the idle time of the machines through the idle time check unit (141) and calculates the actual work time through the task process work time calculation unit (142). Then, the grinding process of mold 1 with the highest priority is assigned to machine 1 among machines 1 and 2, which can finish the work the fastest. The wire process is assigned to machine 4, and the high-speed machining process is assigned to machine 3.

The above process task allocation unit (143) allocates process tasks to the machine and then provides the allocation information to the idle time confirmation unit (141).

Then, the idle time checking unit (141) updates the initial idle time of machine 1 in Table 1 from 8:00AM to 11:00AM as the 3-hour grinding electrode process is assigned to machine 1, updates the idle time of machine 4 from 8:00AM to 9:00AM as the wire process is assigned to machine 4, and updates the idle time of machine 3 from 8:00AM to 9:00AM as the high-speed machining process is assigned to machine 3.

When the work scheduling for mold 1 is completed, the process work allocation unit (143) allocates the grinding electrode process work for mold 2, which is the next processing sequence. At this time, the idle time of machine 1 is 11:00 AM, and the idle time of machine 2 is 8:00 AM. Therefore, the process work allocation unit (143) adds the actual work time of each machine 1 and machine 2 to their respective idle times. Since the idle time of machine 1 is 11:00 AM + 2 = 1:00 PM, and the idle time of machine 2 is 8:00 AM + 2.5 = 10:30 AM, the grinding electrode process work is allocated to machine 2.

Then, the idle time check unit (141) will update the idle time of machine 2 to 10:30AM.

As described above, the process work allocation unit (143) schedules process work for tasks of an ordered mold by applying a greedy algorithm that uses the mold processing sequence, actual work time reflecting efficiency, and idle time as multiple constraints, and creates process scheduling information for each machine and stores it in the work DB (112).

The task scheduling unit (150) generates scheduling information including process scheduling information for each process and stores it in the scheduling DB (114).

The scheduling output unit (160) outputs the scheduling information so that the corresponding work processing unit (10) processes the work according to the corresponding process scheduling information of the scheduling information.

The process progress confirmation unit (170) is configured according to an embodiment, and can receive progress information and work progress time directly from each of the work processing units (10) or from the progress confirmation unit (30), calculate the efficiency of the corresponding machine according to the progress information and work progress time, and update the efficiency of the corresponding machine in the machine DB (113).

FIG. 3 is a flowchart illustrating a multi-constraint greedy work scheduling method according to a first embodiment of the present invention.

Referring to FIG. 3, first, the information acquisition unit (120) acquires machine status information including efficiency of each machine and stores it in the machine DB (112) (S111), and creates an idle table for each machine for recording and updating the machine-specific working time (including the work start time and work end time: the time from the work start time to the work end time) and the machine-specific idle time, and stores it by including it in the machine status information (S113).

The information acquisition unit (120) acquires task information on the task of the ordered mold and stores it in the mold DB (111), and acquires the process for producing the mold for each task of the ordered mold and the basic work time for each process and stores it in the work DB (112) (S115).

After the above-described task information, process information, and machine status information are stored, the task sequencing unit (130) monitors whether a scheduling event occurs (S117), and when a scheduling event occurs, the tasks of the ordered molds are sequenced according to preset sequencing conditions to determine the processing order for each mold, and the processing order information is stored in the mold DB (111) (S119). The above-described scheduling event may be generated by a command from an administrator, may be automatically generated at a preset time with a preset cycle, and may be generated when the above-described information (task information, machine status information, and process information) is stored due to the occurrence of a new mold order.

When a scheduling event occurs and the ordered molds are sequenced, the work scheduling unit (140) assigns each task of the mold to a machine by applying a greedy algorithm based on the basic work time of each process task and the idle time of the machine as multiple constraints (S121). At this time, the idle time of the machine status information of each machine is updated by adding the basic work time.

When a task is assigned, the task scheduling unit (140) calculates the actual working time of the machine for the basic working time by reflecting the efficiency of the machine (S123), and then updates the idle time recorded in the idle time field of the idle table to an idle time that reflects the difference between the basic working time and the actual working time (S125).

The above work scheduling unit (140) will generate process scheduling information for each machine when work scheduling is completed, and the task scheduling unit (50) will generate scheduling information including process scheduling information.

FIG. 4 is a flowchart illustrating a multi-constraint greedy work scheduling method according to a second embodiment of the present invention. In the following description with reference to FIG. 4, steps performing the same operations as S111 to S119 of FIG. 3 described above are given the same reference numerals, and their descriptions are omitted.

When the mold work order is determined in S119, the work scheduling unit (140) determines the type of machine corresponding to the process work type of the mold to be scheduled (S211). For example, if the process work type is a grinding electrode, a grinding electrode machine corresponding to the operating electrode is determined.

When the machine type for each process operation is determined, the work scheduling unit (140) determines whether there are two or more machines determined for the operation of one mold (S213). For example, in Table 2 described above, there are machines 1 and 2 that can be applied to the grinding electrode operation, so this will be applicable.

If there is one, the work scheduling unit (140) designates a machine (work processing unit (10)) to process each work task of the mold (S215) and calculates the actual work time according to the efficiency of the designated work processing unit (S217).

When the actual working time of the work assigned to the machine is calculated, the work scheduling unit (140) updates the idle time by reflecting the actual working time in the idle time field of the machine status information of the machine idle table as the actual working time (S219).

On the other hand, if there are two or more tasks, the task scheduling unit (140) calculates the actual work time of the corresponding task for each machine reflecting the efficiency of each of the two or more machines (S221).

When the actual working time for each of two or more machines is calculated, the work scheduling unit (140) applies a greedy algorithm that uses the mold work start time, the machine idle time, and the actual working time of each machine as multiple constraints to assign process tasks to the machines (S223). For example, the work scheduling unit (140) adds the calculated actual working time to each idle time to determine which machine finishes the task faster, and assigns the task preferentially to the machine that finishes the task faster. In addition, at this time, the work scheduling unit (140) may be configured to assign the machine that finishes the task faster by further considering the mold work time.

When the task assignment is completed, the task scheduling unit (140) updates the idle time by reflecting the actual work time in the idle time of the machine to which the task is assigned (S225).

Meanwhile, it will be readily understood by those skilled in the art that the present invention is not limited to the above-described typical preferred embodiments, but can be implemented by improving, changing, replacing or adding in various ways without departing from the gist of the present invention. If the implementation by such improvement, change, replacement or addition falls within the scope of the appended claims below, the technical idea thereof should also be considered to belong to the present invention.

10: Work Processing Unit (Machine) 20: Scheduling Unit
30: Progress check section 110: Storage section
111: Mold DB 112: Work DB
113: Machine DB 114: Scheduling DB
120: Information acquisition section 130: Task sequencing section
140: Job scheduling section 141: Idle time checking section
142: Task Processing Time 143: Processing Work Allocation Section
150: Task scheduling section 160: Scheduling output section
170: Process progress check section

Claims (14)

A plurality of work processing units configured to perform work on one of a plurality of processes for producing a product, the plurality of work processing units processing the corresponding process work on materials loaded into the machine according to process scheduling information corresponding to the machine; and
A multiple constraint greedy job scheduling system, characterized by including a scheduling unit which stores and manages tasks for at least one product ordered for production, at least one process for producing the product of each task, basic working time for each process, and the efficiency of the machine, determines a processing order for each product, applies the determined processing order, the idle time and the efficiency of the machine to a greedy algorithm as multiple constraints, allocates the processing work for each product to the corresponding machine, generates scheduling information including process scheduling information for each machine, and outputs the scheduling information so that the corresponding job processing unit processes the work by the corresponding process scheduling information of the scheduling information.
In the first paragraph,
The above product is a mold,
The above scheduling section,
A mold DB that stores task information for at least one mold that has been ordered for production.
A work DB that stores process information for each mold, including at least one process and basic work time for each process required to manufacture the mold for each mold.
Machine DB that stores machine status information including machine-specific efficiency, machine-specific working time, and machine-specific idle time.
A storage unit including a scheduling DB that stores scheduling information including process scheduling information for each machine;
A task sequencing unit that determines the processing order for each mold by sequencing tasks of task information registered in the mold DB according to preset sequencing conditions and stores the sequenced processing order information in the mold DB;
A work scheduling unit that applies the determined mold processing order, idle time and efficiency of the machine as multiple constraints to a greedy algorithm, assigns the product-specific process work to the corresponding machine, generates machine-specific process scheduling information according to the machine-specific process scheduling, and stores it in the work DB;
A task scheduling unit that generates scheduling information including process scheduling information for each process and stores it in a scheduling DB; and
A multi-constraint greedy job scheduling system characterized by including a scheduling output unit that outputs the scheduling information so that the corresponding job processing unit processes the job according to the corresponding process scheduling information of the scheduling information.
In the second paragraph,
The above work scheduling section,
An idle time verification unit that checks the idle time of each machine in the machine DB when scheduling a process for a task of an arbitrary mold, outputs the checked idle time, and after assigning a process task for the task to an arbitrary machine, updates the idle time of the machine to which the process task is assigned in the machine DB to the time after the actual work time of the process task according to the assignment of the process task;
A task process work time calculation unit that calculates and outputs the actual work time for the basic work time of the process work assigned to the machine by reflecting the efficiency of each machine in the above machine DB; and
A multiple constraint greedy job scheduling system characterized by including a process job allocation unit that applies a greedy algorithm with the mold processing sequence, actual work time, and idle time of each machine as multiple constraints to allocate process tasks to the machines, thereby generating machine-specific process scheduling information according to machine-specific process scheduling and storing the information in the job DB.
In the third paragraph,
The above work processing unit further includes a progress confirmation unit that checks the progress of the process work being performed according to the work time allocated to each machine and provides the progress information and work time to the above scheduling unit.
The above scheduling section,
A multi-constraint greedy work scheduling system characterized by further including a process progress verification unit that receives progress information and work progress time from the progress verification unit, calculates the efficiency of the corresponding machine according to the progress information and work progress time, and updates the efficiency of the corresponding machine in the machine DB.
In the third paragraph,
The above task process work time calculation section is,
A multi-constraint greedy work scheduling system characterized by calculating the actual working time by the following mathematical formula.
[Mathematical formula]

Here, t' _j,A is the time added according to efficiency, p' _bh,j is the standard working time for j operation of h (mold) assigned to machine A, and e _j,A is the efficiency of machine A.
In the second paragraph,
The above task sequencing unit,
A multi-constraint greedy job scheduling system characterized by sequencing according to the order of mold tasks.
In the second paragraph,
The above task sequencing unit,
A multi-constraint greedy job scheduling system characterized by sequencing the order of mold tasks and the importance of applying at least one of the order job price, mold task delivery speed, and delivery period.
In the first paragraph,
A multi-constraint greedy work scheduling system characterized by further including an information acquisition unit that acquires machine status information including task information for an ordered mold, process information required for mold processing of an ordered task, basic work time for each process, and operation start time and operation end time for each machine, and stores the acquired information in the mold DB, work DB, and machine DB of the storage unit.
A scheduling process in which a scheduling unit stores and manages tasks for at least one mold for which a production order has been made, at least one process for producing a product of each task, a basic working time for each process, and the efficiency of the machine, determines a processing order for each product, and applies the determined processing order, idle time, and efficiency of the machine to a greedy algorithm as multiple constraints, thereby allocating the processing work for each product to the corresponding machine and generating scheduling information including process scheduling information for each machine; and
A multi-constraint greedy job scheduling method characterized in that the job processing unit configured in a machine performing a job for one of a plurality of processes for producing a product includes a process job performing process for performing a corresponding public job according to the corresponding process scheduling information of the scheduling information.
In Article 9,
The above scheduling process is,
A task sequencing step in which the scheduling unit determines the processing order for each mold by sequencing tasks of task information registered in the mold DB according to a pre-set sequencing condition through the task sequencing unit, and stores the sequenced processing order information in the mold DB;
A work scheduling step in which the above scheduling section applies the mold processing order determined through the work scheduling section, the idle time and efficiency of the machine to a greedy algorithm as multiple constraints, and allocates the product-specific process work to the corresponding machine, thereby generating machine-specific process scheduling information according to the machine-specific process scheduling and storing it in the work DB;
A task scheduling step in which the scheduling unit generates scheduling information including process scheduling information for each process through the task scheduling unit and stores it in the scheduling DB; and
A multi-constraint greedy job scheduling method characterized in that it includes a scheduling output step in which the scheduling unit outputs the scheduling information through a scheduling output unit so that the corresponding job processing unit processes the job according to the corresponding process scheduling information of the scheduling information.
In Article 10,
The above task scheduling steps are:
An idle time checking step in which the above scheduling unit checks the idle time of each machine in the machine DB when scheduling a process for a task of an arbitrary mold through the idle time checking unit of the work scheduling unit and outputs the checked idle time;
An actual work time calculation step in which the scheduling unit calculates and outputs the actual work time for the basic work time of the process work assigned to the corresponding machine by reflecting the efficiency of each machine in the machine DB through the task process work time calculation unit of the work scheduling unit;
The above scheduling unit applies the mold processing order, actual working time, and idle time of each machine as multiple constraints to a greedy algorithm through the process work allocation unit of the work scheduling unit to allocate process work to the machine, thereby generating process scheduling information for each machine according to the process scheduling of each machine and storing it in the work DB; and
A multi-constraint greedy task scheduling method characterized in that the scheduling unit includes an idle time updating step of updating the idle time of a machine to which the process task is assigned in the machine DB to a time after the actual working time of the process task according to the assignment of the process task after the process task is assigned to an arbitrary machine through the idle time checking unit.
In Article 11,
The progress confirmation unit further includes a progress confirmation process in which the progress of the process work in progress is confirmed according to the work time allocated to each machine through the work processing unit, and the progress confirmation process provides the progress information and work time to the scheduling unit.
The above scheduling process is,
A multi-constraint greedy work scheduling method characterized in that the scheduling unit further includes a process progress confirmation step in which the scheduling unit receives the progress work time and progress information from the progress confirmation unit, calculates the efficiency of the corresponding machine according to the progress work time and progress information, and updates the efficiency of the corresponding machine in the machine DB.
In Article 11,
The above task process work time calculation section,
A multi-constraint greedy work scheduling method characterized by calculating the actual work time by the following mathematical formula.
[Mathematical formula]

Here, t' _j,A is the time added according to efficiency, p' _bh,j is the standard working time for j operation of h (mold) assigned to machine A, and e _j,A is the efficiency of machine A.
In Article 10,
The above task sequencing unit,
A multi-constraint greedy job scheduling method characterized by sequencing the order of mold tasks and the importance applied to at least one of the order work price, mold task delivery speed, and delivery days.