JP4589164B2 - PCB mounting optimization method for component mounters - Google Patents

Description

  The present invention relates to a method for optimizing board production of a component mounting machine, and more specifically, in a mounting line in which a plurality of component mounting machines are connected, the allocation of a plurality of production programs for producing boards to the mounting line is optimized. It is about the method.

  In an electronic component mounter that mounts electronic components (hereinafter simply referred to as components) on a circuit board, board production (component mounting) is performed by creating a production program for producing the board for each board type. Each production program includes various data for producing a board on a mounting machine, for example, data about a board, data about a mounting position, data about a part (for example, height and width dimensions), data about a suction position, image recognition, etc. Information, data on application of adhesive, and the like.

  Conventionally, production programs have been optimized in order to improve substrate production efficiency. For example, the order of component adsorption and mounting has been optimized to shorten production tact.

  In addition, the production tact time is also shortened by linearizing the path through which the head unit that picks up the component moves from the component supply unit to a predetermined position on the circuit board (Patent Document 1).

In addition, multiple random arrays of feeders that supply parts are generated, and each of these arrays is made into an individual.Each individual is evaluated using a genetic algorithm, and each individual is evaluated. As an optimal feeder arrangement, component mounting is also optimized (Patent Documents 2 and 3).
JP 2001-94295 A JP 10-209681 A JP 2000-261190 A

  Normally, board production is performed on a mounting line in which multiple component mounters are connected, and the conventional method for optimizing a production program is that the user selects only one type of mounting line, and the selected mounting line On the other hand, one production program or a plurality of production programs are optimized. However, for users with multiple mounting lines, there is no way to automatically determine which production program is to be produced efficiently on which mounting line, and board production can be performed on multiple mounting lines. When performing, there is a problem that efficient board production cannot be performed unless the line allocation of the production program is optimal.

  Accordingly, an object of the present invention is to provide a method capable of improving substrate production efficiency when a substrate is produced by a plurality of production programs in a plurality of mounting lines.

The present invention
In the board production optimization method of a component mounter that provides a plurality of mounting lines connecting one or more component mounters and assigns a plurality of production programs to each mounting line to produce a board.
For each production program, find the production time required to produce a board using the production program in each mounting line.
Sequentially assign production programs to the mounting line with the shortest cumulative production time on the mounting line to which the production program is assigned,
When optimizing the allocation of production programs to the mounting line by evaluating the total production time on each mounting line after all production programs have been allocated ,
Incorporating the setup time required when different production programs are used in one mounting line, obtaining the cumulative production time,
The setup time is calculated from a preset default setup time and a similarity of feeders of each production program indicating a ratio of parts having the same packing form .

  According to the present invention, when board production is shared by a plurality of mounting lines, the allocation of a plurality of production programs to the plurality of mounting lines can be optimized, so that all the boards produced by the plurality of production programs are finished in production. The production time can be shortened and the productivity of the substrate can be improved.

  In the present invention, a plurality of production programs are collectively optimized for a plurality of mounting lines and assigned to each mounting line. At the time of optimization, considering the operation time of the user and changeover time, the line assignment of each production program is made so that the time until all production programs finish the board production is shortest, and those productions are also made. Schedule the order. Hereinafter, the present invention will be described in detail with reference to the accompanying drawings.

  In FIG. 1, three mounting lines (1) to (3) are provided, and the five production programs (A) to (E) are optimized at once, and the production program is assigned to each mounting line for board production. The state to be performed is illustrated.

  The mounting line (1) includes component mounting machines 10, 11, and 12, the mounting line (2) includes component mounting machines 20, 21, and 22, and the mounting line (3) includes a component mounting machine 30. , 31 and 32 are connected and arranged, and in each mounting line, the board is divided and produced by a plurality of mounting machines. In the example of FIG. 1, the same number of plural (three) component mounters are connected to each mounting line, but there may be a case where only one component mounter is arranged on the mounting line. In some cases, board production is performed by providing different numbers of component mounting machines.

  FIG. 2 illustrates a detailed configuration of the component mounter 10, and the other component mounters 11, 12, 20 to 22, and 30 to 32 have the same configuration.

  The component mounter 10 includes a control unit 41 including a CPU 41a that controls the entire component mounting, a ROM 41c that stores various control programs and data, and a RAM 41b that stores control data and processing data and provides a work area. Yes. Further, the component mounter 10 is provided with a data transmission / reception unit 46 capable of transmitting / receiving data to / from a host computer (not shown), and a production program transmitted from the host computer uses the data transmission / reception unit 46. And is stored in the data storage unit 45. The control unit 41 drives the X / Y drive unit and other drive units 42 in accordance with the production program data transmitted from the host computer and the data input via the data input unit 43, and the suction head (not shown). ) Is moved to the feeder, where the electronic components supplied from the feeder are sucked by the suction head. The sucked component is recognized by the image recognition unit 44 equipped with a camera, and the suction position deviation is corrected. Then, the sucked part is moved to a predetermined position on the board and mounted on the board.

  Each production program (A) to (E) in FIG. 1 includes various data for producing a board on a mounting machine. For example, data relating to a board, data relating to a mounting position, data relating to a component (for example, height and width height) Dimension), data on the suction position, information for image recognition, data on application of adhesive, and the like. One board type is produced according to one production program. As shown in FIG. 1, when a production program is assigned to each mounting line, three component mounting machines 10 are provided on the mounting line (1). , 11 and 12, a plurality of two types of boards are produced according to the production programs (A) and (B), respectively, and in the mounting line (2), the production program (C) is created by three component mounting machines 20, 21 and 22. , (D), a plurality of two types of boards are produced, and in the mounting line (3), a plurality of one type of boards are produced by the three component mounting machines 30, 31, 32 according to the production program (E). The

  In the present invention, the efficiency of board production is improved by optimizing the allocation of a plurality of production programs to a mounting line. First, before execution of optimization, project information necessary for execution of optimization is set by the user. The project information to be set is shown as a table in FIG. 3A. The project information includes “the number of boards scheduled to be produced for each production program”, “setup time”, “operation time”, “each line”. The date and time at which production can be started ”is described on the right side of each item.

  Also, at the time of optimization execution, the user can incorporate control conditions by setting some constraint conditions for the allocation output by optimization. The settable constraint conditions are shown in a table in FIG. 3B. The constraint conditions include “delivery date of each production program”, “inventory status of parts necessary for production and delivery date of parts”, “ There are "context relationship between production programs" and "line designation of each production program", and the contents are explained on the right side of each item.

  Based on the above conditions, allocation of multiple production programs to multiple mounting lines is optimized. This optimization flow is shown in FIG. 4, and a genetic algorithm (GA), which will be described later, is used for generating an allocation pattern.

  First, in steps S1 and S2 in FIG. 4, the information on the mounting line and the information on the production program are read, and then the production time in each mounting line of each production program is calculated and matrixed (steps S3 and S4). . In the present embodiment, this is called a production time matrix, and its contents are shown in FIG. For the calculation of the production time, predetermined optimization software and simulation are used. For production programs that have already been optimized, the simulation time is reduced by executing only the simulation.

  The values in the matrix of FIG. 5 (A) are obtained by multiplying the scheduled production quantity of each production program by one production time. For example, in the mounting line (1), board production is performed according to the production program (A). When it is done, it is calculated as “3” time, which corresponds to the product of the time that the production program (A) produces one board in the mounting line (1) and the total number of produced boards. . In addition, there are combinations that cannot be produced depending on the mounting status of the mounting line. In this case, “production impossible” is marked.

  Next, a matrix of setup times required for feeder rearrangement when each production program is switched is generated (step S5). This is calculated from the similarity of the feeders of each production program (ratio of parts having the same package appearance) and the default setup time set by the user in the project information.

Setup time = default setup time × (1−similarity): similarity is 0.0 to 1.0
The default setup time is 1h (1 hour), the similarity is 1.0 for programs (A) and (B), 0.5 for programs (C) and (D), and 0. A setup time matrix in the case of 0 is shown as a table in FIG. For example, from this chart, when the production programs (A) and (B) are switched, the similarity of the feeders is 1.0, so that feeder rearrangement is unnecessary and the setup time is “0”. I understand that.

Based on the above matrix, FIG. 3 (A), the project reads information and constraints (step S6, S7) of (B), to evaluate the substrate production time due to some scheduling pattern. The scheduling pattern is generated by GA (step S8).

  The general flow of GA and its contents are shown in FIGS. 6 and 7. In the GA process, an initial population is first generated (step S21) and generated in a gene pool 80 of sufficient individuals. The initial value of the chromosome is randomly determined according to the problem specific coding. Subsequently, two individuals to be left in the next generation are determined from the gene pool 80 based on a certain selection criterion (step S22). For this selection criterion, for example, roulette selection, tournament selection, rank method or the like is used. Next, a partial exchange of crossing the genes of the selected parents is performed to generate two children (step S23). This crossover includes one-point crossover, two-point crossover, uniform crossover, and the like. Subsequently, a mutation process is performed to randomly mutate a gene at an arbitrary position for two children obtained by crossover (step S24). This mutation is provided to randomly change the locus of an individual and prevent it from falling into a local solution. As shown in FIG. 7, the next generation gene pool 81 is generated by mutation. Then, an end condition check is performed (step S25), and the process ends when the planned generation has ended or when no improvement in solution has been observed for a certain period of time. If the termination condition is not satisfied, the new gene pool 81 is updated and the process is repeated from step S22.

  In the present invention, a sequence of production programs is expressed as a sequence of genes (chromosomes), and these are subjected to generation change operation by GA to try to improve the solution.

  In the present invention, GA coding is performed by directly coding the sequence of each production program as gene information. That is, as shown in FIG. 8 (A), by performing unique numbering <1> to <5> for each production program (A) to program (E), a gene is assigned to each production program. Codes <1> to <5> are assigned (coding). Then, by generating gene sequences by GA, gene patterns (1), (2),... Having various gene code sequences (production program sequences) as shown in FIG. .. (n) is generated.

  Next, the gene pattern is interpreted and an actual scheduling pattern is generated as a Gantt chart and evaluated (step S8).

  FIG. 9A shows an example in which the setup time does not occur when the order of the production programs is changed, that is, the setup time matrix in FIG. 5B is not considered.

  First, when the gene pattern (1) shown in FIG. 8B is generated, the first gene code of the gene pattern (1) is <3>. Therefore, the third production program (C) is first displayed on the Gantt chart. assign. At that time, the allocation destination is determined from the production time matrix of FIG. 5A to the mounting line where the production program (C) ends most quickly, that is, the mounting line where the production time is the shortest. Since the program (C) is 10 hours for the mounting line (1), 2 hours for the mounting line (2), and 6 hours for the mounting line (3), assign the assignment destination to the mounting line (2) with the shortest production time. decide. When the numbers indicating the order of determination are indicated by numbers in circles, the production programs (A) to (E) are indicated by alphabets attached to the end thereof, and written on the Gantt chart, FIG. 9 (A) Display corresponding to the first two hours of the mounting line (2).

  Next, since the next gene code of the gene pattern (1) is <2>, the second production program (B) is assigned. At that time, the second production program is sequentially assigned to each mounting line, the total production time in each assigned mounting line is obtained, and the second production program ( B) is assigned. The production program (B) cannot be produced on the mounting line (1), the production time on the mounting line (2) is 4 hours, and the production line (3) is 6 hours. The production program (B) is transferred to the mounting line (2). In the case of allocation, the cumulative production time on the mounting line is 2 hours already allocated time + 4 hours = 6 hours, and when allocated to the mounting line (3), the allocation to that line is the first time, The total production time is 6 hours, the same as the production time by the production program (B). As described above, when there are a plurality of lines having the shortest accumulated time, if the mounting line is not yet allocated, the mounting line is allocated to the mounting line, and if not, it is allocated to any of them. In the above example, the mounting line (3) has not been assigned yet, so it is assigned to the mounting line (3). Therefore, in the Gantt chart of FIG. 9A, the number 2 indicating the order of determination is attached within the circle of the mounting line (3), and “B” (production program (B)) is entered for 6 hours. ing.

  Next, the gene code is <4> and is assigned to the fourth production program (D). When this is assigned to the mounting line (1), the accumulated production time is the same as the production time since it is the first assignment to that line, and is one hour. On the other hand, in the case of allocation to the mounting line (2), the allocated time is 2 hours + 8 hours = 10 hours, and production on the mounting line (3) is impossible, and allocation cannot be performed. Therefore, the mounting lines (1) and (2) are assigned to the mounting line (1) with a short accumulated production time.

  Further, when the production program (E) corresponding to the next gene code <5> is assigned to the mounting line (1), the total production time is 1 hour + 12 hours = 13 hours, and the mounting line (2) Since the total production time is 6 hours + 10 hours = 16 hours when assigned to the mounting line (3), the shorter of the mounting lines (1) and (3) Assign to line (1).

  Further, when the production program (A) corresponding to the next gene code <1> is assigned to the mounting line (1), the cumulative production time is 1 hour + 12 hours + 3 hours = 16 hours, and the mounting line ( When assigned to 2), the cumulative production time is 2 hours + 2 hours = 4 hours, and when assigned to the mounting line (3), the cumulative production time is 6 hours + 5 hours = 11 hours. Are assigned to the shortest mounting line (2).

  In this way, when all the production programs are sequentially assigned to the mounting lines in a predetermined order, that is, in order of the gene pattern (1), the scheduling pattern shown in FIG. 9A is generated.

  The production end time of all production programs is adopted as the evaluation value of the gene pattern. This means that when all the production programs are assigned to the mounting lines in a predetermined order, the longest cumulative production time is obtained from the total production times in each mounting line after all production program assignments are completed. When the order of allocation is the gene pattern (1), the cumulative production time in the mounting line (1) is the longest, and that time, 13 hours, is determined as the evaluation value. Hereafter, Gantt charts are created using the same algorithm as gene pattern (1) for the order of gene patterns (2), ... (n) generated by GA, and the totals for each mounting line Evaluate for production time.

  When evaluation values for all gene patterns are obtained in this way, the output result is displayed (step S9), and the Gantt chart with the highest tabular value, that is, the shortest total production time (production by all production programs). Determine the allocation of the production program to the mounting line. For example, if the example of the Gantt chart shown in FIG. 9A has the highest evaluation value, the production program (D), (E) is assigned to the mounting line (1), and the mounting line (2 ) Is assigned production programs (C) and (A), and the production program (B) is assigned to the mounting line (3).

  The above is an example in which no setup is required for feeder arrangement when the production program is replaced in one mounting line. For this reason, an example of optimization when setup time is required will be described.

  For example, when the assignment is performed in the order shown in the gene pattern (1), the first gene code of the gene pattern (1) is <3>, so the third production program (C) is first displayed on the Gantt chart. assign. At that time, the assignment destination is determined from the production time matrix to the mounting line where (C) finishes earliest, and since it is the mounting line (2), it is recorded in the Gantt chart as shown in FIG. 9B. . Next, the second production program (B) is assigned. At this time, the production program (B) cannot be produced on the mounting line (1), is 4 hours on the mounting line (2), and 6 hours on the mounting line (3), but the setup time matrix shown in FIG. Therefore, in the case of allocation to the mounting line (2), the setup time is 1 hour + the production time is 4 hours = 5 hours. Therefore, since the total production time is 2 hours + 5 hours = 7 hours in the mounting line (2) and 6 hours in the mounting line (3), the total production time including the setup time is the shortest in the mounting line (3). Determine the assignment. Similarly, the scheduling shown in FIG. 9B can be obtained by sequentially assigning the production program corresponding to the gene code to the mounting line with the shortest cumulative production time. As the evaluation value of this assignment, the production end time of all production programs, that is, 14 hours in the mounting line (1) having the longest total production time including the setup time among the mounting lines is adopted.

  Next, a Gantt chart is created with the same algorithm as the gene pattern (1) in the order of gene pattern (2) assignment (FIG. 9C).

  Since the first gene code of the gene pattern (2) is <5>, the fifth production program (E) is first assigned to the mounting line (3) with the minimum production time. Since the next gene code is <1>, the first production program (A) is assigned. Since the total production time in the mounting lines (1) to (3) is 3 hours, 2 hours, 10 hours + 5 hours = 15 hours, it is assigned to the shortest mounting line (2). When the production program (C) corresponding to the next gene code <3> is assigned to the mounting lines (1) to (3), the cumulative production time is 10 hours; 2 hours + 1 hour (setup time) +2 hours = 5 Since time: 10 hours + 1 hour (setup time) +6 hours = 17 hours, the cumulative production time is assigned to the smallest mounting line (2). Similarly, when the production programs (D) and (B) are assigned, the Gantt chart shown in FIG. 9C is obtained, and the table value is the mounting line (2), ( 3) 10 hours. Since the evaluation with the smaller table value is higher, the Gantt chart based on the gene pattern (2) is higher than the evaluation based on the gene pattern (1).

  Below, Gantt charts with gene patterns (3), ..... (n) are created, and the Gantt charts with the shortest cumulative production time are compared. In accordance with the above, the allocation of the production program to the mounting line is determined.

On the other hand, when the constraint conditions as shown in FIG. 3B are set, these are taken into account when assigning to the Gantt chart. For example, in the case of the above gene pattern (2),
Constraint (1): Production program (E) must be completed before (A) Constraint (2): Production program (D) has a constraint that production can only be started after 4 hours If so, the assignment of production program (A) must wait for completion of production program (E). Therefore, as shown in FIG. 9D, after the production program (E) is assigned to the mounting line (3), the production time of the production program (A) is 10 hours after the production by the program is finished. The production program (A) is assigned to the smallest mounting line (2).

  The production program (C) corresponding to the next gene code <3> is assigned to the vacant time of the mounting line (2) having the smallest production time, and the production program (D) of the next gene code <4> is assigned. However, according to the constraint condition (2), the assignment to the mounting line (1) having the minimum production time is determined after 4 hours. Similarly, when assignment is determined for all gene codes, a scheduling pattern as shown in FIG. 9D is obtained.

  As for the other constraint conditions, the assignment is executed in consideration of the arrangement in the Gantt chart as in the above example. Here, when scheduling that satisfies the constraint conditions cannot be performed, such a genetic code is regarded as a lethal gene in GA, and the evaluation value of the individual is the worst (maximum value that the evaluation value can take) Operate so that By selecting a lethal gene, only normal solutions remain due to the selection and selection mechanism in GA.

It is a block diagram which shows the structure of the mounting system used by this invention. It is a block diagram which shows the detailed structure of a component mounting machine. (A) is a chart showing project information, and (B) is a chart showing constraints. It is a flowchart which shows the flow of the optimization by this invention. (A) is a chart showing a production time matrix, and (B) is a chart showing a setup time matrix. It is a flowchart which shows the basic flow of a genetic algorithm. It is a flowchart which shows the detailed flow of a genetic algorithm. It is a table | surface explaining the production | generation of a gene pattern. It is a Gantt chart which shows the allocation of the production program by various genetic patterns.

Explanation of symbols

10, 11, 12 Component mounter 20, 21, 22 Component mounter 30, 31, 32 Component mounter

Claims (2)

In a method for optimizing board production of a component mounter that provides a plurality of mounting lines in which one or a plurality of component mounting machines are connected and allocates a plurality of production programs to each mounting line to produce a board.
For each production program, find the production time required to produce a board using the production program in each mounting line.
Sequentially assign production programs to the mounting line with the shortest cumulative production time on the mounting line to which the production program is assigned,
When optimizing the allocation of production programs to mounting lines by evaluating the total production time on each mounting line after all production programs have been allocated ,
Incorporating the setup time required when different production programs are used in one mounting line, obtaining the cumulative production time,
A board production optimization method for a component mounting machine, wherein the setup time is calculated from a preset default setup time and a similarity of feeders of each production program indicating a proportion of parts having the same packaging shape . It allocates considering multiple constraints set by the user, the substrate production method of optimizing component mounting apparatus according to claim 1, characterized in that determining the total production time.