JP7221830B2 - Manufacturing support device, manufacturing support method and program - Google Patents

Description

The present invention relates to a manufacturing support apparatus, a manufacturing support method, and a program.

When manufacturing products or intermediate products in the manufacturing industry, it is often the case that manufacturing lots are created in each manufacturing process, and the manufacturing order within the manufacturing lots is determined before manufacturing. A production lot is a process that is assembled when manufacturing multiple materials (including raw materials, parts, and products) that are grouped into a certain economical and efficient scale according to the production flow and are manufactured in a common process. Units. The processing unit is determined by balancing the order of processing priority and the amount of in-process inventory, etc., for the purpose of reducing variations in quality and improving productivity and reducing costs by reducing set-up work for manufacturing equipment.

In addition, the order of production is determined based on production standards that define the conditions to be observed during production for purposes such as ensuring quality and improving productivity. For example, in a factory in the manufacturing industry, when manufacturing products or intermediate products, the manufacturing order is determined so that the changes in the attributes (quality, properties, etc.) of the products or intermediate products manufactured continuously within a certain lot are small. do. At this time, the quality is ensured by, for example, setting the limit values of changes in attributes that should be observed by continuously manufactured products or intermediate products. The same applies to the determination of manufacturing lots. Limit values for attributes that cannot be grouped together in the same lot are set, and attributes (quality, properties, etc.) that products or intermediate products manufactured in the same lot must protect are specified. By keeping it within the range, it is possible to improve productivity, reduce costs, and reduce variations in quality.

On the other hand, Patent Literature 1 describes a technique for automatically determining the manufacturing order by defining an objective function and constraints and formulating an optimization problem.

JP 2013-151021 A

However, in many cases, a product or an intermediate product has multiple attributes, each of which has constraints, and consideration is given to the priority among constraints, the difference between absolute constraints and soft constraints, etc. Therefore, it is not easy to formulate such that an appropriate manufacturing order and manufacturing lot can be automatically determined as described in Patent Document 1. In addition, in many cases, there are evaluation indexes and constraints held as know-how by the operator, and it is difficult to consider all of these to perform appropriate formulation and determine parameters.

Therefore, the present invention provides a cost function that makes it possible to statistically process the feature amount from the manufacturing results of products or intermediate products, and to determine an appropriate manufacturing schedule (manufacturing sequence or manufacturing lot) based on the statistically processed feature amount. It is an object of the present invention to provide a manufacturing support device, a manufacturing support method, and a program for creating In this specification, the “cost function” is a kind of evaluation function for judging the quality of the manufacturing schedule, and when the manufacturing schedule planning problem is treated as a minimization problem, can be identified as the optimal manufacturing schedule.

According to one aspect of the present invention, for at least two products or intermediate products having a predetermined temporal relationship in manufacturing performance, a performance data aggregating unit for aggregating relationships of attribute values of at least two products or intermediate products. and a cost function determination unit that determines a cost function relating to the relationship of attribute values based on the frequency of occurrence of the relationship of attribute values.

According to another aspect of the present invention, for at least two products or intermediate products having a predetermined temporal relationship in manufacturing performance, aggregating relationships of attribute values of the at least two products or intermediate products; determining a cost function for the attribute value relationships based on the frequency of occurrence of the value relationships.

According to yet another aspect of the present invention, for at least two products or intermediate products having a predetermined temporal relationship in manufacturing performance, performance data aggregation for aggregating relationships of attribute values of at least two products or intermediate products. and a cost function determination unit that determines a cost function related to the relationship of attribute values based on the frequency of occurrence of the relationship of attribute values.

According to the above configuration, the cost function is appropriately determined based on the frequency of occurrence of the relationships of the aggregated attribute values, and the appropriate manufacturing schedule (manufacturing sequence or manufacturing lot) is determined using this cost function. be able to.

It is a figure which shows the example of the constraint conditions in a manufacturing schedule. 2 is a diagram showing an example of a cost function that expresses the constraint conditions shown in FIG. 1; FIG. 4 is a flow chart showing an example of processing at a cost function setting stage of the manufacturing support method according to one embodiment of the present invention; It is a figure which shows the example of a production record file. It is a figure which shows the example of frequency distribution of the relationship of an attribute value. 6 is a diagram showing an example of a cost function determined from the frequency distribution shown in FIG. 5; FIG. FIG. 10 is a diagram showing another example of the frequency distribution of relationships between attribute values; FIG. 8 is a diagram showing an example of a cost function determined after correcting the frequency distribution shown in FIG. 7; FIG. 10 illustrates an example of modifying the cost function considering absolute constraints and dead zones; FIG. 4 is a diagram for explaining a method of creating a cost table from a cost function; FIG. 4 is a flow chart showing an example of processing in a manufacturing schedule determination stage of a manufacturing support method according to an embodiment of the present invention; 1 is a diagram showing a schematic configuration of a manufacturing support apparatus according to one embodiment of the present invention; FIG.

Preferred embodiments of the present invention will be described in detail below with reference to the accompanying drawings. In the present specification and drawings, constituent elements having substantially the same functional configuration are denoted by the same reference numerals, thereby omitting redundant description.

FIG. 1 is a diagram showing an example of constraints in a manufacturing schedule. FIG. 1 conceptually shows the manufacturing order of a product and the values of attribute 1 and attribute 2 of each product. Attribute 1 and Attribute 2 are, for example, dimensions of the product, and it is preferable that Attribute 1 has a smaller difference between the products that follow it in the manufacturing order. In addition, when a difference occurs for attribute 1, the difference of "-" which is smaller in the product after the product before than the difference of "+" which is larger in the product after the product than in the product before is preferred. As a result, the manufacturing order of the products is determined so that the products having no difference in attribute 1 are consecutive and the difference of "-" is more than the difference of "+".

FIG. 2 is a diagram showing an example of a cost function that expresses the constraint conditions shown in FIG. As described above, for the relationship of attribute 1, it is preferable that the difference between the preceding and succeeding products in the manufacturing order is small. Therefore, in the illustrated example, if the difference of attribute 1 is 0 in the cost table, it becomes "0", and a cost value that increases as the absolute value of the difference increases is set. Furthermore, if a difference occurs for attribute 1, the difference of "-" is preferable to the difference of "+", so the cost value is set larger in the "+" range than in the "-" range. be done. The manufacturing sequence can be optimized for the Attribute 1 constraint by minimizing the sum of the costs incurred by the difference in Attribute 1 values for two products or intermediate products that are manufactured in sequential order. In other words, the manufacturing order can be optimized by setting appropriate cost values in the cost table. The same applies to manufacturing schedules other than the manufacturing order, such as determination of manufacturing lots.

However, determining a manufacturing schedule based on cost values expressing constraints is not easy in practice. In many cases, a product (or intermediate product) has multiple attributes. In the example shown in FIGS. 1 and 2, the product also has attribute 2 which is different from attribute 1. FIG. In the attribute 1 cost table illustrated in FIG. 2, attribute 2 is incorporated as an element that affects attribute 1 cost. Specifically, the larger the attribute 2 value, the smaller the cost value for the attribute 1 difference. In such a case, the cost table is stratified (four layers in the example of FIG. 2) by the value of attribute 2. FIG. Furthermore, as shown in the example of FIG. 1, attribute 2 may also have a constraint condition that it is preferable that the difference between the preceding and succeeding products in the manufacturing order is small. In this case, a cost table such as that illustrated in FIG. 2 is created for attribute 2 as well. When the number of attributes is large and the attributes influence each other, it is not easy to set the optimum cost value in the cost table.

FIG. 3 is a flow chart showing an example of processing in the cost function setting stage of the manufacturing support method according to one embodiment of the present invention. First, in the performance data input step (S110), the manufacturing performance file 1 is read. The production record file 1 is an example of information indicating attribute values of at least two products or intermediate products having a predetermined temporal relationship in the production record. In the example shown in FIG. 4, the production record file 1 includes a product ID, attribute values (attributes 1 to 3 in the illustrated example, but may be more), production lot ID, and Contains information on the manufacturing sequence within a manufacturing lot. For example, a manufacturing lot ID is information indicating that at least two products or intermediate products were manufactured within the same time frame. Also, the manufacturing order within a manufacturing lot is information indicating that at least two products or intermediate products were manufactured in consecutive order. Thus, as used herein, if at least two products or intermediate products are manufactured in consecutive order or within the same time frame in a manufacturing experience, then these products or intermediate products are defined as It is also said to have a temporal relationship.

Next, in the stratified condition determination step (S120), stratified conditions for the cost function of each attribute are determined from the actual values of the attributes indicated by the manufacturing performance file 1 read in step S110. Here, the stratification condition refers to a condition for stipulating how to stratify (classify) the cost function relating to the relationship of certain attribute values according to other attribute values. For example, as in the example shown in FIG. 2, when multiple attribute values are defined for a product or an intermediate product, a cost table that stratifies a cost function relating to the relationship of the first attribute value according to the second attribute value. may be configured (in the illustrated example, the second attribute value is stratified according to which value of "2", "3", "5", and "10"). In step S120, for example, the stratification conditions defined in existing manufacturing standards may be imported, or the stratification conditions may be extracted by specifying the correlation of the actual values of attributes by cluster analysis or the like. . When determining the stratification conditions for the cost table of attribute 1 in the example shown in FIG. 4, either or both of attributes 2 and 3 are candidates. For example, if there is a correlation between the performance values of attribute 1 and the performance values of attribute 2, a more appropriate cost function can be defined by stratifying the cost function of attribute 1 according to attribute 2.

Next, in the performance data totaling step (S130), according to the stratification conditions determined in step S120, the feature values from the manufacturing performance file 1 read in step S110, that is, the relationship of the attribute values indicated by the manufacturing performance file 1 Aggregate (statistically process) the characteristics. For example, if the manufacturing order, that is, the order in which products or intermediate products are manufactured, the differences in the attribute values of the products or intermediate products manufactured in consecutive order in the manufacturing results may be aggregated. In the example shown in FIGS. 1 and 2 above, with respect to attribute 1, differences in attribute values between the preceding and succeeding products are aggregated. In addition, when targeting the manufacturing lot, that is, the time frame in which products or intermediate products are manufactured, the distribution state represented by the distribution of the attribute values of the products or intermediate products manufactured within the same time frame in the manufacturing results may be aggregated. The aggregation calculates the frequency of occurrence of attribute value relationships of at least two products or intermediate products having a predetermined temporal relationship. The occurrence frequency is represented by a frequency distribution as shown in FIG. 5, for example. As in the constraint condition of attribute 1 shown in FIG. 2 above, when a smaller difference in attribute values is preferable, the peak of the frequency distribution is near the difference 0, and the frequency increases as the absolute value of the difference increases. descend. To simplify the explanation, one attribute value has been described here, but it can be extended to a multidimensional space. Specifically, after performing stratification, differences in multiple attributes are taken for products or intermediate products that have temporal relationships, and by treating them as frequency distributions in a multidimensional space, expansion to multiple dimensions is also possible. It is possible. At this time, by stratifying, it is possible to deal with the case where the probability distribution differs greatly in each layer and it is difficult to handle with one function.

Next, in a cost function determination step (S140), a cost function related to attribute value relationships is determined based on the feature amount aggregated (statistically processed) in step S130, that is, the frequency of occurrence of attribute value relationships. For example, considering the frequency distribution as shown in FIG. 5 as a probability density function, calculating the inverse function of the probability density function as shown in FIG. may be Statistical processing and numerical calculation can be facilitated by converting to a probability density function. Here, in many cases, products (or intermediate products) are manufactured according to orders, and the attributes of ordered products are biased. There may be a dip in the frequency distribution that is not reflected. In such a case, as shown in FIG. 8, after performing correction (for example, linear interpolation) to fill the drop in the frequency distribution by user operation or by automatic determination using the differential value of the probability density function may determine the cost function. Further, in the cost function determination step S140, the cost function may be stratified according to the stratification conditions determined in the stratification condition determination step S120.

Then, preferably, in a cost function modification step (S150), the cost function determined in step S140 is modified. For example, as shown in FIG. 9, if there is an absolute constraint condition determined from facility constraints, etc., the cost You can also maximize the value of the function. This makes it possible to express soft constraints that should be observed as much as possible within a range that satisfies absolute constraints. On the other hand, the value of the cost function may be leveled in the vicinity of the range where the cost function is minimized. In the illustrated example, the values of the cost function are leveled in the neighboring regions where the attribute value difference is zero. The range in which the values of the cost function are leveled becomes a dead zone, and the relationship between attributes can be arbitrarily set within this range. As a result, when the cost function for a certain attribute is within the dead band, priority is given to the cost functions for other attributes, and a reasonable manufacturing schedule can be determined.

Through the above cost function correction step (S150), a cost function for the relationship of each attribute is output. In this embodiment, the cost function is output as a cost table 2 stratified according to the stratification condition determined by the stratification condition determination step (S120). Several methods are conceivable for creating the cost table. As shown in FIG. 10, it is conceivable to express by averaging the values of the cost function in an arbitrary section. Representing the cost function as a cost table improves readability, facilitates comparison with conventionally used standards, and facilitates manual correction.

FIG. 11 is a flow chart showing an example of processing in the manufacturing schedule determination stage of the manufacturing support method according to one embodiment of the present invention. Although an example of determining the order of manufacturing will be described below as an example of the manufacturing schedule, the same applies to determining the manufacturing lot. In the illustrated example, first, the manufacturing schedule file 3 is read in the schedule data input step (S160). The manufacturing schedule file 3 includes information such as product IDs of products (or intermediate products) scheduled to be manufactured and attribute values.

Next, in the manufacturing order determination step (S170), the cost function for the relationship of each attribute is read from the cost table 2, and based on the cost function, the manufacturing schedule of the products to be manufactured included in the manufacturing schedule file 3, specifically determines the order of production. The manufacturing order determined by scheduling is recorded in the manufacturing order file 4 . Here, for scheduling, a known optimization technique, specifically, mixed integer programming, local search, simulated annealing, or the like, can be used. In any optimization method, the manufacturing order can be determined by searching for the manufacturing order that minimizes the total value of this cost function.

When determining the manufacturing plan, specifically the manufacturing sequence, using the optimization method, the optimization parameters reflect the evaluation indexes and constraints that the workers have in terms of know-how, which do not appear in the visualized manufacturing standard. It is not easy to let Therefore, in this embodiment, the feature amount is statistically processed from the manufacturing results by the manufacturing schedule determination process as described above, and the optimization parameter, specifically the cost function, is calculated from the statistically processed feature amount. It makes it possible to determine a more appropriate manufacturing sequence.

FIG. 12 is a diagram showing a schematic configuration of a manufacturing support apparatus according to one embodiment of the invention. The manufacturing support apparatus 10 shown in FIG. 12 is configured by a large computer such as a mainframe or an open system, or an information processing apparatus such as a personal computer. The production record file 1, the cost table 2, the production schedule file 3, and the production order file 4 are stored in various memories such as ROM and RAM, hard disks, and recording media such as CD-ROM. Alternatively, part or all of these data may be stored in another information processing apparatus and transmitted to the manufacturing support apparatus 10 by communication using an electric communication line such as a LAN or the Internet.

In the manufacturing support apparatus 10, the performance data input unit 11 executes the performance data input step S110 of the manufacturing support method described above. Similarly, the stratified condition determination unit 12 executes the stratified condition determination step S120, the performance data aggregation unit 13 executes the performance data aggregation step S130, and the cost function determination unit 14 executes the cost function determination step S140. , the cost function correction unit 15 executes a cost function correction step S150. The cost function correction unit 15 may automatically correct the cost function as described above, or may execute the correction according to a user's operation acquired via the input/output device 5 . The scheduled data input unit 21 executes the scheduled data input step S160 of the manufacturing support method, and the manufacturing order determination unit 22 executes the manufacturing order determination step S170. It should be noted that, as described above, the example of the manufacturing schedule is not limited to the manufacturing order, and the manufacturing lot may be determined in the same manner. The functions of the above units are realized by the processor of the information processing device (computer) operating according to the program stored in the memory or recording medium.

Note that each part of the manufacturing support apparatus 10 does not need to be implemented by a single information processing apparatus, and may be implemented by being distributed among a plurality of information processing apparatuses. Specifically, for example, a performance data input unit 11, a stratified condition determination unit 12, a performance data aggregation unit 13, a cost function determination unit 14, a cost function correction unit 15, a schedule data input unit 21, and a manufacturing order determination unit 22 may be implemented by different information processing devices.

Although the preferred embodiments of the present invention have been described in detail above with reference to the accompanying drawings, the present invention is not limited to such examples. It is obvious that a person having ordinary knowledge in the technical field to which the present invention belongs can conceive of various modifications or modifications within the scope of the technical idea described in the claims. It is understood that these also naturally belong to the technical scope of the present invention.

Reference Signs List 1 production record file, 2 cost table, 3 production schedule file, 4 production sequence file, 5 input/output device, 10 production support device, 11 performance data input unit, 12 stratified condition determination unit, 13 -- Actual data totalization section, 14 -- Cost function determination section, 15 -- Cost function correction section, 21 -- Planned data input section, 22 -- Manufacturing order determination section.

Claims (10)

a performance data aggregating unit for aggregating relationships between attribute values of at least two products or intermediate products having a predetermined temporal relationship in manufacturing performance;
A manufacturing support apparatus comprising: a cost function determination unit that determines a cost function related to the attribute value relationship based on the frequency of occurrence of the attribute value relationship. 2. The manufacturing support apparatus according to claim 1, wherein said predetermined temporal relationship includes that said at least two products or intermediate products are manufactured in sequential order. 3. The manufacturing support apparatus according to claim 2, wherein said relationship of attribute values includes a difference in attribute values of products or intermediate products manufactured in said consecutive order. 4. The manufacturing support apparatus according to any one of claims 1 to 3, wherein said predetermined temporal relationship includes that said at least two products or intermediate products are manufactured within the same time frame. 5. The manufacturing support apparatus according to claim 4, wherein said relationship of attribute values includes a distribution state of attribute values of products or intermediate products manufactured within said same time frame. 6. The manufacture according to any one of claims 1 to 5, further comprising a cost function correction unit that maximizes the value of the cost function within a range in which the relationship of the attribute values does not satisfy absolute constraints. support equipment. 7. The manufacturing support apparatus according to any one of claims 1 to 6, further comprising a cost function correction unit that levels the value of the cost function in the vicinity of a range in which the cost function is minimized. first attribute values and second attribute values are defined for the at least two products or intermediate products;
The performance data aggregation unit aggregates relationships of the first attribute values,
wherein the cost function determining unit stratifies the cost function regarding the relationship of the first attribute values according to the second attribute value of at least one product or intermediate product among the two products or intermediate products; The manufacturing support apparatus according to any one of claims 1 to 7. for at least two products or intermediate products having a predetermined temporal relationship in manufacturing performance, aggregating relationships of attribute values of the at least two products or intermediate products;
determining a cost function for the attribute value relationship based on the frequency of occurrence of the attribute value relationship. a performance data aggregating unit for aggregating relationships between attribute values of at least two products or intermediate products having a predetermined temporal relationship in manufacturing performance;
A program for causing a computer to function as a manufacturing support device, comprising: a cost function determination unit that determines a cost function related to the attribute value relationship based on the frequency of occurrence of the attribute value relationship.

Images