JP2009054766A - Monitoring method, monitoring device and monitoring program of manufacturing process, and manufacturing method of semiconductor device - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a monitoring method of a manufacturing process, capable of automatically detecting an executing time of maintenance of a manufacturing device. <P>SOLUTION: Wavelet transform signal are obtained for a plurality of frequency bands by executing wavelet transform to a measurement signal output from an LSI manufacturing device (S3). When the wavelet transform signals for al the frequency bands are not smaller than a threshold value, it is determined that the maintenance of the manufacturing device has been executed, and in the cases other than that, it is determined that the maintenance has not been executed (S4). When it is determined that the maintenance has been executed, an alarm is not issued (S5), and a new management range is set based on a measurement signal output after the maintenance (S6). When it is determined that the maintenance has not been executed, whether the measurement signal is within the management range is determined (S7), and it is determined that abnormality has occurred in the manufacturing device when the measurement device is not within the management range, and an alarm is issued (S8). <P>COPYRIGHT: (C)2009,JPO&INPIT

Description

  The present invention relates to a manufacturing process monitoring method, a monitoring apparatus, a monitoring program, and a semiconductor device manufacturing method, and in particular, a manufacturing process for monitoring occurrence of an abnormality in a manufacturing apparatus based on a measurement signal output from a product manufacturing apparatus. The present invention relates to a monitoring method, a monitoring apparatus, a monitoring program, and a semiconductor device manufacturing method to which the monitoring method is applied.

  2. Description of the Related Art Conventionally, in a semiconductor device manufacturing line, occurrence of an abnormality in each manufacturing apparatus is monitored based on a measurement signal output from each manufacturing apparatus. For example, the average value μ and standard deviation σ are obtained from the history of the measurement signal, the value of (μ ± 3σ) is set as the control limit, and when the measurement signal exceeds the control limit, an alarm is issued uniformly and manufactured. The apparatus is inspected (for example, refer to Patent Document 1).

  However, depending on the manufacturing apparatus, even when maintenance such as replacement of parts or cleaning is performed, the measurement signal may fluctuate and exceed the control limit. In the conventional method described above, an alarm is issued even in such a case, and as a result, there is a problem that unnecessary device inspection is performed.

  As a workaround for this problem, a method may be considered in which a maintenance person communicates information on the timing of performing maintenance in each manufacturing apparatus to an apparatus inspection person. However, in the semiconductor device manufacturing line, the number of manufacturing apparatuses is enormous, hundreds to thousands, and it is difficult for the person in charge of apparatus inspection to grasp in real time that the maintenance is performed by this method.

JP 2007-65883 A

  An object of the present invention is to provide a manufacturing process monitoring method, a monitoring apparatus and a monitoring program, and a semiconductor device manufacturing method capable of automatically detecting the maintenance execution time of the manufacturing apparatus.

  According to one aspect of the present invention, a wavelet transform is performed on a measurement signal output from a product manufacturing apparatus to obtain a wavelet transform signal in a plurality of frequency bands, and the wavelet transform signal in the plurality of frequency bands. If a certain number or more of the wavelet transform signals are equal to or greater than a threshold value, it is determined that maintenance of the manufacturing apparatus has been performed, and otherwise, it is determined that maintenance of the manufacturing apparatus has not been performed. A maintenance determination step, and an abnormality detection step for determining whether or not the measurement signal is within a management range. When it is determined that the maintenance has been performed, an alarm is not issued and the maintenance is performed after the maintenance. A new management range is set based on the output measurement signal and it is determined that the maintenance is not performed. When the measurement signal is within the control range, the manufacturing apparatus determines that the manufacturing apparatus is normal, does not issue an alarm, and determines that the maintenance is not performed. When the measurement signal is not within the control range, it is determined that an abnormality has occurred in the manufacturing apparatus, and an alarm is issued, and a manufacturing process monitoring method is provided.

  According to another aspect of the present invention, there is provided a method for manufacturing a semiconductor device, wherein the product is a semiconductor device, and the manufacturing process is monitored by the method described above.

  According to still another aspect of the present invention, a determination unit that determines whether an abnormality has occurred in the manufacturing apparatus based on a measurement signal output from the product manufacturing apparatus, and the determination unit includes the manufacturing unit An alarm means for issuing an alarm when it is determined that an abnormality has occurred in the apparatus, the determination means performs wavelet transform on the measurement signal, and outputs wavelet transform signals in a plurality of frequency bands. Determining that when a certain number or more of the wavelet transform signals of the plurality of frequency bands are equal to or greater than a threshold value, it is determined that maintenance of the manufacturing apparatus has been performed; otherwise, the manufacturing is performed. When it is determined that the apparatus is not maintained, it is determined whether the measurement signal is within a management range, and when it is determined that the maintenance is performed, When a new management range is set based on the measurement signal output after tenancy and it is determined that the maintenance is not performed, and the measurement signal is within the management range, When it is determined that the manufacturing apparatus is normal and the maintenance is not performed and the measurement signal is not within the management range, it is determined that an abnormality has occurred in the manufacturing apparatus. A manufacturing process monitoring device is provided.

  According to still another aspect of the present invention, a computer performs a wavelet transform on a measurement signal output from a product manufacturing apparatus, and obtains a wavelet transform signal of a plurality of frequency bands, If a certain number or more of the wavelet transform signals in the frequency band are equal to or greater than the threshold value, it is determined that the manufacturing apparatus has been maintained. In other cases, the manufacturing apparatus is maintained. A maintenance determination procedure for determining that the maintenance signal has not been received, and an SPC procedure for determining whether or not the measurement signal is within a management range. If it is determined that the maintenance has been performed, an alarm is issued. The maintenance range is set based on the measurement signal output after the maintenance, and the maintenance is performed. If it is determined that the measurement signal is within the control range, the manufacturing apparatus is determined to be normal and no alarm is issued, and the maintenance is being performed. If it is determined that the measurement signal is not within the control range, it is determined that an abnormality has occurred in the manufacturing apparatus and an alarm is issued. Is provided.

  According to the present invention, it is possible to realize a manufacturing process monitoring method, a monitoring apparatus and a monitoring program, and a semiconductor device manufacturing method capable of automatically detecting the maintenance execution time of the manufacturing apparatus.

Hereinafter, embodiments of the present invention will be described with reference to the drawings.
FIG. 1 is a block diagram illustrating a manufacturing process monitoring system according to this embodiment.
As shown in FIG. 1, in the monitoring system 1 according to the present embodiment, a manufacturing apparatus 2 for manufacturing a product is provided. In the present specification, the term “product” refers to a member after a predetermined process is performed in the manufacturing apparatus, and includes an intermediate product in addition to a so-called finished product. The manufacturing apparatus 2 is an apparatus constituting a part of a manufacturing line for manufacturing a semiconductor device such as a large scale integrated circuit (LSI). The manufacturing apparatus 2 measures processing conditions performed by the manufacturing apparatus 2, for example, temperature or pressure in the chamber at regular intervals, and outputs these measured values as measurement signals.

  The monitoring system 1 is provided with a monitoring device 3. The monitoring device 3 is a manufacturing process monitoring device that monitors whether or not an abnormality has occurred in the LSI manufacturing process performed by the manufacturing device 2. The monitoring device 3 is provided with a determination unit 4 that receives the measurement signal output from the manufacturing device 2 and determines whether or not an abnormality has occurred in the manufacturing device 2 based on the measurement signal. In addition, the monitoring device 3 is provided with a notification unit 5 that issues an alarm when the determination unit 4 determines that an abnormality has occurred in the manufacturing apparatus 2. For convenience of explanation, only one manufacturing apparatus 2 is shown in FIG. 1, but the monitoring system 1 includes a plurality of manufacturing apparatuses 2, and one monitoring apparatus 3 includes a plurality of manufacturing apparatuses 2. The manufacturing apparatus 2 may be monitored.

  Furthermore, the monitoring system 1 is provided with a terminal computer 6. The terminal computer 6 is used by a person in charge of equipment inspection of the manufacturing apparatus 2. The terminal computer 6 receives an alarm output from the monitoring device 3 and displays it when an alarm is input.

Next, an operation of the monitoring apparatus according to the present embodiment configured as described above, that is, a manufacturing process monitoring method according to the present embodiment will be described.
FIGS. 2A to 2H are graphs illustrating the history of the measurement signal or the change of the wavelet transform signal corresponding to the history of the measurement signal, with time on the horizontal axis and the measurement signal or its wavelet transform signal on the vertical axis. It is.
The vertical axis of FIG. 2 (a) represents the measurement signal indicating the pressure, the vertical axis of (b) represents the high-frequency wavelet transform signal, the vertical axis of (c) represents the medium-frequency wavelet transform signal, The vertical axis of (d) represents the low-frequency wavelet transform signal. The vertical axis in FIG. 2 (e) represents the measurement signal indicating temperature, the vertical axis in (f) represents the high-frequency wavelet transform signal, and the vertical axis in (g) represents the medium-frequency wavelet transform signal. The vertical axis of (h) represents the low-frequency wavelet transform signal. In each diagram of FIG. 2, the values on the horizontal axis (time axis) are common. Further, FIG. 2 shows only wavelet transform signals in three frequency bands of “high frequency”, “medium frequency”, and “low frequency”, but the number of frequency bands is not limited to three, but four. It may be the above.
FIG. 3 is a flowchart illustrating the manufacturing process monitoring method according to this embodiment.

As shown in FIGS. 2A and 2E, the manufacturing apparatus 2 receives a measurement signal P (t) indicating the pressure in the chamber at time t and a measurement signal T indicating the temperature in the chamber at time t. (T) is output to the monitoring device 3. The value of time t takes discrete values arranged at regular intervals, for example, takes a value every second. A time t ₀ shown in FIG. 2 is a time when maintenance is performed on the manufacturing apparatus 2. The maintenance is, for example, part replacement or cleaning of the manufacturing apparatus 2.

As shown in FIGS. 2A and 2E, in the period before the maintenance execution timing, that is, in the period of t <t ₀ , the measurement signals P (t) and T (t) It is approximately within a predetermined range centered on the average value. For example, when the average value of the measurement signal P (t) is μ _P and the standard deviation is σ _P , the measurement signal P (t) is in the range of (μ _P −3σ _P ) to (μ _P + 3σ _P ). Almost settled. Therefore, the monitoring device 3 sets this range as a management range, and performs SPC (Statistical Process Control). That is, when the measurement signal is out of the management range, it is determined that an abnormality has occurred in the manufacturing apparatus 2 and an alarm is issued to the terminal computer 6.

However, when maintenance is performed on the manufacturing apparatus 2 at time t ₀ , the measurement signals P (t) and T (t) are large, discontinuously, and irreversibly change. In the period after the maintenance execution timing, that is, in the period of t ₀ <t, the measurement signals P (t) and T (t) are each stabilized within a new range. In general, the range in which the measurement signal is stable in the period after maintenance is significantly different from the range in which the measurement signal was stable in the period before maintenance. Accordingly, when performing continued SPC after maintenance, it is determined that abnormality in the manufacturing apparatus 2 at the maintenance time t ₀ has occurred, an alarm will be alarm. However, since no abnormality has actually occurred in the manufacturing apparatus 2 and no inspection is required, this alarm is false.

  Therefore, in the present embodiment, when the measurement signal fluctuates, it is determined whether or not this is due to maintenance, and when it is determined that it is due to maintenance, an alarm is not issued. In addition, a new management range is set based on the measurement signal after maintenance, and thereafter, SPC is performed based on the new management range.

Whether or not the fluctuation of the measurement signal is caused by maintenance is determined by wavelet transforming the measurement signal. As shown in FIG. 2 (a) and (e), the measurement signal changes greatly and discontinuously at time _{t 0,} as shown in FIG. 2 (b) ~ (d) , (f) ~ (h) In addition, the wavelet transform signals of these measurement signals take high values at all frequencies. Therefore, it can be estimated that maintenance has been performed when all the wavelet transform signals have values greater than or equal to the threshold.

Hereinafter, a specific example of an algorithm for realizing the monitoring method according to the above-described embodiment, that is, the monitoring method for performing SPC after determining the presence or absence of maintenance will be described.
In the monitoring system 1 shown in FIG. 1, the manufacturing apparatus 2 continuously manufactures products, sequentially measures the pressure and temperature in the chamber, and obtains a measurement signal P (t) and temperature representing the pressure in the chamber at time t. The expressed measurement signal T (t) is sequentially output to the monitoring device 3. In this state, the monitoring device 3 operates as follows.

First, as shown in step S1 of FIG. 3, from the history of past measurement signals P (t) and T (t), the average value μ _P and standard deviation σ _{P of} the measurement signal P (t) and the measurement signal T It calculates an average value mu _T and standard deviation sigma _T of (t). Based on the average value and the standard deviation, the management range of the measurement signal is set. For example, the management range of the measurement signal P (t) is (μ _P −3σ _P ) to (μ _P + 3σ _P ), and the management range of the measurement signal T (t) is (μ _T −3σ _T ) to (μ _T + 3σ). _T ). The measurement signal P (t) is a wavelet transform, to determine the threshold alpha _P based on the obtained wavelet transform signal. Further, the measurement signal T (t) is subjected to wavelet transform, and a threshold value α _T is determined based on the obtained wavelet transform signal. At this time, when the maintenance of the manufacturing apparatus 2 is performed, the threshold values α _P and α _T are equal to or greater than a certain number, for example, wavelet transform signals for all frequencies exceed the threshold values α _P and α _T. Value.

  In this state, as shown in step S2 of FIG. 3 and FIGS. 2A and 2E, the measurement signal P (t) and the measurement signal T at time t are read from the manufacturing apparatus 2.

Next, as shown in step S3, the wavelet transform is sequentially performed on the measurement signals P (t) and T (t) read during a certain period until time t, and FIGS. , (F) to (h), wavelet transform signals W _P (t, f) and W _T (t, f) in a plurality of frequency bands are obtained. W _P (t, f) is a wavelet transform signal of frequency f obtained for the measurement signal P (t) up to time t, and W _T (t, f) is the measurement signal T (t up to time t. ) Is a wavelet transform signal of frequency f obtained for.

Next, as shown in step S4, it is determined whether or not the manufacturing apparatus 2 has been maintained. Specifically, the wavelet transform signal W _P (t, f) is compared with the threshold α _P, and the wavelet transform signal W _T (t, f) is compared with the threshold α _T. Then, as shown in FIG. 2, among a plurality of frequency band wavelet transform signals W _P (t, f), a certain number or more of wavelet transform signals are equal to or greater than a threshold value α _P and a plurality of frequency band wavelets. It is determined that the maintenance of the manufacturing apparatus 2 has been performed at time t when a certain number or more of the wavelet transform signals among the transform signals W _T (t, f) are greater than or equal to the threshold value α _T. For example, the wavelet transform signal W _P (t, f) is greater than or equal to α _{P for} all frequencies f, and the wavelet transform signal W _T (t, f) is greater than or equal to α _T for all frequencies f. In some cases, it is determined that maintenance has been performed. On the other hand, in cases other than the above, it is determined that maintenance of the manufacturing apparatus 2 is not performed.

In step S4, while if it is determined that the maintenance of the production apparatus 2 is performed, as shown in step S5, that several points counted from the time t ₀ when maintenance was performed is measured, i.e., the time t ₀ Until several unit time elapses, no alarm is issued even if the measurement signal exceeds the control limit. As shown in step S6, the average values μ _P and μ _T and the standard deviations σ _P and σ _T are obtained for the measurement signals P (t) and T (t) output after the maintenance execution time t _0. calculate. Thus, a new management range is set based on the measurement signal output after the maintenance. Thereafter, the process proceeds to step S9.

On the other hand, if it is determined in step S4 that maintenance has not been performed, it is determined whether or not the measurement signal is within the management range, as shown in step S7. That is,
| P (t) −μ _P |> 3σ _P
Or | T (t) −μ _T |> 3σ _T
If so, it is determined that an abnormality has occurred in the manufacturing apparatus 2, and otherwise, it is determined that the manufacturing apparatus 2 is normal.

  If it is determined in step S7 that an abnormality has occurred in the manufacturing apparatus 2, the process proceeds to step S8 and an alarm is issued. For example, the monitoring device 3 transmits a warning mail to the terminal computer 6. This alerts the person in charge of inspection of the manufacturing apparatus 2. Thereafter, the process proceeds to step S9.

  On the other hand, if it is determined in step S7 that the manufacturing apparatus 2 is normal, the process proceeds to step S9 without passing through step S8, that is, without issuing an alarm.

  When monitoring the manufacturing apparatus 2 is continued, the process proceeds from step S9 to step 10, the value of time t is increased by 1, and the process returns to step S2. Thereafter, the above steps are repeated. Through such steps, a semiconductor device, for example, an LSI can be manufactured. That is, the manufacturing method of the semiconductor device according to the present embodiment is a manufacturing method in which the manufacturing process is monitored by the above-described method.

Next, the effect of this embodiment will be described.
As described above, when the maintenance is performed on the manufacturing apparatus, the measurement signal may be large and discontinuously change before and after the maintenance. In such a case, the wavelet transform signal of the measurement signal shows a high value in the entire frequency band at the time when the measurement signal suddenly changes. When this manufacturing apparatus outputs a plurality of types of measurement signals, all measurement signals change simultaneously. On the other hand, when an abnormality occurs in the manufacturing apparatus, in many cases, the measurement signal changes continuously or suddenly, and immediately returns to the original level. In addition, when a plurality of types of measurement signals are output, only some of the measurement signals change. For this reason, when an abnormality occurs in the manufacturing apparatus, the wavelet transform signals of all frequency bands of all measurement signals do not take high values at the same time.

  Therefore, in the present embodiment, it is possible to estimate whether or not maintenance has been performed by checking whether all wavelet transform signals or a certain number of wavelet transform signals are equal to or greater than a threshold value. And, only when the maintenance is not performed and the measurement signal is out of the management range, the alarm is issued to prevent the generation of the false information due to the maintenance, and the manufacturing apparatus is abnormal. Only when the alarm can be issued accurately. As described above, according to the present embodiment, it is possible to automatically detect the maintenance execution time of the manufacturing apparatus based on the measurement signal, and to make the occurrence probability of the false alarm constant regardless of the presence or absence of the maintenance. Can do. As a result, the semiconductor device can be manufactured efficiently.

  The determination unit 4 and the reporting unit 5 of the monitoring device 3 may be configured by dedicated hardware, but the monitoring device 3 executes a series of monitoring operations shown in FIG. 3, for example, a program Can also be implemented. In this case, the monitoring device 3 has a built-in computer, and the monitoring device 3 stores a program for causing the computer to execute the following procedures (1) to (3).

(1) A conversion procedure (step S3) in which wavelet transform is performed on the measurement signal output from the manufacturing apparatus 2 to obtain wavelet transform signals W (f, t) in a plurality of frequency bands.
(2) If a certain number or more of the wavelet transform signals W (f, t) in a plurality of frequency bands are equal to or greater than the threshold value α, it is determined that the maintenance of the manufacturing apparatus 2 has been performed, and otherwise In the case of, a maintenance determination procedure for determining that maintenance of the manufacturing apparatus 2 is not performed (step S4).
(3) SPC procedure for determining whether or not the measurement signal is within the management range (step 7).

  When this program determines that the maintenance of the manufacturing apparatus 2 has been performed on the computer, the program does not issue an alarm (step S5), and a new management range based on the measurement signal output after the maintenance. Is set (step S6), and when it is determined that maintenance is not performed and the measurement signal is within the control range, the manufacturing apparatus 2 is determined to be normal and an alarm is issued. If it is determined that maintenance has not been performed and the measurement signal is not within the management range, the manufacturing apparatus 2 is determined to have an abnormality and an alarm is issued (step S8).

  Further, when there are a plurality of types of measurement signals, for example, a measurement signal P (t) indicating pressure and a measurement signal T (t) indicating temperature, this program is stored in the computer of the monitoring device 3 with the above-mentioned maintenance. In the determination procedure, for all types of measurement signals, when a certain number or more of wavelet transform signals W (f, t) in a plurality of frequency bands, for example, all wavelet transform signals are greater than or equal to a threshold value α, the manufacturing apparatus 2 It is determined that maintenance has been performed.

  Further, the program may cause the computer of the monitoring device 3 to further execute a procedure for calculating the average value and the standard deviation of the measurement signal, and determine the management range based on the average value and the standard deviation.

  The present invention has been described above with reference to the embodiment. However, the present invention is not limited to this embodiment. For example, as long as a person skilled in the art adds, deletes, or changes a design, or adds, deletes, or changes a process as appropriate to the above-described embodiment, the gist of the present invention is included. , Within the scope of the present invention.

  For example, in the above-described embodiment, an example in which the manufacturing apparatus outputs the measurement signal P (t) indicating the pressure and the measurement signal T (t) indicating the temperature is shown, but the present invention is not limited to this. That is, the manufacturing apparatus may output one type or three or more types of measurement signals. When the manufacturing apparatus outputs a plurality of types of measurement signals, in the step of determining the presence or absence of maintenance (step S4 in FIG. 3), for all types of wavelet transform signals W, a certain number or more of wavelet transform signals. Is greater than or equal to the threshold value α, it is determined that maintenance of the manufacturing apparatus has been performed, and in the process of determining abnormality of the manufacturing apparatus (step S7 in FIG. 3), a certain number of one or more types of wavelet transform signals W are determined. What is necessary is just to determine with abnormality having occurred in the manufacturing apparatus, when the above wavelet transformation signal remove | deviates from the management range.

  In the above-described embodiment, the example in which the management range of the measurement signal is set to the range of (μ ± 3σ) is shown, but the present invention is not limited to this, and the management range includes the characteristics of the manufacturing process, the manufacturing apparatus What is necessary is just to set suitably according to the performance of this, the tendency of the dispersion | variation in a measurement signal, etc.

  Furthermore, in the above-described embodiment, an example in which the monitoring device 3 outputs an alarm to the terminal computer 6 has been shown. However, the present invention is not limited to this, and for example, the monitoring device 3 outputs an alarm by voice output or a screen. It may be displayed or printed out by a printing device.

  Furthermore, in the above-described embodiment, the example in which the monitoring device 3 monitors the LSI manufacturing process has been described. However, the present invention is not limited to this, and the present invention is applicable to any product manufacturing process. Can do.

It is a block diagram which illustrates the monitoring system of the manufacturing process which concerns on embodiment of this invention. (A)-(h) is a graph which takes the time on a horizontal axis, takes a measurement signal or its wavelet transform signal on a vertical axis, and shows the history of a measurement signal or the change of the wavelet transform signal corresponding to it. . It is a flowchart figure which illustrates the monitoring method of the manufacturing process which concerns on this embodiment.

Explanation of symbols

1 monitoring system, 2 manufacturing device, 3 monitoring device, 4 judging means, 5 reporting means, 6 terminal computer

Claims (10)

A wavelet transform is performed on the measurement signal output from the product manufacturing device, and a wavelet transform signal for a plurality of frequency bands is obtained.
When the wavelet transform signals of a certain number or more among the wavelet transform signals of the plurality of frequency bands are greater than or equal to a threshold value, it is determined that maintenance of the manufacturing apparatus has been performed, and otherwise, the manufacturing apparatus A maintenance determination step for determining that maintenance has not been performed;
An abnormality detection step of determining whether or not the measurement signal is within a management range;
With
If it is determined that the maintenance has been performed, an alarm is not issued, and a new management range is set based on the measurement signal output after the maintenance,
When it is determined that the maintenance has not been performed, and the measurement signal is within the control range, the manufacturing apparatus determines that it is normal and does not issue an alarm.
When it is determined that the maintenance has not been performed and the measurement signal is not within the control range, it is determined that an abnormality has occurred in the manufacturing apparatus and an alarm is issued. To monitor the manufacturing process. There are multiple types of measurement signals,
In the maintenance determination step, for all types of the measurement signals, it is determined that the maintenance has been performed when a predetermined number or more of wavelet transform signals of the plurality of frequency bands are equal to or greater than the threshold value. The manufacturing process monitoring method according to claim 1, wherein: Further comprising calculating an average value and a standard deviation of the measurement signal,
3. The manufacturing process monitoring method according to claim 1, wherein the management range is determined based on the average value and the standard deviation.   A manufacturing method of a semiconductor device, wherein the product is a semiconductor device, and the manufacturing process is monitored by the method according to claim 1. Based on the measurement signal output from the product manufacturing apparatus, a determination means for determining whether an abnormality has occurred in the manufacturing apparatus;
When the determination unit determines that an abnormality has occurred in the manufacturing apparatus, a notification unit that issues an alarm;
With
The determination means includes
Wavelet transform is performed on the measurement signal to obtain a wavelet transform signal of a plurality of frequency bands,
When the wavelet transform signals of a certain number or more among the wavelet transform signals of the plurality of frequency bands are equal to or greater than a threshold, it is determined that the manufacturing apparatus has been maintained, and otherwise, the manufacturing apparatus Determine that maintenance is not being performed,
Determining whether the measurement signal is within a control range;
If it is determined that the maintenance has been performed, a new management range is set based on the measurement signal output after the maintenance,
If it is determined that the maintenance has not been performed and the measurement signal is within the control range, the manufacturing apparatus is determined to be normal,
A manufacturing process monitoring device characterized in that, when it is determined that the maintenance is not performed and the measurement signal is not within the control range, it is determined that an abnormality has occurred in the manufacturing device. . There are multiple types of measurement signals,
The determination means determines that the maintenance has been performed for all types of measurement signals when a certain number or more of wavelet transform signals of the plurality of frequency bands are equal to or greater than the threshold value. The manufacturing process monitoring apparatus according to claim 5. The determination means includes
Calculate the average value and standard deviation of the measurement signal,
7. The manufacturing process monitoring apparatus according to claim 5, wherein the management range is determined based on the average value and the standard deviation. On the computer,
A wavelet transform is performed on the measurement signal output from the product manufacturing equipment, and a conversion procedure for obtaining a wavelet transform signal of a plurality of frequency bands,
When the wavelet transform signals of a certain number or more among the wavelet transform signals of the plurality of frequency bands are greater than or equal to a threshold value, it is determined that maintenance of the manufacturing apparatus has been performed, and otherwise, the manufacturing apparatus A maintenance determination procedure for determining that maintenance has not been performed;
An SPC procedure for determining whether the measurement signal is within a management range;
And execute
If it is determined that the maintenance has been performed, an alarm is not issued, and a new management range is set based on the measurement signal output after the maintenance,
When it is determined that the maintenance has not been performed, and the measurement signal is within the management range, the manufacturing apparatus is determined to be normal and does not issue an alarm.
When it is determined that the maintenance is not performed, and the measurement signal is not within the control range, it is determined that an abnormality has occurred in the manufacturing apparatus and an alarm is issued. Manufacturing process monitoring program. There are multiple types of measurement signals,
On the computer,
In the maintenance determination procedure, for all types of the measurement signals, when a certain number or more of wavelet transform signals of the plurality of frequency bands are equal to or greater than the threshold, it is determined that the maintenance has been performed. 9. The manufacturing process monitoring program according to claim 8, wherein: On the computer,
Further executing a procedure for calculating an average value and a standard deviation of the measurement signal,
The manufacturing process monitoring program according to claim 8 or 9, wherein the management range is determined based on the average value and the standard deviation.

Images