JP3362751B2 - Parts supply instruction method - Google Patents

Description

Description: BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a component supply instruction method for instructing a production line to supply various parts to a production line while leveling their supply amounts. It is. 2. Description of the Related Art When various parts are supplied to a machine assembly process line, the supply amount of each part is leveled.
We can increase the utilization rate of equipment and personnel involved in this,
Several methods or devices have been proposed for this. In the conventional methods disclosed in Japanese Patent Application Laid-Open Nos. 2-224954 and 2-250756, the load of a process and the flow rate of parts are leveled by manipulating the production order of varieties in a mixed production line. It becomes
This is not applicable to medium / mass production lines where lot production takes place. Also, when using conveyors, trolleys, trucks, and other transportation means to supply parts to the machine assembly process, if you try to increase the loading efficiency, by supplying parts that do not yet have demand first, Space is needed to store near the machine assembly line until demand arises. [0004] Now, consider a medium / mass production machine assembly process in which lot production is performed. Here, we proceed with the following assumptions. (1) Each part is contained in a box by a predetermined quantity, and the supply to the assembling process is performed in units of this box. (2) The number of parts accommodated in each box is set to a divisor of one lot. That is, the number of boxes for one lot of each part is an integer. (3) Parts are paid out at regular time intervals.
Parts are supplied from a warehouse to an assembly line using a truck or the like. In order to reduce the inventory on the assembly line side, it is desirable to supply parts immediately before demand arises. Therefore, it is better not to perform advance supply. 10 kinds of parts required for assembling a product, 1
It is assumed that when parts of a lot are paid out in ten times, parts supply as shown in Table 1 is to be performed. [Table 1] [0007] In the case shown in Table 1, the number of dispensed boxes is large every time, such as 10 boxes for the first time and 1 box for the second time, and the operation rate of required equipment (or workers) becomes extremely poor. . This is due to the fact that the demand for parts is concentrated in a certain time zone (first and seventh times). If the demand for parts can be shifted to a supply cycle with fewer boxes, In addition, workload peaks can be suppressed. [0008] Therefore, if some parts are supplied to the assembly line as initial stock in advance before the start of assembly, the demand timing of the parts can be shifted, whereby the supply workload can be leveled. This is shown in Table 2. [Table 2] [0010] In Table 2, those marked with an asterisk indicate that the product will be for the next lot when several lots of this product are continuously assembled. The initial stock is required only when starting to assemble this product. If several lots are to be continuously produced, the first to tenth supply patterns may be repeated. By issuing a component supply instruction based on the component supply timing determined in this way, it is possible to minimize equipment and personnel required for component supply. However, as described above, it is possible to shift the demand timing of parts depending on the initial stock, but how to shift the demand remains as a problem. The number of combinations of how to shift the demand timing is enormous, and an extremely large number of operations are required to find the optimum one among all the combinations. That is, when the number of times of supply is N and there are M types of parts, the number of combinations is N ^M. Table 1,
In the example shown in Table 2, 10 ¹⁰ = 10 billion patterns, and it is not realistic to perform a round robin trial. The present invention has been made in view of the above, and has as its object to provide a component supply instruction method capable of obtaining a high-quality leveling result with a small amount of calculation. [0014] To achieve the above object, a component supply instruction method according to the present invention is directed to a component supply instruction method for supplying various parts to a production line. Calculate the variance of the supply pattern to the line, classify the calculation result of this variance in descending order, calculate the variance of the sum of the supply patterns of the parts in order from the part with the largest variance, and change the phase of the supply pattern of the various parts. By calculating the variance of the sum of the supply patterns of the components with a predetermined shift, and determining the component supply pattern of the minimum value of the variance calculation result, the supply order of the various components to the production line is determined. I have to. [Work] If the pattern of the total number of boxes in each supply cycle is leveled, it is not important which component is shifted. It only concerns which parts are in the initial inventory. Therefore, there is a case where it is not always necessary to shift by the supply pattern from the relation between the supply pattern and the number of component boxes. For example, in Tables 1 and 2, the component 2 is the same as the original supply pattern if it is shifted twice, so that no further trial is necessary. In other words, it is only necessary to try for the period of the component supply pattern. Further, in Tables 1 and 2, the case where only the part 2 is shifted once is the same as the case where only the part 3 is shifted once, and the total number of boxes is exactly the same. In other words, if the number of boxes for one lot is the same, that is, if there are parts with the same supply pattern cycle, the number of trials can be reduced. Table 3 shows the cycle of each component in this case. [Table 3] Here, assuming that there are m types of components having a period p, the number of trials for the m types of components can be expressed by the following equation. Calculating the total number of trials using the above equation in the example shown in p ^H m Table 3, as follows. 1 × ₂ H ₃ × ₁₀ H ₃ × ₅ H ₂ × 10 = 1 × 4 × 220 × 15 × 10 = 132,000 patterns. However, this depends on the type of component (the number of accommodated components), and the number of trials does not always decrease as described above. The payout pattern and the pattern of the total number of boxes for each component are considered to be discrete time series data. Therefore, "to level the pattern of the total number of boxes for each dispensing time by shifting the component demand timing" means, as shown in FIG. 1, that "the waveform represented by the dispensing pattern of each component is shifted in phase. Are synthesized and the vibration energy of the synthesized waveform, that is, the variance is minimized. " This is called a sequential variance minimization method. Therefore, a simple calculation method is adopted in which component supply patterns are taken out one by one and synthesized so that the variance of the pattern of the total number of boxes is minimized. However, an important point at this time is that "the components are selected and synthesized in descending order of the supply pattern of the components". If the variance of the last remaining components is large, the variance of the total number of boxes is large, and there is a high possibility that sufficient leveling cannot be performed. FIG. 2 is a flowchart showing the procedure of the sequential variance minimization. FIG. 3 shows a configuration of an apparatus for carrying out the above method, which comprises an information storage unit, a component supply timing generation unit, and a component supply instruction output unit. Embodiments Specific embodiments of the present invention will be described below. By the above calculation method, the number of trials is represented by the sum of the cycles of the supply pattern of all parts. In the example shown in Table 1, Equation 1 is obtained. [Formula 1] Table 4 shows the result of leveling the pattern shown in Table 1 by this method. [Table 4] In Table 4, *, **, *** have the same variance value, and therefore may be started from any of the same marks. In the above embodiment, as shown in Table 4, the optimum value (the one with the smallest variance of the pattern of the total number of boxes) was obtained, but the successive variance minimization method does not always provide the optimum value. However, since a high-quality leveling result can be obtained with a small amount of calculation, an extremely effective method is used when a strictly optimum leveling result is not required, or when the number of parts supplied and the number of parts are large. is there. Next, the procedure for performing leveling will be described. By the above-described calculation method, the number of trials is represented by the sum of the supply pattern periods of all components. When the number of trials is calculated in an example as shown in Table 3 having ten types of component configurations, the number of trials is as shown in Expression 1 above, and there are 52 types. Now, the procedure up to this leveling will be described. (1) Determining the Number of Parts Supply The appropriate number of supplies is determined from the time required to supply the parts once, the production time for one lot, and the total number of parts (boxes) that can be supplied at one time. In this case, it is assumed that parts for one lot are supplied in ten times. (2) Determining the payout pattern of each component The number of boxes required for each component is divided by the number of times of supply to calculate the payout timing. For example, in the case of the supply number: 10 times and the parts 5: 4 boxes / lot, as shown in the column of the parts 5 in Table 3, the first time, the third time, the sixth time, and the eighth time,
It may be supplied once every 2.5 times. Other parts are also supplied in the patterns shown in the respective columns. (3) Calculation of supply pattern variance for each component The variation of the supply pattern obtained in (2) above, that is, the variance is calculated. The variance is represented by Equation 2. [Equation 2] When the variance of the component 5 described above is obtained, the average value = 4/10 = 0.4 variance = (1−0.4) ² + (0−0.4) ² + (1−
0.4) ² + (0−0.4) ² + (0−0.4) ² +
(1−0.4) ² + (0−0.4) ² + (1−0.4)
² + (0-0.4) Table If shows a ^{2 + (0-0.4) = 0.35 ×} 4 + 0.16 × 6 = 2.4 results obtained dispersion for each component in this way The result is the same as that shown in FIG. (4) Select the two with the largest variance. The two with the largest variance found in (3) above are selected.
Referring to Table 5, it can be seen that they are part 2 and part 3. As described in Table 4, the components 2, 3,
Any two of the parts 4 can be selected. Here, parts 2 and 3 are selected. [Table 5] (5) Calculation of variance of combined pattern The supply pattern of the selected two components is combined for each supply count. Component 2 1 0 1 0 1 0 1 1 0 1 0 Component 3 1 0 1 0 1 0 1 0 1 1 0 Combined result 2 0 2 0 2 0 2 0 2 0 Dispersion is calculated from the combined result and stored. Average value = 1 Dispersion = 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 = 10 (6) Shifting the pattern, shifting the phase of the pattern of the component 2 (or component 3) by one and synthesizing it in the manner described in (5) above. Component 2 1 0 1 0 1 0 1 0 1 0 Component 3 0 1 0 1 0 1 0 1 0 1 1 Synthetic result 1 1 1 1 1 1 1 1 1 1 1 1 (7) Repeat shifting by the pattern period The patterns are synthesized by being shifted one by one by the period of the pattern, and the variance of the synthesis result is obtained. Since the cycle of the part 3 is 2, it is sufficient to execute the first synthesis and the synthesis shifted by one, a total of two times.
It ends with the above (5) and (6). If this is 3, the first synthesis, the synthesis with the phase of the pattern shifted by one,
The variance of a total of three synthesis results is obtained by shifting the phase of the pattern by two. (8) Find a composite pattern with the minimum variance. From all the synthesis results, the one with the smallest variance, that is, the leveled pattern is selected. In the examples so far, when the component 3 is shifted by one, the synthesis result is in a leveled state. (9) Selection and Combination of Next Part Next, from the remaining parts, that is, the parts which have not been combined yet, the one having the largest variance is selected, and the result of the leveled combination up to the previous time is selected. Combine with Looking at Table 4, it can be seen that this is part 4. Combined result up to the previous time 1 1 1 1 1 1 1 1 1 1 1 1 Part 4 1 0 1 0 1 0 1 0 1 0 1 0 Combined result 2 1 2 1 2 1 2 1 2 1 When variance is calculated in the same manner, average value = 1 0.5 dispersion = 0.5 ² × 10 = 2.5. (10) Pattern Shift The pattern of the component 4 is shifted and synthesized in the same manner as in the above (6). Combined results up to the previous time 1 1 1 1 1 1 1 1 1 1 1 1 Parts 4 0 1 0 1 0 1 0 1 0 1 0 1 Combined results 1 2 1 2 1 2 1 2 1 2 When the variance is similarly calculated, the average value is 1. 5 min dispersion = 0.5 ² × 10 = 2.5 and heals. (11) Repetition of shifting by the period of the pattern Repeating one by one by the period of the pattern in the same manner as in the above item (7), and synthesizing the result of the synthesis. (12) Obtaining a Composite Pattern with the Minimum Variance A pattern with the minimum variance is selected in the same manner as in the above item (8). Hereinafter, the above (8)-(1) is applied to all remaining parts.
By repeating 2), the component supply pattern is leveled. The results are shown in Table 6. [Table 6] According to the present invention, a high-quality leveling result can be obtained with a small amount of computation by using the successive variance minimization method, and a strictly optimum leveling result is not required. This is an effective method when the number of parts supplied and the number of types of parts are large. By using this method, various parts are supplied to the production line for assembling machines, etc. Without increasing the number of parts, the flow rate of parts can be leveled, and equipment and personnel for supplying parts can be suppressed.

BRIEF DESCRIPTION OF THE DRAWINGS FIG. 1 is a schematic diagram for explaining a successive variance minimization method. FIG. 2 is a flowchart showing a procedure for calculating leveling. FIG. 3 is an explanatory view showing the concept of an apparatus for carrying out the method of the present invention.

────────────────────────────────────────────────── ─── Continuation of the front page (56) References JP-A-7-299681 (JP, A) JP-A-3-234449 (JP, A) (58) Fields investigated (Int. Cl. ⁷ , DB name) B23P 21/00 307 B23Q 41/08

Claims (1)

(57) [Claim 1] In a parts supply instruction method for supplying various parts to a production line, a variance of a supply pattern of the various parts to the production line is calculated, and the variance is calculated. The results are sorted in descending order, the variance of the sum of the supply patterns of the components is calculated in order from the component having the largest variance, and the variance of the sum of the supply patterns of the components is calculated by shifting the phase of the supply pattern of the various components by a predetermined amount. A component supply instruction method characterized by determining the supply order of the various components to the production line by determining the component supply pattern of the minimum value of the dispersion calculation result.