Data processing method and storage medium

Abstract

A data processing method and a storage medium are provided. The data processing method includes a step of acquiring time-series data, a step of acquiring evaluation values, a classification step, and an extraction step. In the step of acquiring time-series data, a plurality of time-series data obtained by a substrate processing apparatus is acquired. In the step of acquiring evaluation values, evaluation values of the plurality of time-series data are acquired, respectively. In the classification step, the plurality of time-series data is classified into any one of a plurality of categories, respectively, based on the evaluation values. In the extraction step, time-series data corresponding to any one of the plurality of categories is extracted as extracted time-series data.

Description

TECHNICAL FIELD

[0001] The present application relates to a data processing method and a storage medium. BACKGROUND

[0002] A substrate processing apparatus that processes a substrate is known. Typically, the substrate processing apparatus is preferably used for processing of a semiconductor substrate. It is being studied to determine an abnormality of the substrate processing apparatus based on data outputted in time series from the substrate processing apparatus (for example, refer to Patent Literature 1). In the semiconductor manufacturing apparatus of Patent Literature 1, a two-axis coordinate system is made based on correlation data of a specific monitoring object and other monitoring objects associated therewith, and an abnormality of the semiconductor manufacturing apparatus is determined based on whether or not it is included in an abnormality region.

[0003] [Related Art Documents]

[0004] [Patent Literature]

[0005] [Patent Literature 1] Japanese Patent Laid-Open No. 2006-228911 SUMMARY

[0006] [Problems to be Solved by the Invention]

[0007] However, in the method of Patent Literature 1, it is necessary to determine two monitoring objects having a desired correlation, but in practice, the determination is difficult, and it is not easy to determine the state of the semiconductor manufacturing apparatus.

[0008] The present application has been achieved in view of the above-described problems, and aims to provide a data processing method and a storage medium that easily extract a specific time series data from a plurality of time series data generated in a substrate processing apparatus. Furthermore, another object of the present application is to provide a data processing method and a storage medium that can reduce the amount of computation when matching an object time series data obtained from a substrate processing apparatus and a time series data contained in a learning database.

[0009] [Means of Solving the Problems]

[0010] According to an aspect of the present application, a data processing method includes a step of acquiring time series data, a step of acquiring an evaluation value, a classification step, and an extraction step. In the step of acquiring time series data, a plurality of time series data obtained by a substrate processing apparatus is acquired. In the step of acquiring an evaluation value, an evaluation value of each of the plurality of time series data is acquired. In the classification step, the plurality of time series data is classified into any one of a plurality of categories based on the evaluation value. In the extraction step, time series data corresponding to any one of the plurality of categories is extracted as extracted time series data.

[0011] In one embodiment, in the step of acquiring the evaluation value, the evaluation value is acquired by comparing each of the plurality of time-series data with the reference data.

[0012] In one embodiment, in the step of acquiring the evaluation value, the evaluation value in a case where a difference between the time-series data and a value of the reference data is large is larger than the evaluation value in a case where the difference is small.

[0013] In one embodiment, the data processing method further includes a step of switching display of an evaluation value chart and the extracted time-series data, the evaluation value chart indicating a change in the evaluation value of the extracted time-series data.

[0014] In one embodiment, the data processing method further includes a step of storing the extracted time-series data.

[0015] In one embodiment, the data processing method further includes a step of performing clustering processing on the extracted time-series data to classify into any one of a plurality of clusters.

[0016] In one embodiment, the data processing method further includes a step of switching display of an evaluation value chart and the extracted time-series data, the evaluation value chart indicating an evaluation value of the extracted time-series data classified into any one of the plurality of clusters.

[0017] In one embodiment, the data processing method further includes a step of generating a learning database, the learning database being a cluster corresponding to the extracted time-series data, to which cause countermeasure information on the substrate processing apparatus corresponding to the time-series data is given.

[0018] Further, according to another aspect of the present application, a data processing method includes a step of acquiring time-series data, a step of acquiring an evaluation value, a classification step, and a matching step. In the step of acquiring the time-series data, a plurality of time-series data obtained by a substrate processing apparatus is acquired. In the step of acquiring the evaluation value, an evaluation value of each of the plurality of time-series data is acquired. In the classification step, the plurality of time-series data is classified into any one of a plurality of categories based on the evaluation value. In the matching step, an object time-series data corresponding to any one of the plurality of categories and time-series data included in a learning database are matched.

[0019] In one embodiment, in the learning database, cause and countermeasure information corresponding to the time series data is stored. The data processing method further includes, in the matching step, a step of reading out the cause and countermeasure information corresponding to the time series data when the matching rate of the object time series data and at least one of the time series data included in the learning database is higher than a threshold value.

[0020] In one embodiment, in the step of acquiring the evaluation value, the evaluation value of each of the plurality of time series data is acquired by comparing each of the plurality of time series data with the reference data.

[0021] In one embodiment, in the step of acquiring the evaluation value, the evaluation value in a case where the difference between the time series data and the value of the reference data is large is larger than the evaluation value in a case where the difference between the time series data and the value of the reference data is small.

[0022] In one embodiment, the data processing method further includes a step of storing the object time series data.

[0023] According to another aspect of the present application, a data processing apparatus includes a data acquisition section, an evaluation value acquisition section, a classification section, and an extraction section. The data acquisition section acquires a plurality of time series data obtained by a substrate processing apparatus. The evaluation value acquisition section acquires an evaluation value of each of the plurality of time series data. The classification section classifies the plurality of time series data into any one of a plurality of categories based on the evaluation value. The extraction section extracts time series data corresponding to any one of the plurality of categories as extracted time series data.

[0024] According to another aspect of the present application, a data processing apparatus includes a data acquisition section, an evaluation value acquisition section, a classification section, and a matching section. The data acquisition section acquires a plurality of time series data obtained by a substrate processing apparatus. The evaluation value acquisition section acquires an evaluation value of each of the plurality of time series data. The classification section classifies the plurality of time series data into any one of a plurality of categories based on the evaluation value. The matching section matches object time series data corresponding to any one of the plurality of categories and time series data included in a learning database.

[0025] According to a further aspect of the present application, a storage medium stores a program that causes a computer to execute a step of acquiring time series data, a step of acquiring evaluation values, a classification step, and an extraction step. In the step of acquiring time series data, a plurality of time series data obtained by a substrate processing apparatus is acquired. In the step of acquiring evaluation values, evaluation values of the plurality of time series data are acquired. In the classification step, the plurality of time series data is classified into any one of a plurality of categories based on the evaluation values. In the extraction step, time series data corresponding to any one of the plurality of categories is extracted as extracted time series data.

[0026] According to a further aspect of the present application, a storage medium stores a program that causes a computer to execute a step of acquiring time series data, a step of acquiring evaluation values, a classification step, and a matching step. In the step of acquiring time series data, a plurality of time series data obtained by a substrate processing apparatus is acquired. In the step of acquiring evaluation values, evaluation values of the plurality of time series data are acquired. In the classification step, the plurality of time series data is classified into any one of a plurality of categories based on the evaluation values. In the matching step, object time series data corresponding to any one of the plurality of categories and time series data contained in a learning database are matched.

[0027] [Effects of the Invention]

[0028] According to the present application, a specific time series data can be easily extracted from a plurality of time series data generated in a substrate processing apparatus. Furthermore, according to the present application, a computation amount at the time of matching object time series data obtained by a substrate processing apparatus and time series data contained in a learning database can be reduced. BRIEF DESCRIPTION OF DRAWINGS

[0029] Figure 1 is a schematic view of a data processing apparatus and a substrate processing apparatus of the present embodiment.

[0030] Figure 2 is a flowchart of a data processing method of the present embodiment.

[0031] Fig. 3(a) is a graph showing time series data used in the data processing method of the present embodiment, and Fig. 3(b) is a graph showing reference data used in the data processing method of the present embodiment.

[0032] Fig. 4(a) is a schematic view showing time series data used in the data processing method of the present embodiment, Fig. 4(b) is a schematic view showing a comparison between time series data and reference data in the data processing method of the present embodiment, and Fig. 4(c) is a schematic view showing a change in evaluation values in the data processing method of the present embodiment.

[0033] Fig. 5(a) is a schematic view showing classification based on evaluation values in the data processing method of the present embodiment, and Fig. 5(b) is a schematic view showing extraction of time series data in the data processing method of the present embodiment.

[0034] Figure 6 is a schematic view of the data processing apparatus and the substrate processing apparatus of the present embodiment.

[0035] Figures 7(a) to 7(c) is a schematic view showing an evaluation value chart showing changes in evaluation values and time series data displayed in the display section in the data processing apparatus of the present embodiment.

[0036] Figures 8(a) to 8(c) is a schematic view showing time series data displayed in the display section in the data processing apparatus of the present embodiment.

[0037] Figures 9(a) to 9(c) is a schematic view showing an extracted evaluation value chart and corresponding time series data displayed in the display section in the data processing apparatus of the present embodiment.

[0038] Figure 10 is a schematic view of the data processing apparatus and the substrate processing apparatus of the present embodiment.

[0039] Figures 11(a) to 11(c) is a schematic view showing time series data extracted in the data processing method of the present embodiment.

[0040] Figure 12 is a flowchart of the data processing method of the present embodiment.

[0041] Figure 13 is a flowchart of the data processing method of the present embodiment.

[0042] Figure 14 is a schematic view showing a table of a learning database in the data processing apparatus of the present embodiment.

[0043] Figure 15 is a schematic view of the data processing apparatus and the substrate processing apparatus of the present embodiment.

[0044] Figure 16 is a flowchart of the data processing method of the present embodiment.

[0045] Figure 17 is a flowchart of the data processing method of the present embodiment.

[0046] Fig. 18(a) is a graph showing a waveform of a control signal, and Fig. 18(b) is a graph showing a waveform of time series data.

[0047] Figure 19 is a diagram representing values of a plurality of time series data.

[0048] Fig. 20(a) is a graph representing a distribution of evaluation values, and Fig. 20(b) is a graph representing a result of standardization of the distribution of evaluation values.

[0049] Figure 21 is a diagram of a data processing apparatus and a substrate processing apparatus of the present embodiment.

[0050] Figure 22 is a diagram of a data processing apparatus and a substrate processing apparatus of the present embodiment.

[0051] Figure 23 is a diagram of a data processing apparatus and a substrate processing apparatus of the present embodiment.

[0052] Figure 24 is a diagram of a data processing apparatus and a substrate processing apparatus of the present embodiment.

[0053] Figure 25 is a diagram of a data processing apparatus and a substrate processing apparatus of the present embodiment.

[0054] [Explanation of symbols]

[0055] 10, 110: processing section

[0056] 11: data acquisition section (acquisition section)

[0057] 12: evaluation value acquisition section (acquisition section)

[0058] 13, 113: classification section

[0059] 14: extraction section

[0060] 15: clustering processing section

[0061] 20, 120, 201b: storage section

[0062] 30, 140: display section

[0063] 32: display screen

[0064] 33: display area

[0065] 34: operation area

[0066] 34a: arrow

[0067] 34b: knob

[0068] 40, 130: input section

[0069] 100: data processing apparatus

[0070] 111: data acquisition section

[0071] 112: evaluation value acquisition section

[0072] 114: matching section

[0073] 115: readout section

[0074] 200, 200A, 200B: substrate processing apparatus

[0075] 201: control device

[0076] 201a: control section

[0077] 210: processing unit

[0078] 220: chamber

[0079] 230, 270: substrate holding section

[0080] 231: rotary base

[0081] 232: chuck member

[0082] 233: shaft

[0083] 234: electric motor

[0084] 235: base

[0085] 240: liquid supply section

[0086] 240b: valve

[0087] 250: cup section

[0088] 260: processing tank

[0089] A: period (rising period)

[0090] Ax: rotation axis

[0091] B: period (stable period)

[0092] B1 to B3: boundary line

[0093] BG: button

[0094] C: period (falling period)

[0095] CR: center robot

[0096] CU: cursor

[0097] EGE: extraction evaluation value chart

[0098] ETD1, ETD2, ETD3: extraction time series data

[0099] Ev, Ev1 to Ev25: evaluation value

[0100] GE: evaluation value chart

[0101] IR: transfer robot

[0102] L: processing liquid

[0103] L0: initial level

[0104] L1: level 1

[0105] L2: level 2

[0106] L3: level 3

[0107] L4: level 4

[0108] LB: fluid tank

[0109] LC: fluid chamber

[0110] LP: loading port

[0111] R1: first range

[0112] R2: second range

[0113] RD: reference data

[0114] S2 to S24, S102 to S114: step

[0115] t0 to t2: time

[0116] TD, TD1 to TD100: time series data

[0117] Th1, Th2: threshold value

[0118] Th3: prescribed value

[0119] TW: tower

[0120] W: substrate

[0121] Wa: upper surface (surface) of the substrate W

[0122] Wb: back surface (lower surface) of the substrate W

[0123] μ: average value

[0124] σ: standard deviation DETAILED DESCRIPTION

[0125] Embodiments of a data processing method, a data processing apparatus, and a program according to the present application will be described below with reference to the accompanying drawings. In the drawings, like or equivalent components are designated by like reference numerals, and the description thereof will not be repeated.

[0126] First, an embodiment of a data processing apparatus 100 according to the present application will be described with reference to Figure 1 Figure 1 is a schematic view of the data processing apparatus 100 and the substrate processing apparatus 200 according to the present embodiment.

[0127] The data processing apparatus 100 performs data processing. Specifically, the data processing apparatus 100 processes time-series data TD generated in the substrate processing apparatus 200.

[0128] Typically, the substrate processing apparatus 200 includes at least one processing unit 210. The processing units 210 respectively generate time-series data TD.

[0129] The time-series data TD is data indicating a temporal change of a physical quantity in the substrate processing apparatus 200. The time-series data TD indicates a temporal change of a physical quantity (value) that changes in a time series over a prescribed period. For example, the time-series data TD is data indicating a temporal change of a physical quantity with respect to a process performed on a substrate by the substrate processing apparatus 200. Alternatively, the time-series data TD is data indicating a temporal change of a physical quantity with respect to a characteristic of a substrate processed by the substrate processing apparatus 200.

[0130] In addition, the value indicated in the time-series data TD can also be a value directly measured in a measuring device. Alternatively, the value indicated in the time-series data TD can also be a value obtained by performing an arithmetic process on a value directly measured in a measuring device. Alternatively, the value indicated in the time-series data TD can also be a value obtained by performing an arithmetic process on values measured in a plurality of measuring devices.

[0131] Typically, the time-series data TD includes 10 or more values. The time-series data TD can also include 100 or more values, and can also include 1000 or more values.

[0132] The data processing apparatus 100 can also be communicably connected to the substrate processing apparatus 200. Alternatively, the time-series data TD generated in the substrate processing apparatus 200 can also be transmitted to the data processing apparatus 100 via a storage element.

[0133] The data processing apparatus 100 includes a processing section 10. The processing section 10 has a processor. The processing section 10 has, for example, a Central Processing Unit (CPU). Alternatively, the processing section 10 can also have a general-purpose processor.

[0134] ​The data processing apparatus 100 can also include a storage section 20. The storage section 20 stores data and computer programs. The storage section 20 includes a main storage device and an auxiliary storage device. The main storage device is, for example, a semiconductor memory. The auxiliary storage device is, for example, a semiconductor memory and / or a hard disk drive. The storage section 20 can also include removable media. The processing section 10 executes the computer programs stored in the storage section 20.

[0135] Here, the computer programs are stored in a non-transitory computer-readable storage medium. The non-transitory computer-readable storage medium includes a Read Only Memory (ROM), a Random Access Memory (RAM), a Compact Disk-Read Only Memory (CD-ROM), a magnetic tape, a magnetic disk, or an optical data storage device.

[0136] The storage section 20 stores the time series data generated in the substrate processing apparatus 200. The data processing apparatus 100 and the substrate processing apparatus 200 can be communicatively connected to each other, and the time series data TD can be transmitted to the data processing apparatus 100 each time it is generated in the substrate processing apparatus 200, and the storage section 20 sequentially stores the time series data TD. Alternatively, a prescribed number of time series data TD generated in the substrate processing apparatus 200 can be collectively transmitted to the data processing apparatus 100 each time, and the storage section 20 collectively stores the time series data TD.

[0137] The processing section 10 includes a data acquisition section 11, an evaluation value acquisition section 12, a classification section 13, and an extraction section 14. The data acquisition section 11 acquires a plurality of time series data. Here, the data acquisition section 11 acquires the time series data TD from the storage section 20. In this specification, the data acquisition section 11 is sometimes referred to simply as the acquisition section 11.

[0138] The evaluation value acquisition section 12 acquires an evaluation value for each of the plurality of time series data TD. In this specification, the evaluation value acquisition section 12 is sometimes referred to simply as the acquisition section 12.

[0139] The evaluation value acquisition section 12 acquires evaluation values in accordance with the same evaluation criterion for a plurality of time series data TD. The evaluation value acquisition section 12 generates one evaluation value in accordance with a specific evaluation criterion for one time series data TD. For example, the evaluation value acquisition section 12 acquires evaluation values in accordance with the same evaluation criterion for the entirety of the time series data TD. Alternatively, the evaluation value acquisition section 12 acquires evaluation values in accordance with the same evaluation criterion for a specific portion of the time series data TD. In addition, the evaluation value acquisition section 12 can also generate evaluation values by synthesizing a plurality of evaluation criteria using the entirety or a specific portion of the time series data TD.

[0140] Typically, the evaluation value acquisition section 12 generates evaluation values by operating the time series data TD in accordance with the evaluation criterion set in the data processing apparatus 100. For example, the evaluation value acquisition section 12 generates evaluation values by numerically processing the waveform of a graph representing the time series data TD. However, the evaluation value acquisition section 12 can also cause other constituent elements to operate evaluation values and acquire only the operated evaluation values. Evaluation values can also be generated by comparing the time series data TD with data serving as a criterion. In addition, it is preferable that the data serving as a criterion be stored in the storage section 20.

[0141] The number of evaluation values that can be used is less than the number of time series data TD. Typically, the number of time series data TD is 100 or more, and in contrast, the number of evaluation values that can be used is 20 or more and less than 100.

[0142] In addition, there is no particular limitation on the upper limit of the time series data TD. The upper limit of the number of time series data TD can also be arbitrary as long as the data processing apparatus 100 is capable of operating or storing. In one example, the upper limit of the time series data TD can be 10,000, and the upper limit of the time series data TD can be 100 million.

[0143] The classification section 13 classifies a plurality of time series data TD into any one of a plurality of categories based on evaluation values. The number of the plurality of categories is less than the number of time series data TD or the number of evaluation values that can be used generated from the time series data TD. Typically, the number of time series data TD is 100 or more, the number of evaluation values that can be used is 20 or more and less than 100, and in contrast, the number of the plurality of categories is 2 or more and 10 or less.

[0144] The evaluation values of the time series data TD are generated in accordance with a fixed evaluation criterion. Therefore, the categories in which the time series data TD is classified using the evaluation values can also indicate the level of the evaluation criterion. In the following description of the present specification, the categories indicating the level of the evaluation criterion will sometimes be referred to as levels.

[0145] For example, the classification section 13 classifies the time series data TD into any one of the levels 1 to 4 based on the evaluation value. Typically, the classification section 13 classifies the time series data TD into any one of the levels 1 to 4 based on the evaluation value after the acquisition section 12 acquires the evaluation value corresponding to the time series data TD based on the time series data TD and the reference data. It is preferable that the reference data be stored in the storage section 20.

[0146] In one example, the level 1 indicates a category in which the evaluation value is the reference data or relatively close to the reference data, and the level 2 indicates a category in which the evaluation value is close to the reference data next to the level 1. Also, the level 3 indicates a category in which the evaluation value is close to the reference data next to the level 2, and the level 4 indicates a category in which the evaluation value is farther from the reference data than the levels 1 to 3.

[0147] In addition, the classification section 13 can classify the time series data TD into two categories based on the evaluation value. For example, the time series data TD set to the category 1 indicates the reference data or is relatively close to the reference data, and the time series data TD set to the category 2 indicates being farther from the reference data than the category 1.

[0148] The extraction section 14 extracts the time series data corresponding to any one of the categories classified in the classification section 13 as the extracted time series data. Thereby, it is possible to collect the specific time series data indicating the evaluation value of a specific range from the plurality of time series data as the extracted time series data.

[0149] For example, it is preferable that the extraction section 14 extract the time series data corresponding to the category farthest from the reference data among the categories classified in the classification section 13. Thereby, it is possible to easily extract the time series data corresponding to the evaluation value far from the ideal value or the average value. Thus, it is possible to grasp the change of the substrate processing device 200 before a substantial abnormality occurs in the substrate processing device 200. At this time, it is preferable that the control of the substrate processing device 200 be changed so that the change of the substrate processing device 200 is suppressed before the substrate processing in the substrate processing device 200 becomes abnormal.

[0150] As described above, it is preferable that the extraction section 14 extract the time series data corresponding to the category farthest from the reference data among the categories classified in the classification section 13. Also, in a case where the time series data indicating the evaluation value far from the reference data is extracted as the extracted time series data, it is preferable that the extracted time series data be further subjected to the clustering process. However, the extracted time series data extracted by the extraction section 14 can not be the time series data corresponding to the category farthest from the reference data.

[0151] According to the data processing apparatus 100 of the present embodiment, it is possible to easily extract a specific time-series data from among a plurality of time-series data generated in a substrate processing apparatus.

[0152] Next, the data processing method of the present embodiment will be described with reference to Figure 1 and Figure 2 . Figure 2 A flowchart showing the data processing method of the present embodiment.

[0153] As shown in Figure 2 , in step S2, time-series data TD is acquired. The time-series data TD is data generated in the substrate processing apparatus 200. The acquisition unit 11 acquires a plurality of time-series data TD. The plurality of time-series data TD can be data generated in the same processing unit 210 or data generated in different processing units 210. Each of the plurality of time-series data TD indicates a time change of a specific unit of a physical quantity in the substrate processing apparatus 200.

[0154] The acquisition unit 11 acquires the plurality of time-series data TD from the storage unit 20. Alternatively, the acquisition unit 11 can directly acquire the time-series data TD from the substrate processing apparatus 200. For example, the time-series data is sequentially generated in association with substrate processing in the substrate processing apparatus 200 or control of the substrate processing apparatus 200. Next, the process proceeds to step S4.

[0155] In step S4, an evaluation value (score) is acquired. The acquisition unit 12 acquires an evaluation value Ev for each of the plurality of time-series data TD. For example, the acquisition unit 12 generates the evaluation value Ev for each of the plurality of time-series data TD in accordance with a specific evaluation criterion. The evaluation criterion can be stored in the data processing apparatus 100 or outside the data processing apparatus 100. One evaluation value Ev is acquired for one time-series data TD. However, the evaluation value Ev can indicate a plurality of values corresponding to the time-series data TD.

[0156] In addition, a range of a maximum value and a minimum value of the evaluation value can be defined. For example, the evaluation value can be normalized so that a difference between the maximum value and the minimum value of the evaluation value is 1.

[0157] The evaluation value is generated in accordance with a specific evaluation criterion. For example, the evaluation value can be defined so that the higher the evaluation value, the higher the evaluation, or the lower the evaluation value, the lower the evaluation. Alternatively, the evaluation value can be defined so that the higher the evaluation value, the lower the evaluation, or the lower the evaluation value, the higher the evaluation. Alternatively, the evaluation value can be defined so that the closer the evaluation value to a specific value, the higher the evaluation, or the evaluation value can be defined so that the farther the evaluation value from the specific value, the higher the evaluation. Next, the process proceeds to step S6.

[0158] In step S6, using the evaluation value Ev, the time series data TD is classified into any one of a plurality of categories. The number of categories that can be classified is less than the number of values that the evaluation value Ev can represent. Next, the processing proceeds to step S8.

[0159] In step S8, from the plurality of time series data TD, a specific time series data TD that is classified into an object category that becomes an object among the plurality of categories is extracted as an extracted time series data. Typically, the extracted time series data is stored in the storage section 20. However, the extracted time series data can also be stored in an external storage element.

[0160] The data processing method of the present embodiment is performed as above. According to the data processing method of the present embodiment, a specific time series data can be easily extracted from a plurality of time series data generated in the substrate processing apparatus 200. In addition, the extracted extracted time series data among the plurality of time series data TD acquired in step S2 is saved to the storage section 20, on the other hand, the time series data TD that is not extracted can also be erased without being saved to the storage section 20. Thereby, the information required can be efficiently saved.

[0161] The evaluation value of the time series data TD can also be generated in comparison with the reference data. For example, the reference data can also be generated based on a plurality of time series data that represent a good substrate processing result. In one example, the reference data is generated by averaging a plurality of time series data that represent a good substrate processing result. In the case where the reference data is ideal data or time series data at the time of a good substrate processing result, the closer the time series data is to the reference data, the higher the evaluation, and the farther the time series data is from the reference data, the lower the evaluation.

[0162] Next, the time series data TD and the reference data RD will be described with reference to FIG. 3(a) and FIG. 3(b). FIG. 3(a) is a graph representing the time series data TD, and FIG. 3(b) is a graph representing the reference data RD. In the graphs of FIG. 3(a) and FIG. 3(b), the horizontal axis represents time, and the vertical axis represents a physical quantity. For example, the time series data TD and the reference data RD represent the time variation of 100 physical quantities (values) respectively.

[0163] The time series data TD and the reference data RD represent the time variation of the same physical quantity, and the shape of the graph representing the time series data TD is substantially the same as the shape of the graph representing the reference data RD. However, strictly speaking, the time series data TD and the reference data RD are not identical.

[0164] As shown in FIG. 3(a), in the time-series data TD, at the start time point, the physical quantity is substantially zero. When a prescribed period elapses, the physical quantity starts to increase from zero. The physical quantity rapidly increases in a short period to reach a peak. From this point, the physical quantity maintains the peak for a prescribed period. Throughout the peak period, the physical quantity is substantially fixed although it has a little variation. Subsequently, the physical quantity rapidly decreases in a short period to return to substantially zero.

[0165] In FIG. 3(b), the reference data RD is made based on the time-series data when the substrate processing in the substrate processing apparatus 200 is advanced in an ideal state. For example, the reference data RD can be the time-series data when the substrate processing in the substrate processing apparatus 200 is advanced in the most ideal state. Alternatively, the reference data RD can represent an average of the time-series data during the period when the substrate processing in the substrate processing apparatus 200 is advanced in the ideal state.

[0166] As shown in FIG. 3(b), in the reference data, at the start time point, the physical quantity is substantially zero. When a prescribed period elapses, the physical quantity starts to increase from zero. The physical quantity rapidly increases in a short period to reach a peak. From this point, the physical quantity maintains the peak for a prescribed period. Subsequently, the physical quantity rapidly decreases in a short period to return to substantially zero.

[0167] Thus, the graph represented by the time-series data TD exhibits the same tendency as the graph represented by the reference data RD, but strictly speaking, the time-series data is not identical to the reference data RD. For example, the rising period until reaching the peak in the graph of the time-series data TD is different from the rising period until reaching the peak in the graph of the reference data RD. Also, the peak value in the graph of the time-series data TD is different from the peak value in the graph of the reference data RD. Furthermore, the peak in the graph of the time-series data TD varies more than the graph of the reference data RD.

[0168] Next, the data processing method of the present embodiment will be described with reference to Figures 1 to 5(b) FIG. 4(a) is a schematic view representing time-series data TD1 to time-series data TD100. The time-series data TD1 to time-series data TD100 each represent the time variation of the same physical quantity. For example, the acquisition section 11 acquires the time-series data TD1 to time-series data TD100 from the storage section 20.

[0169] The acquisition unit 12 acquires evaluation values Ev from the time-series data TD1 to the time-series data TD100. For example, the acquisition unit 12 acquires the evaluation values Ev1 to Ev100 based on differences between the time-series data TD1 to TD100 and the reference data RD. For example, the acquisition unit 12 generates the evaluation values Ev by accumulating squares of differences between the values of the time-series data TD and the values of the reference data RD. At this time, the closer the time-series data is to the reference data, the smaller the evaluation value is, and the farther the time-series data is from the reference data, the larger the evaluation value is.

[0170] Fig. 4(b) is a graph showing a result of comparing the time-series data TD with the reference data RD. The oblique line in Fig. 4(b) indicates a difference between the time-series data TD and the reference data RD. In a case where the evaluation value is acquired by accumulation of squares of differences, the evaluation value Ev corresponds to an area of the oblique line portion in Fig. 4(b).

[0171] Fig. 4(c) is an evaluation value graph GE showing changes in the evaluation values Ev1 to Ev25 corresponding to the time-series data TD1 to TD25. In Fig. 4(c), the evaluation value graph GE shows changes in the evaluation values Ev1 to Ev25 corresponding to the time-series data TD1 to TD25 among the time-series data TD1 to TD100 in the drawing. As shown in Fig. 4(c), the evaluation values Ev change in order corresponding to the time-series data TD. Here, the evaluation value Ev6 is relatively high compared with the evaluation values Ev1 to Ev5 and the evaluation values Ev7 to Ev10. Also, the evaluation value Ev20 is relatively high compared with the evaluation values Ev15 to Ev19 and the evaluation values Ev21 to Ev25.

[0172] Fig. 5(a) is a graph for explaining classification of the evaluation value graph GE showing changes in the evaluation values Ev1 to Ev25. As shown in Fig. 5(a), the evaluation values Ev are classified according to magnitude. For example, the evaluation values are classified into any one of a plurality of categories based on magnitude.

[0173] Here, the classification section 13 classifies the time-series data TD1 to TD25 into any one of four levels based on the evaluation values Ev1 to Ev25. The classification section 13 classifies the time-series data TD whose magnitude of the evaluation value Ev does not exceed the threshold value Th1 as level 1. The classification section 13 classifies the time-series data TD whose magnitude of the evaluation value Ev is equal to or more than the threshold value Th1 and does not exceed the threshold value Th2 as level 2. The classification section 13 classifies the time-series data TD whose magnitude of the evaluation value Ev is equal to or more than the threshold value Th2 and does not exceed Th3 as level 3. The classification section 13 classifies the time-series data TD whose magnitude of the evaluation value Ev is equal to or more than a prescribed value Th3 as level 4.

[0174] In the graph of Fig. 5(a), a boundary line B1 is displayed at the boundary between level 1 and level 2. Also, a boundary line B2 is displayed at the boundary between level 2 and level 3, and a boundary line B3 is displayed at the boundary between level 3 and level 4.

[0175] In the graph of Fig. 5(a), the evaluation value Ev3, the evaluation value Ev8, the evaluation value Ev12, the evaluation value Ev13, the evaluation value Ev15, the evaluation value Ev17, the evaluation value Ev21, and the evaluation value Ev23 correspond to level 1. The evaluation value Ev1, the evaluation value Ev2, the evaluation value Ev4, the evaluation value Ev5, the evaluation value Ev7, the evaluation value Ev9, the evaluation value Ev10, the evaluation value Ev14, the evaluation value Ev16, the evaluation value Ev18, the evaluation value Ev19, the evaluation value Ev22, the evaluation value Ev24, and the evaluation value Ev25 correspond to level 2. The evaluation value Ev11 corresponds to level 3. The evaluation value Ev6 and the evaluation value Ev20 correspond to level 4.

[0176] Fig. 5(b) is a graph for explaining that the extraction section 14 extracts the time-series data TD6 and the time-series data TD20 corresponding to the evaluation value Ev6 and the evaluation value Ev20. As shown in Fig. 5(b), if the evaluation values Ev obtained in accordance with a certain evaluation criterion are compared, the evaluation value Ev6 and the evaluation value Ev20 have values larger than the other evaluation values. At this time, it can be considered that the time-series data TD6 and the time-series data TD20 corresponding to the evaluation value Ev6 and the evaluation value Ev20 have a tendency to exhibit abnormality compared to the other time-series data TD. Therefore, the extraction section 14 extracts the time-series data TD6 and the time-series data TD20 corresponding to the evaluation value Ev6 and the evaluation value Ev20.

[0177] According to this embodiment, the evaluation values ​​corresponding to multiple time series data TD1 to TD100 can be used as a benchmark to better extract time series data TD6 and time series data TD20 that exhibit an abnormal tendency from the multiple time series data TD1 to TD100. Therefore, by extracting time series data TD6 and time series data TD20 classified as object categories from the multiple time series data TD1 to TD100, the abnormal state of the substrate processing apparatus 200 can be understood and studied without excessive calculation.

[0178] In addition, the description (especially the reference) Figures 3(a) to 5(b) In the description, when generating the evaluation value Ev, the time series data TD is compared with the reference data RD, but this implementation is not limited to this. It is also possible to generate the evaluation value Ev from the time series data TD without using the reference data RD.

[0179] For example, the acquisition unit 12 may also generate an evaluation value Ev by comparing a portion of the time series data TD with a specific value stored in the storage unit 20. For example, the storage unit 20 may store values ​​(peak equivalent values) that correspond to the peak values ​​of the graph of the time series data TD, and the acquisition unit 12 may generate the evaluation value Ev by comparing the peak values ​​of the time series data TD with the peak equivalent values ​​stored in the storage unit 20.

[0180] Alternatively, the storage unit 20 may store values ​​(equivalent values ​​of rising periods) corresponding to the rising period of a graph of time series data, and the acquisition unit 12 may compare the rising period of the time series data TD with the equivalent values ​​of rising periods stored in the storage unit 20 to generate an evaluation value Ev. Alternatively, the storage unit 20 may store values ​​(equivalent values ​​of peak fluctuations) corresponding to the peak fluctuations of a graph of time series data, and the acquisition unit 12 may compare the peak fluctuations of the time series data TD with the equivalent values ​​of peak fluctuations stored in the storage unit 20 to generate an evaluation value Ev.

[0181] Furthermore, in the data processing apparatus 100, it is preferable to be able to input instructions from operators or managers. Moreover, in the data processing apparatus 100, it is preferable to be able to display information to operators or managers. For example, it is particularly preferable to be able to display evaluation values ​​or time series data (especially extracted time series data) to operators or managers.

[0182] Next, refer to Figure 6 The data processing apparatus 100 of this embodiment will be described. Figure 6 This is a schematic diagram of the data processing device 100 and the substrate processing device 200. Figure 6 The data processing device 100 also includes a display unit 30 and an input unit 40, in addition to having a reference...Figure 1 The data processing apparatus 100 described above has the same structure. Therefore, in order to avoid redundancy, the repeated description is omitted.

[0183] The display section 30 displays an operation screen or a result of various processing. Also, the display section 30 displays a graph representing time series data or an evaluation value graph GE representing a change in evaluation value.

[0184] The display section 30 has a display. For example, the display includes a liquid crystal display or an organic electroluminescence (EL) display.

[0185] Also, it is preferable that the display section 30 display the evaluation value graph GE representing a change in evaluation value. Also, it is preferable that, when the operator or the administrator specifies one value (evaluation value) of the evaluation value graph GE via the input section 40 in a case where the evaluation value graph GE is displayed, the display section 30 displays time series data corresponding to the specified value.

[0186] The input section 40 includes, for example, various keys for instructing a kind of job and a content of the job. The input section 40 includes a keyboard and a mouse. Alternatively, the input section 40 can include a touch sensor. Also, the display section 30 and the input section 40 can be a touch screen in which both are integrated.

[0187] Next, the display section 30 switches and displays the evaluation value graph GE and a graph representing time series data TD will be described with reference to Figures 4(a) to 7(c)

[0188] Figs. 7(a) and 7(b) represent the evaluation value graph GE displayed by the display section 30. As shown in Fig. 7(a), the display section 30 displays the evaluation value graph GE representing a change in values of the evaluation values Evl to EvlOO corresponding to the time series data TDl to TDlOO. The operator or the administrator specifies the time series data TD corresponding to the evaluation value Ev in order to confirm the time series data TD corresponding to the evaluation value Ev via the input section 40.

[0189] As shown in Fig. 7(b), the operator or the administrator moves the cursor CU to the evaluation value Ev to be an object via the input section 40 to specify a point representing the specific evaluation value Ev in the evaluation value graph GE. Here, the cursor CU specifies the evaluation value Ev6. At this time, the display section 30 displays the time series data TD6 corresponding to the specified evaluation value Ev6.

[0190] ​Fig. 7(c) is a chart showing the time-series data TD displayed by the display section 30. Here, the display section 30 displays the time-series data TD6 corresponding to the evaluation value Ev6.

[0191] In addition, it is preferable that the display section 30, in a case where the time-series data TD is displayed, further displays a button BG for returning to the evaluation value chart GE. When the button BG is selected, the display section 30 switches to a screen where the evaluation value chart GE is displayed again. It is preferable that the display section 30 switches the display between the evaluation value chart GE showing the change in the evaluation value Ev and the chart showing the time-series data TD as such.

[0192] In addition, typically, the storage section 20 stores a plurality of time-series data. It is preferable that the storage section 20 stores the plurality of time-series data in the order of generation of the time-series data. At this time, it is preferable that the display section 30 switches the display of the plurality of time-series data corresponding to a prescribed operation.

[0193] Next, the display of the time-series data by the display section 30 will be described with reference to Figures 1 to 8(c) Figures 8(a) to 8(c) is a schematic view showing the change in the display of the plurality of time-series data by the display section 30. Figures 8(a) to 8(c)

[0194] The display screen 32 includes a display region 33 and an operation region 34. The display region 33 displays one time-series data. Here, the display region 33 extends in the horizontal direction. The horizontal length of the display region 33 is larger than the vertical length of the display region 33.

[0195] The time-series data displayed in the display region 33 can be switched by the operation region 34. Here, the operation region 34 is disposed below the display region 33.

[0196] Figures 8(a) to 8(c) In the embodiment, the operation region 34 is a scroll bar. The operation region 34 includes an arrow 34a and a knob 34b. The arrow 34a extends linearly in the horizontal direction. The knob 34b is movable along the arrow 34a in a manner that a portion of the knob 34b coincides with the arrow 34a. The time-series data displayed in the display region 33 is switched corresponding to the position of the knob 34b within the arrow 34a.

[0197] The knob 34b is moved in the horizontal direction within the arrow 34a corresponding to the input from the input section 40. For example, when the knob 34b is moved in the left direction or the right direction in a state where the knob 34b is selected, the time-series data TD displayed in the display region 33 is switched.

[0198] ​​For example, as shown in Fig. 8(a), when the knob 34b is positioned at the left end of the arrow 34a, the time series data TDl generated first is displayed in the display area 33. The knob 34b is moved in the horizontal direction within the arrow 34a in response to an input from the input section 40.

[0199] As shown in Fig. 8(b), when the knob 34b is moved in the right direction in the state where the knob 34b is selected, the time series data displayed in the display area 33 is switched. For example, when the knob 34b is moved to the center of the arrow 34a, the time series data displayed in the display area 33 is switched to the time series data TD50.

[0200] Further, as shown in Fig. 8(c), when the knob 34b is further moved in the right direction in the state where the knob 34b is selected, the time series data displayed in the display area 33 is switched to the time series data TD100. As described above, the display section 30 can switch the display of a plurality of time series data in response to a prescribed operation.

[0201] Next, the switching display of the chart indicating the change of the evaluation value and the time series data by the display section 30 will be described with reference to Figures 9(a) to 9(c) Figs. 9(a) and 9(b). Fig. 9(a) and 9(b) show the extracted evaluation value chart EGE displayed by the display section 30. The extracted evaluation value chart EGE indicates the time change of the evaluation value corresponding to the extracted time series data extracted from the plurality of time series data TD. At this time, the time change of the evaluation value of the extracted time series data can be displayed.

[0202] As shown in Fig. 9(a), the display section 30 displays the extracted evaluation value chart EGE indicating the change of the evaluation value Ev corresponding to the extracted time series data. The operator or the administrator specifies the time series data corresponding to the evaluation value Ev in order to confirm the extracted time series data corresponding to the evaluation value Ev via the input section 40.

[0203] As shown in Fig. 9(b), the operator or the administrator moves the cursor CU so as to move to the evaluation value Ev set as the object to specify the particular evaluation value Ev. Here, the cursor CU specifies the evaluation value Ev. At this time, the display section 30 displays the extracted time series data corresponding to the specified evaluation value Ev.

[0204] Fig. 9(c) is a chart showing the extracted time series data displayed by the display section 30. It is preferable that the display section 30 switch the display of the chart indicating the change of the evaluation value Ev and the extracted evaluation value chart EGE indicating the extracted time series data as described above. Thereby, the time change of the evaluation value with respect to the extracted time series data can be displayed.

[0205] In addition, the switching display of the chart indicating the change of the evaluation value and the time series data by the display section 30 will be described with reference to Figures 1 to 9(c)In the above-described explanation, the extracted time series data is extracted, but preferably the extracted time series data is processed. For example, the extracted time series data is preferably subjected to clustering processing.

[0206] Next, the data processing apparatus 100 according to the present embodiment will be described with reference to Figure 10 Next, the data processing apparatus 100 according to the present embodiment will be described with reference to Figure 10 is a schematic view of the data processing apparatus 100 and the substrate processing apparatus 200. Figure 10 The data processing apparatus 100 according to the present embodiment further includes the clustering processing section 15, and has the same structure as the data processing apparatus 100 described with reference to Figure 6 The data processing apparatus 100 described above has the same structure as the data processing apparatus 100 described with reference to

[0207] The data processing apparatus 100 further includes the clustering processing section 15. The clustering processing section 15 is included in the processing section 10.

[0208] The clustering processing section 15 performs clustering processing on the extracted time series data. For example, the clustering processing section 15 classifies the extracted time series data into any one of a plurality of clusters, based on a feature of the extracted time series data. In one example, the clustering processing section 15 classifies the extracted time series data into any one of a plurality of clusters, based on a feature of the graph shown by the extracted time series data.

[0209] Preferably, the clustering processing section 15 performs clustering processing on the extracted time series data by so-called unsupervised learning. For example, the clustering processing section 15 can perform clustering processing using all values of the extracted time series data. For example, the clustering processing section 15 can perform clustering processing by any one of hierarchical clustering, k-means clustering, Gaussian mixture model clustering, self-organizing map clustering, and hidden Markov model clustering.

[0210] For example, the plurality of clusters include an upward delay cluster of upward delay of the graph, a peak variation cluster of large peak variation of the graph, and a low peak cluster of low peak of the graph. At this time, the clustering processing section 15 performs clustering processing to classify the extracted time series data into at least one of the upward delay cluster, the peak variation cluster, and the low peak cluster. In addition, the plurality of clusters can further include other clusters. For example, the plurality of clusters can include a downward delay cluster of downward delay of the graph.

[0211] In addition, the clustering processing section 15 can perform clustering processing using all values of the extracted time series data. For example, the clustering processing section 15 can perform processing on all values of the extracted time series data by vector analysis. Alternatively, the clustering processing section 15 can perform clustering processing using a part of values of the extracted time series data.

[0212] Next, the data processing apparatus 100 according to the present embodiment will be described with reference to Figures 11(a) to 11(c)The clusters into which the extracted time-series data is classified will be described. FIG. 11(a) is a graph showing extracted time-series data ETD1 classified into an up-delay cluster, FIG. 11(b) is a graph showing extracted time-series data ETD2 classified into a peak fluctuation cluster, and FIG. 11(c) is a graph showing extracted time-series data ETD3 classified into a low-peak cluster. In addition, Figures 11(a) to 11(c) In each of the graphs, the reference data RD is shown together with the extracted time-series data.

[0213] As shown in FIG. 11(a), in a case where the extracted time-series data is compared with the reference data, the up of the graph of the extracted time-series data ETD1 is delayed with respect to the graph of the reference data RD. Such extracted time-series data is classified into an up-delay cluster.

[0214] As shown in FIG. 11(b), in a case where the extracted time-series data is compared with the reference data, the peak of the graph of the extracted time-series data ETD2 fluctuates greatly compared with the graph of the reference data RD. Such extracted time-series data is classified into a peak fluctuation cluster.

[0215] As shown in FIG. 11(c), in a case where the extracted time-series data is compared with the reference data, the peak of the graph of the extracted time-series data ETD3 is lower than the reference data RD. Such extracted time-series data is classified into a low-peak cluster.

[0216] In addition, with reference to Figures 11(a) to 11(c) clusters into which the extracted time-series data is classified are exemplified, but in the present embodiment, the clusters into which the extracted time-series data is classified are not limited to these. The extracted time-series data can also be classified into other clusters.

[0217] Next, with reference to Figures 1 to 12 the data processing method of the present embodiment will be described. Figure 12 is a flowchart of the data processing method of the present embodiment. Figure 12 The flowchart of is appended with the clustering process of step S10, and is otherwise the same as the flowchart with reference to Figure 2 the above-described flowchart. Therefore, in order to avoid redundancy, the overlapping description is omitted.

[0218] As shown in Figure 12 extracted time-series data is extracted as shown in step S8. The extraction section 14 extracts the extracted time-series data from the plurality of time-series data based on the classification result of the evaluation value. Next, the process proceeds to step S10.

[0219] In step S10, the extracted time-series data is subjected to a clustering process. The clustering process section 15 classifies the extracted time-series data into any one of a plurality of clusters.

[0220] The data processing method of the present embodiment is performed as above. According to the data processing method of the present embodiment, the extracted time series data that exhibits abnormal evaluation values can be classified into each cluster from among the plurality of time series data.

[0221] In general, if clustering processing is performed on all of the time series data, since the time series data also includes normal time series data, a cluster that is not abnormal can be formed unnecessarily, and thus it can be impossible to grasp the abnormal state of the substrate processing apparatus sufficiently. In contrast, according to the data processing method of the present embodiment, the extracted time series data that exhibits abnormal evaluation values is subjected to clustering processing from among the plurality of time series data, and thus the abnormal state of the substrate processing apparatus 200 can be classified into clusters more accurately. Furthermore, according to the data processing method of the present embodiment, clustering processing is not performed on all of the time series data, and thus unnecessary computation can be avoided.

[0222] In addition, it is preferable that the display section 30 display, for each cluster, an evaluation value chart that indicates the change in the corresponding evaluation value. At this time, it is preferable that, when an operator or manager specifies one evaluation value of the chart, the display section 30 displays the time series data corresponding to the specified evaluation value.

[0223] For example, in the description with reference to Figures 9(a) to 9(c) , the display section 30 displays the extracted evaluation value chart EGE that indicates the temporal change in the evaluation value with respect to the extracted time series data extracted by the extraction section 14, but the present embodiment is not limited thereto. The display section 30 can also display a chart that indicates the temporal change in the evaluation value with respect to the time series data classified into a cluster by the clustering processing section 15. At this time, the temporal change in the evaluation value within a specific cluster can be displayed.

[0224] In addition, it is preferable that, as with reference to Figures 10 to 12 , the extracted time series data is subjected to clustering processing as described above. Needless to say, the extracted time series data reflects the abnormal state of the substrate processing apparatus 200, and the result of the clustering processing indicates the state of the substrate processing apparatus 200 corresponding to the kind of the abnormal state of the substrate processing apparatus 200. Therefore, it is preferable that the result of the clustering processing is used to learn the state of the substrate processing apparatus 200.

[0225] Next, the data processing method of the present embodiment will be described with reference to Figures 1 to 13 . Figure 13 is a flowchart of the data processing method of the present embodiment. Figure 13The flowchart shown in FIG. 8A determines whether the extracted time series data has been extracted in step S8A, and steps S2B to S6B, step S9, and steps S22 to S24 are added thereto, other than that, the same as the flowchart shown in FIG. 7 is followed. Figure 12 The flowchart shown in FIG. 8A determines whether the extracted time series data has been extracted in step S8A, and steps S2B to S6B, step S9, and steps S22 to S24 are added thereto, other than that, the same as the flowchart shown in FIG. 7 is followed. Figure 13 The data processing method shown in FIG. 8A can preferably be used for creating a learning database in the substrate processing apparatus 200.

[0226] In step S2, the time series data is acquired. Here, the time series data is used to learn the state of the substrate processing apparatus 200. Therefore, in this specification, the time series data is sometimes referred to as learning time series data.

[0227] Here, the acquisition unit 11 acquires one learning time series data generated in the substrate processing apparatus 200. Next, the process proceeds to step S4. Further, steps S4 and S6 are the same as the flowchart shown in FIG. 7. Figure 12 The flowchart shown in FIG. 8A determines whether the extracted time series data has been extracted in step S8A, and steps S2B to S6B, step S9, and steps S22 to S24 are added thereto, other than that, the same as the flowchart shown in FIG. 7 is followed.

[0228] In step S8A, it is determined whether the learning time series data is extracted as the extracted time series data. In a case where the classification unit 13 classifies the learning time series data into the target category based on the evaluation value Ev, the extraction unit 14 extracts the learning time series data as the extracted time series data. On the other hand, in a case where the classification unit 13 classifies the learning time series data into the non-target category based on the evaluation value Ev, the extraction unit 14 does not extract the learning time series data as the extracted time series data.

[0229] For example, in a case where the classification unit 13 classifies the learning time series data into the level 4 based on the evaluation value Ev of the learning time series data, the extraction unit 14 extracts the learning time series data as the extracted time series data. Also, in a case where the classification unit 13 classifies the learning time series data into any one of the levels 1 to 3, the extraction unit 14 does not extract the learning time series data as the extracted time series data.

[0230] If the extracted time series data is not extracted (NO in step S8A), the process returns to step S2. If the extracted time series data is extracted (YES in step S8A), the process proceeds to step S10. Further, if the clustering process is not performed in advance, it is also possible to proceed to step S10 after the data of the number that enables unsupervised learning is extracted.

[0231] In step S10, the extracted time series data is subjected to a clustering process. Thereby, the learning time series data is classified into any one of a plurality of clusters. Next, the process proceeds to step S22. In step S2, the time series data is acquired. Here, the time series data is used to learn the state of the substrate processing apparatus 200. Therefore, in this specification, the time series data is sometimes referred to as learning time series data.

[0232] In step S22, the control of the substrate processing apparatus 200 is changed. Specifically, the control of the substrate processing apparatus 200 or the processing unit 210 for which the learning time series data is generated is changed. The change of the control of the substrate processing apparatus 200 can also be made in accordance with a program that controls the driving of the substrate processing apparatus 200. Alternatively, the change of the control of the substrate processing apparatus 200 can also be input by hand in the substrate processing apparatus 200.

[0233] Typically, the control of the substrate processing apparatus 200 is made in a state in which the substrate processing in the substrate processing apparatus 200 is temporarily stopped. For example, if the amount of the supply of the processing liquid for processing the substrate changes as indicated by the time series data TD, and the amount of the supply of the processing liquid is in an abnormal state, the substrate processing is temporarily stopped and the valve is adjusted so as to return the amount of the supply to normal. Alternatively, the change of the control of the substrate processing apparatus 200 can also be made while continuing the substrate processing in the substrate processing apparatus 200. Next, the process proceeds to step S2B.

[0234] In step S2B, the time series data is acquired from the substrate processing apparatus 200 whose control is changed. In this specification, the time series data acquired from the substrate processing apparatus 200 whose control is changed is sometimes referred to as changed time series data. Here, the acquisition unit 11 acquires one of the time series data generated in the substrate processing apparatus 200 as the changed time series data. Next, the process proceeds to step S4B.

[0235] In step S4B, the evaluation value (score) is acquired. The acquisition unit 12 acquires the evaluation value with respect to the changed time series data for the substrate processing apparatus 200 whose control is changed. For example, the acquisition unit 12 generates the evaluation value with respect to the changed time series data in accordance with a specific evaluation criterion, similarly to the learning time series data. Next, the process proceeds to step S6B.

[0236] In step S6B, the changed time series data is classified into any one of a plurality of categories using the evaluation value. Next, the process proceeds to step S9.

[0237] In step S9, it is determined whether or not the changed time series data is extracted as the extracted time series data. In a case where the classification unit 13 classifies the changed time series data into the target category, the extraction unit 14 extracts the changed time series data as the extracted time series data. On the other hand, in a case where the classification unit 13 classifies the changed time series data into the non-target category, the extraction unit 14 does not extract the time series data that becomes the target as the extracted time series data.

[0238] For example, in a case where the classification section 13 classifies the change time series data into the level 4 based on the evaluation value Ev of the change time series data, the extraction section 14 extracts the change time series data as the extraction time series data. Also, in a case where the classification section 13 classifies the change time series data into any one of the levels 1 to 3, the extraction section 14 does not extract the change time series data as the extraction time series data.

[0239] If the extraction time series data is extracted (Yes in step S9), the processing returns to step S22. At this time, the substrate processing apparatus 200 is controlled again to further change the control of the substrate processing apparatus 200. On the other hand, if the extraction time series data is not extracted (No in step S9), the processing proceeds to step S24.

[0240] In step S24, information related to the change content of the control of the substrate processing apparatus 200 performed in step S22 is stored in the storage section 20 as cause countermeasure information. The cause countermeasure information is information related to the change content of the control of the substrate processing apparatus 200 performed in step S22. For example, the cause countermeasure information can also be information indicating the change content of the control of the substrate processing apparatus 200 itself. Alternatively, the cause countermeasure information can also be information considered to be a cause of the learning time series data presenting an abnormal state.

[0241] For example, an operator or a manager of the data processing apparatus 100 and / or the substrate processing apparatus 200 inputs the cause countermeasure information via the input section 40, and the storage section 20 stores the cause countermeasure information. At this time, it is preferable that, in the storage section 20, the cause countermeasure information is stored together with the learning time series data for each cluster subjected to the clustering processing in step S10. In addition, the cluster, the learning time series data, and the cause countermeasure information stored in the storage section 20 can preferably be used as a learning database.

[0242] The data processing method of the present embodiment is performed as described above. According to the data processing method of the present embodiment, it is possible to determine an abnormal state of the substrate processing apparatus 200 using the acquisition section 11, the acquisition section 12, the classification section 13, and the extraction section 14 of the data processing apparatus 100, and it is possible to determine whether or not the abnormal state of the substrate processing apparatus 200 has been eliminated after the control of the substrate processing apparatus 200 is changed.

[0243] Also, according to the data processing method of the present embodiment, the cause countermeasure information is stored after the time series data when the substrate processing apparatus 200 extracts a specific state is subjected to the clustering processing. Therefore, it is possible to store the cause countermeasure information considered to be effective to restore the time series data of the substrate processing apparatus 200 from an abnormal state to an original state together.

[0244] Further, preferably, the storage section 20 stores the clusters, the extracted time series data, and the cause countermeasure information as a learning database.

[0245] Next, the learning database of the storage section 20 will be described with reference to Figure 14 Figure 14 is a schematic table for illustrating the learning database of the storage section 20.

[0246] As Figure 14 illustrated, the learning database includes clusters classified by the clustering processing section 15, time series data included in the clusters, and cause countermeasure information corresponding to the clusters. For example, Figure 14 In the table illustrated in

[0247] Specifically, the rise delay cluster includes the time series data TD6 and the time series data TD40 as the time series data, and includes the process liquid concentration drop / valve retightening as the cause countermeasure information. The wave peak variation cluster includes the time series data TD20 and the time series data TD41 as the time series data, and includes the substrate holding force drop / overcurrent suppression as the cause countermeasure information. The low wave peak cluster includes the time series data TD80 and the time series data TD95 as the time series data, and includes the environment unevenness / sensor restart as the cause countermeasure information.

[0248] Further, in the description with reference to Figure 14 , in order to avoid the invention from becoming overly complex, a set of cause countermeasure information is indicated corresponding to one cluster, but one cluster can also correspond to two or more cause countermeasure information. At this time, preferably, the generation of the cluster and the correspondence of the cause countermeasure information are appropriately learned using so-called machine learning.

[0249] In the present embodiment, the learning database is used for control modification of the substrate processing apparatus 200 and / or the processing unit 210 using time series data generated by the substrate processing apparatus 200 and / or the processing unit 210 later. The data processing apparatus 100 can also modify the control of the substrate processing apparatus 200 based on the learning contents stored in the learning database of the storage section, according to time series data generated in the substrate processing apparatus 200.

[0250] Next, the data processing apparatus 100 of the present embodiment will be described with reference to Figures 1 to 15 Figure 15 is a schematic view of the data processing apparatus 100 of the present embodiment. The data processing apparatus 100 includes a processing section 110, a storage section 120, a display section 130, and an input section 140. Here, the learning database is stored in the storage section 120. The display section 130 and the input section 140 correspond to the description with reference to Figure 6 ​​The display section 30 and the input section 40 described above.

[0251] The processing section 110 includes a data acquisition section 111, an evaluation value acquisition section 112, a classification section 113, a matching section 114, and a readout section 115. The data acquisition section 111, the evaluation value acquisition section 112, and the classification section 113 correspond to the data acquisition section 11, the evaluation value acquisition section 12, and the classification section 13 described above. Figure 6 The data acquisition section 11, the evaluation value acquisition section 12, and the classification section 13 described above.

[0252] The matching section 114 matches the time series data classified as the target category by the classification section 113 with the extracted time series data stored in the learning database of the storage section 120. In this specification, the time series data classified as the target category by the classification section 113 is sometimes referred to as target time series data. Also, the time series data of the learning database similar to the target time series data is sometimes referred to as similar time series data.

[0253] The matching section 114 determines the time series data of the learning database similar to the target time series data as the similar time series data. Typically, the time series data of the learning database closest to the target time series data becomes the similar time series data. However, in the learning database, there are also cases where there is no time series data that becomes the similar time series data.

[0254] Typically, the matching section 114 matches the target time series data with the time series data stored in the learning database in accordance with a specific evaluation criterion. For example, the matching section 114 matches the target time series data with the time series data of the learning database for the entire time series data. In one example, the matching section 114 can also process the target time series data and the time series data of the learning database by vector analysis. Alternatively, the matching section 114 can also match the target time series data with the time series data stored in the learning database for a specific part of the time series data. Through the matching section 114, the similar time series data close to the target time series data among the time series data stored in the learning database of the storage section 120 is determined. In addition, the matching of the target time series data and the similar time series data can be performed in accordance with the same evaluation criterion as the evaluation value acquisition section 112, or in accordance with an evaluation criterion different from the evaluation value acquisition section 112.

[0255] Subsequently, the readout section 115 determines the cluster corresponding to the similar time series data in the learning database of the storage section 120, and reads out the cause countermeasure information corresponding to the determined cluster. Typically, the display section 130 displays the cause countermeasure information.

[0256] The data processing apparatus 100 according to the present embodiment reads out similar time series data of the object time series data and corresponding cause countermeasure information from the learning database. Thus, the data processing apparatus 100 can acquire appropriate cause countermeasure information for the object time series data classified into a specific category.

[0257] Next, the data processing method according to the present embodiment will be described with reference to Figures 1 to 16 Figure 16 A flowchart showing the data processing method according to the present embodiment. Figure 16 Steps S102 to S106 of the flowchart of Figure 12 correspond to steps S2 to S6 of the flowchart of

[0258] As shown in Figure 16 , in step S102, time series data is acquired. The time series data is data generated in the substrate processing apparatus 200. Next, the process proceeds to step S104.

[0259] In step S104, an evaluation value (score) of the time series data is acquired. The acquisition unit 12 acquires the evaluation value of the time series data. For example, the acquisition unit 12 acquires the evaluation value of the time series data in accordance with the same evaluation criteria as when the learning database is created. Next, the process proceeds to step S106.

[0260] In step S106, the time series data is classified into any one of a plurality of categories using the evaluation value of the time series data. For example, the classification unit 13 classifies the time series data into any one of a plurality of categories using the evaluation value in the same manner as when the learning database is created. Next, the process proceeds to step S106a.

[0261] In step S106a, it is determined whether the time series data has been classified into an object category. For example, the object category is level 4.

[0262] If the time series data has not been classified into the object category (NO in step S106a), the process returns to step S102. If the time series data has been classified into the object category (YES in step S106a), the process proceeds to step S108.

[0263] In step S108, the object time series data is matched with time series data stored in the learning database of the storage unit 120. The matching unit 114 matches the object time series data with time series data stored in the learning database of the storage unit 120. The matching unit 114 determines similar time series data in the extracted time series data stored in the learning database of the storage unit 120. Next, the process proceeds to step S110.

[0264] ​In step S110, the cause countermeasure information corresponding to the similar time series data is read out from the learning database. The reading unit 115 determines the cluster corresponding to the similar time series data in the learning database of the storage unit 120, and reads out the cause countermeasure information corresponding to the determined cluster. Typically, the display unit 130 displays the cause countermeasure information thereafter. In addition, as needed, the operator or the manager of the data processing apparatus 100 and / or the substrate processing apparatus 200 can also change the control of the substrate processing apparatus 200 based on the displayed cause countermeasure information.

[0265] The data processing method of the present embodiment is performed as above. According to the data processing method of the present embodiment, the subject time series data determined to be abnormal via the evaluation value is matched. Therefore, the amount of calculation related to the matching processing can be reduced.

[0266] Furthermore, according to the present embodiment, the cause countermeasure information corresponding to the cluster of the subject time series data can be acquired from the learning database. Therefore, even if the time series data is abnormal, the cause countermeasure information corresponding to the time series data can be acquired from the learning database. Therefore, the abnormal state of the substrate processing apparatus 200 can be effectively grasped.

[0267] Typically, the operator skilled in the operation of the substrate processing apparatus can infer the condition of the substrate processing apparatus from the time series data acquired from the substrate processing apparatus, but the unskilled operator cannot infer the condition of the substrate processing apparatus from the time series data acquired from the substrate processing apparatus. However, according to the data processing method of the present embodiment, if the subject time series data of the substrate processing apparatus 200 is abnormal, the cause countermeasure information of the substrate processing apparatus 200 can be acquired using the learning database made in the past. Therefore, even if the operator of the substrate processing apparatus 200 is unskilled, the undesirable condition of the substrate processing apparatus 200 can be eliminated.

[0268] In addition, in the description with reference to Figure 16 , in step S110, the cause countermeasure information is read out, and thereafter, the operator or the manager changes the control of the substrate processing apparatus 200 based on the displayed cause countermeasure information, but the present embodiment is not limited thereto. The substrate processing apparatus 200 can also change the control of the substrate processing apparatus 200 based on the read-out cause countermeasure information without going through the operator or the manager.

[0269] Furthermore, in the description with reference to Figure 16 , in the matching in step S108, the similar time series data is determined from the learning database, but if the time series data acquired in step S102 is extremely abnormal, it can also be considered that the similar time series data is not determined from the learning database. At this time, the time series data acquired in step S102 can also be stored in the storage unit 120 separately.

[0270] Next, the data processing method of the present embodiment will be described with reference to Figures 1 to 17 Figure 17 A flowchart showing the data processing method of the present embodiment. Figure 17 The flowchart of the data processing method of the present embodiment is supplemented with steps S108a, steps S112-S114, and is otherwise the same as the flowchart of the data processing method of the first embodiment, and thus the overlapping description is omitted to avoid redundancy. Figure 16

[0271] As described above, in step S108, the target time series data is matched with the time series data stored in the learning database of the storage section 120. The matching section 114 matches the target time series data with the time series data stored in the learning database of the storage section 120. For example, the matching is performed by obtaining a matching rate of the target time series data with the time series data of the learning database. Next, the process proceeds to step 108a.

[0272] In step S108a, it is determined whether there is similar time series data similar to the target time series data within the learning database. For example, the matching section 114 can determine whether there is similar time series data similar to the target time series data within the learning database depending on whether the matching rate exceeds a threshold value. For example, if the matching rate is higher than the threshold value, the matching section 114 determines that there is similar time series data. On the other hand, if the matching rate is equal to or lower than the threshold value, the matching section 114 determines that there is no similar time series data.

[0273] If it is determined that there is similar time series data (Yes in step S108a), the process proceeds to step S110. If it is determined that there is no similar time series data (No in step S108a), the process proceeds to step S114.

[0274] In step S110, the cause countermeasure information corresponding to the similar time series data is read out from the learning database. The reading section 115 reads out the cause countermeasure information corresponding to the similar time series data from the learning database of the storage section 120. The process proceeds to step S112.

[0275] In step S112, the control of the substrate processing apparatus 200 is changed based on the cause countermeasure information. For example, the processing section 110 notifies the cause countermeasure information to the operator or the administrator. In one example, the display section 130 displays the cause countermeasure information.

[0276] ​​In step S114, the target time series data is stored in the learning database of the storage section 120. At this time, it is preferable that the storage section 120 stores information indicating a new cluster corresponding to the target time series data together with the target time series data. Also at this time, the display section 130 can display that the matching rate is below the threshold value.

[0277] The data processing method of the present embodiment is performed as described above. According to the data processing method of the present embodiment, the cause countermeasure information corresponding to the cluster of the target time series data is read out from the learning database and displayed, so the operator or the administrator of the data processing apparatus 100 can acquire the cause countermeasure information about the substrate processing apparatus 200.

[0278] Next, the time series data TD will be described in detail with reference to Figures 18(a) to 20(b) Typically, the time series data TD can be divided into an ascending portion, a stable period portion, and a descending portion according to the shape of the graph. Also typically, the time series data TD indicates a temporal change of a physical quantity in response to a prescribed control signal.

[0279] First, the time series data TD will be described with reference to Figure 1 Figs. 18(a) and 18(b). Fig. 18(a) is a graph indicating a temporal change of a control signal CS in the substrate processing apparatus 200. Fig. 18(b) is a graph indicating the time series data TD representing a temporal change of a physical quantity controlled in accordance with the control signal CS.

[0280] As shown in Fig. 18(a), the control signal CS is at a low level in an initial state (time tO). The control signal CS changes from the low level to a high level at time tl and changes from the high level to the low level at time t2.

[0281] As shown in Fig. 18(b), the time series data TD changes between an initial level L0 and a target level LI (where L0

[0282] The time series data TD has the initial level L0 in the initial state (time tO). When the control signal CS changes from the low level to the high level at time tl, the time series data TD starts to ascend from the initial level L0 toward the target level LI. Here, the time series data TD descends after ascending beyond the target level LI. The time series data TD repeatedly ascends and descends around the target level LI and eventually stabilizes around the target level LI.

[0283] Subsequently, when the control signal CS changes from the high level to the low level at time t2, the time series data TD starts to decrease from the vicinity of the target level Ll toward the initial level L0. The time series data TD increases after decreasing to the initial level L0 or the vicinity thereof. The time series data TD repeatedly increases and decreases in the vicinity of the initial level L0, and finally stabilizes at the initial level L0.

[0284] For example, in the time series data TD, a first range Rl including the target level Ll and a second range R2 including the initial level L0 are set. For example, a range of 90% to 110% of the target level Ll is set as the first range Rl, and a range of -10% to 10% of the target level Ll is set as the second range R2. The upper and lower limits of the first range Rl and the upper and lower limits of the second range R2 are arbitrarily determined by the user.

[0285] The evaluation value acquisition section 12 can also acquire the evaluation value from the time series data TD as follows. First, the evaluation value acquisition section 12 obtains a period from when the control signal CS changes from the low level to the high level until the time series data TD converges within the first range Rl as "period A (rise period)", obtains a period from when the control signal CS changes from the high level to the low level until the time series data TD converges within the second range R2 as "period C (fall period)", and obtains a period between the rise period and the fall period as "period B (stabilization period)". Further, by "the time series data converges within a certain range" is meant that the time series data does not take a value outside the range after this time point.

[0286] The evaluation value acquisition section 12 can also acquire the evaluation value from any one of the "period A", "period B", and "period C". For example, the evaluation value acquisition section 12 can also acquire the evaluation value from the "period A". Alternatively, the evaluation value acquisition section 12 can also acquire the evaluation value from the "period B" or "period C".

[0287] For example, the evaluation value acquisition section 12 obtains a score of the rise period, a score of the stabilization period, and a score of the fall period by a predetermined method. For example, the evaluation value acquisition section 12 can also obtain a length of the rise period as the score of the rise period, and obtain a length of the fall period as the score of the fall period. In this way, the evaluation value acquisition section 12 can also use the lengths of the "period A", "period B", and "period C" themselves as the evaluation value. Alternatively, the evaluation value acquisition section 12 can also use an average or a composite value of the lengths of the "period A", "period B", and "period C" as the evaluation value.

[0288] Alternatively, the evaluation value acquisition section 12 can also acquire any one value (score) from the "period A", "period B", and "period C" respectively, and determine, compare, and / or composite the scores, thereby acquiring the evaluation value.

[0289] The evaluation value acquisition section 12 can also calculate a statistical value of the time series data TD in the period A, the period B, and the period C as the evaluation value or score of the period A, the period B, and the period C.

[0290] For example, the evaluation value acquisition section 12 can also calculate a burst value of the stable period using a plurality of time series data as the score of the stable period. The evaluation value acquisition section 12 calculates a mean value, a median value, or a variance of the time series data TD in the stable period as the score of the stable period.

[0291] Figure 19 is a graph showing values of a plurality of time series data. Figure 19 The n time series data TD1, TD2,..., TDn shown in the graph each include m values. Here, when i is an integer of 1 or more and n or less, and j is an integer of 1 or more and m or less, the jth data included in the time series data TDi is referred to as x ij At this time, the evaluation value acquisition section 12 calculates the score Sp of the time series data TDp in accordance with the following equations (1) to (4).

[0292] [Equation 1]

[0293]

[0294] In addition, in the equation (1), the value μ pj represents a mean value of the jth data included in the (n-1) time series data other than the time series data TDp that is the object. In the equation (2), the value μ p represents a mean value of all data included in the (n-1) time series data other than the time series data TDp. In the equation (3), the value σ p 2 represents a variance of the (n-1) time series data other than the time series data TDp.

[0295] Further, the evaluation value acquisition section 12 can also calculate an overshoot amount of the time series data TD as the score of the rising period. When a target level of the time series data TD is L1, and a maximum value of the time series data TD is M, the evaluation value acquisition section 12 can also calculate the overshoot amount V of the time series data in accordance with the equation (5).

[0296] V = (M - L1) / L1 x 100 (5)

[0297] Alternatively, the evaluation value acquisition section 12 can also calculate the overshoot amount V of the time series data in accordance with the equation (6).

[0298] V = M - L1... (6)

[0299] In addition, the evaluation value acquisition method is merely an example, and is not limited to these methods. For example, the evaluation value can be acquired by using the reference data RD as a reference, as described with reference to FIG. 17. Figures 4(a) to 4(c) As described above, the evaluation value is acquired by accumulating the square of the difference between the value of the time series data TD and the value of the reference data RD.

[0300] In addition, the classification section 13 (e.g., the classification section 13) can classify the evaluation value Ev of each of the plurality of time series data TD into any one of a plurality of classes by processing the evaluation value Ev. Figure 1

[0301] Next, an example of the classification processing performed by the classification section 13 will be described with reference to FIG. 20(a) and FIG. 20(b). FIG. 20(a) is a graph showing an evaluation value distribution. The evaluation value distribution shows the frequency (number of times) of the evaluation value Ev of each of the plurality of time series data TD. In the graph of FIG. 20(a), the horizontal axis shows the magnitude of the evaluation value Ev, and the vertical axis shows the frequency (number of times or occurrence frequency) of the evaluation value Ev.

[0302] In FIG. 20(a), μ shows the average value of the plurality of evaluation values Ev, and σ shows the standard deviation. As shown in FIG. 20(a), the evaluation value distribution has a peak around the average value μ. The evaluation value distribution has a shape close to a normal distribution.

[0303] The classification section 13 normalizes the evaluation value distribution. Specifically, the classification section 13 generates a normalized distribution obtained by normalizing the evaluation value distribution in accordance with Equation (7).

[0304] [Equation 2]

[0305]

[0306] Here, Sold shows the evaluation value distribution, and Snew shows the normalized distribution.

[0307] FIG. 20(b) is a graph showing the normalized distribution. The normalized distribution is generated by performing normalization processing such that the average value μ of the evaluation value distribution becomes 0 and the standard deviation becomes 1.

[0308] For example, the normalized distribution is divided into L1 to L4. In the normalized distribution, if it is less than 1, it is set to level 1 (L1), if it is 1 or more and less than 2, it is set to level 2 (L2), if it is 2 or more and less than 3, it is set to level 3 (L3), and if it is 3 or more, it is set to level 4 (L4). Here, the level indicates the degree of abnormality.

[0309] In addition, here, the division is performed using the standard deviation. Therefore, the operator or the administrator can not input the threshold value at the time of abnormality determination individually.​

[0310] In addition, the description of Figures 18(a) to 20(b) is mainly described with respect to the evaluation value acquisition section 12 and the classification section 13, but can also be applied to the evaluation value acquisition section 112 and the classification section 113. Furthermore, it can also be applied to the clustering processing section 15 and / or the matching section 114.

[0311] In addition, Figure 10 In the data processing apparatus 100 of Figure 15 In the data processing apparatus 100 of

[0312] Next, the data processing apparatus 100 and the substrate processing apparatus 200 of the present embodiment will be described with reference to Figure 21 Figure 21 is a schematic view of the data processing apparatus 100 and the substrate processing apparatus 200.

[0313] As shown in Figure 21 , the data processing apparatus 100 includes the processing section 10 and the storage section 20. The processing section 10 includes, in addition to the data acquisition section 11, the evaluation value acquisition section 12, the classification section 13, the extraction section 14, and the clustering processing section 15, the data acquisition section 111, the evaluation value acquisition section 112, the classification section 113, the matching section 114, and the readout section 115. The storage section 20 stores the learning database.

[0314] The processing section 10 is capable of causing the learning database to be stored in the storage section 20 by the data acquisition section 11, the evaluation value acquisition section 12, the classification section 13, the extraction section 14, and the clustering processing section 15. Furthermore, the processing section 10 is capable of effectively utilizing the learning database stored in the storage section 20 by the data acquisition section 111, the evaluation value acquisition section 112, the classification section 113, the matching section 114, and the readout section 115.

[0315] In addition, generally, the substrate processing apparatus 200 is roughly classified into a single-wafer type or a batch type. The substrate processing apparatus 200 can be either a single-wafer type or a batch type.

[0316] Next, the data processing apparatus 100 and the substrate processing apparatus 200 of the present embodiment will be described with reference to Figure 22 Figure 22 is a schematic view of the data processing apparatus 100 and the substrate processing apparatus 200.

[0317] ​​The data processing apparatus 100 processes the time-series data TD generated in the substrate processing apparatus 200. Here, the substrate processing apparatus 200 is single-wafer type.

[0318] The substrate processing apparatus 200 processes the substrate W. The substrate processing apparatus 200 processes the substrate W in at least one of etching, surface treatment, property imparting, processing film formation, removal of at least a part of a film, and cleaning.

[0319] The substrate W includes, for example, a semiconductor wafer, a substrate for liquid crystal display devices, a substrate for plasma display devices, a substrate for Field Emission Display (FED), a substrate for optical discs, a substrate for magnetic discs, a substrate for magneto-optical discs, a substrate for photomasks, a ceramic substrate, and a substrate for solar batteries. The substrate W is, for example, a substantially circular plate. The substrate processing apparatus 200 processes the substrate W piece by piece.

[0320] The substrate processing apparatus 200 includes a chamber 220, a substrate holding portion 230, and a liquid supply portion 240 as the processing unit 210. The chamber 220 houses the substrate W. The substrate holding portion 230 holds the substrate W.

[0321] The chamber 220 is a substantially box shape having an internal space. The chamber 220 houses the substrate W. Typically, the chamber 220 is an air environment. However, a gas flow can be formed by a down flow or the like in the chamber 220.

[0322] Here, the substrate processing apparatus 200 is single-wafer type that processes the substrate W piece by piece, and the substrate W is housed in the chamber 220 piece by piece. The substrate W is housed in the chamber 220 and is processed in the chamber 220. In the chamber 220, at least a part of each of the substrate holding portion 230 and the liquid supply portion 240 is housed.

[0323] The substrate holding portion 230 holds the substrate W. For example, the substrate holding portion 230 clamps an end portion of the substrate W. The substrate holding portion 230 holds the substrate W horizontally in such a manner that an upper surface (a surface) Wa of the substrate W faces upward and a back surface (a lower surface) Wb of the substrate W faces downward vertically. Also, the substrate holding portion 230 rotates the substrate W while holding the substrate W. For example, the substrate W is rotated in a counterclockwise direction when viewed from above vertically.

[0324] The substrate holding portion 230 includes, for example, a spin base 231, a chuck member 232, a shaft 233, an electric motor 234, and a base 235. The spin base 231 is, for example, a plate shape extending along the XY plane. Here, the spin base 231 is a disc shape (a thin circular plate shape). The spin base 231 faces the substrate W.

[0325] The chuck member 232 is provided on the spin base 231. The chuck member 232 fixes (chucks) the substrate W. Typically, a plurality of chuck members 232 are provided on the spin base 231.

[0326] The shaft 233 can also be a hollow shaft. The shaft 233 extends in the vertical direction along the rotation axis Ax. The spin base 231 is coupled to the upper end of the shaft 233. The substrate W is placed above the spin base 231.

[0327] The shaft 233 extends downward from the central portion of the spin base 231. The electric motor 234 imparts a rotational force to the shaft 233. The shaft 233 rotates with respect to the base 235. The base 235 rotatably supports the shaft 233. The electric motor 234 rotates the shaft 233 in the rotation direction, thereby rotating the substrate W and the spin base 231 about the rotation axis Ax. The electric motor 234 is an example of a rotating member.

[0328] The liquid supply portion 240 supplies a liquid to the substrate W. Typically, the liquid supply portion 240 supplies a liquid to the upper surface Wa of the substrate W. The liquid includes, for example, a rinsing liquid or a chemical liquid.

[0329] The rinsing liquid can include any one of deionized water (DIW), carbonated water, electrolytic ion water, ozone water, ammonia water, hydrochloric acid water of a dilute concentration (for example, about 10 ppm to 100 ppm), or reduced water (hydrogen water).

[0330] The chemical liquid includes hydrofluoric acid. For example, the hydrofluoric acid can be heated to 40°C or higher and 70°C or lower, or can be heated to 50°C or higher and 60°C or lower. However, the hydrofluoric acid can not be heated. Also, the chemical liquid can include water or phosphoric acid.

[0331] Further, the chemical liquid can further include hydrogen peroxide water. Also, the chemical liquid can include SC1 (ammonia hydrogen peroxide water mixture), SC2 (hydrochloric acid hydrogen peroxide water mixture), or aqua regia (a mixture of concentrated hydrochloric acid and concentrated nitric acid).

[0332] The substrate processing apparatus 200 further includes a cup 250. The cup 250 recovers liquid scattered from the substrate W. The cup 250 is raised and lowered. For example, the cup 250 is raised to the side of the substrate W in the vertical upward direction during the period in which the liquid supply section 240 supplies liquid to the substrate W. At this time, the cup 250 recovers liquid scattered from the substrate W due to rotation of the substrate W. Further, the cup 250 is lowered from the side of the substrate W in the vertical downward direction when the period in which the liquid supply section 240 supplies liquid to the substrate W ends.

[0333] The control device 201 includes a control section 201a and a storage section 201b. The control section 201a controls the substrate holding section 230, the liquid supply section 240, and / or the cup 250. In one example, the control section 201a controls the electric motor 234, the valve 240b, and / or the cup 250.

[0334] The substrate processing apparatus 200 of the present embodiment is preferably used for manufacturing of semiconductor devices provided with a semiconductor. The substrate processing apparatus 200 can be preferably used for cleaning and / or processing (e.g., etching, property change, etc.) of semiconductor devices at the time of manufacturing of semiconductor devices.

[0335] As described above, the substrate processing apparatus 200 can also be a single-wafer type. In this case, it is preferable that the substrate processing apparatus 200 include a plurality of processing units 210.

[0336] Next, the data processing apparatus 100 and the substrate processing apparatus 200 of the present embodiment will be described with reference to Figure 23 Figure 23 is a schematic view of the data processing apparatus 100 and the substrate processing apparatus 200. Here, the substrate processing apparatus 200 includes a plurality of processing units 210. The processing units 210 each process a substrate W. The plurality of processing units 210 are arranged in a prescribed arrangement.

[0337] As shown in Figure 23 , the substrate processing apparatus 200 includes a plurality of processing units 210, a fluid cabinet LC, a fluid box LB, a plurality of load ports LP, an indexer robot IR, a center robot CR, and a control device 201. The control device 201 controls the load ports LP, the indexer robot IR, and the center robot CR. The control device 201 includes a control section 201a and a storage section 201b.

[0338] ​The load ports LP each stack-hold a plurality of substrates W. The transfer robot IR carries the substrates W between the load ports LP and the central robot CR. The central robot CR carries the substrates W between the transfer robot IR and the processing units 210. The processing units 210 each spout a liquid to the substrate W to process the substrate W. The liquid includes, for example, a processing liquid, a rinse liquid, and / or a medicine liquid. The fluid chamber LC holds the liquid. In addition, the fluid chamber LC can also hold a gas.

[0339] Specifically, the plurality of processing units 210 form a plurality of towers TW configured in a manner that the towers TW surround the central robot CR in plan view. Figure 1 Each tower TW includes a plurality of processing units 210 stacked vertically (three processing units 210 in the embodiment). Figure 1 The fluid tank LB corresponds to each of the plurality of towers TW. The liquid in the fluid chamber LC is supplied to all of the processing units 210 included in the tower TW corresponding to the fluid tank LB via any one of the fluid tanks LB. Also, the gas in the fluid chamber LC is supplied to all of the processing units 210 included in the tower TW corresponding to the fluid tank LB via any one of the fluid tanks LB.

[0340] The substrate processing apparatus 200 further includes a control device 201. The control device 201 controls various actions of the substrate processing apparatus 200.

[0341] The control device 201 includes a control section 201a and a storage section 201b. The control section 201a has a processor. The control section 201a has, for example, a central processing unit (CPU). Alternatively, the control section 201a can have a general-purpose processor.

[0342] The storage section 201b stores data and a computer program. The data includes recipe data. The recipe data includes information indicating a plurality of recipes. The plurality of recipes each specify a processing content and a processing flow of the substrate W.

[0343] The storage section 201b includes a main storage device and an auxiliary storage device. The main storage device is, for example, a semiconductor memory. The auxiliary storage device is, for example, a semiconductor memory and / or a hard disk drive. The storage section 201b can also include a removable medium. The control section 201a executes the computer program stored in the storage section 201b, thereby performing a substrate processing action.

[0344] In addition, in the description referring to Figure 22 and Figure 23 , the substrate processing apparatus 200 and / or the processing unit 210 are single-wafer type, but the embodiment is not limited thereto. The substrate processing apparatus 200 and / or the processing unit 210 can be batch type.

[0345] Referring to Figure 24 the data processing apparatus 100 and the substrate processing apparatus 200 of the present embodiment will be described. Figure 24 is a schematic view of the data processing apparatus 100 and the substrate processing apparatus 200 of the present embodiment. Here, the substrate processing apparatus 200 is of a batch type, and is capable of collectively processing a plurality of substrates W.

[0346] The substrate processing apparatus 200 includes a processing tank 260, a substrate holding section 270, and a control apparatus 201. The processing tank 260 stores a processing liquid L used for processing the substrate W.

[0347] The substrate holding section 270 holds the substrate W. A normal line direction of a main surface of the substrate W held by the substrate holding section 270 is parallel to the Y direction. The substrate holding section 270 moves the substrate W in a state of holding the substrate W. For example, the substrate holding section 270 moves in a vertical direction toward a vertically upper side or a vertically lower side in a state of holding the substrate W.

[0348] Typically, the substrate holding section 270 collectively holds a plurality of substrates W. Here, the plurality of substrates W are arranged in a line along the Y direction. Alternatively, the substrate holding section 270 can hold only one substrate W.

[0349] Figures 1 to 24 The data processing apparatus 100 illustrated is one that processes time-series data generated in one substrate processing apparatus 200, but the present embodiment is not limited thereto. The data processing apparatus 100 can process time-series data generated in substrate processing apparatuses 200 arranged at different sites.

[0350] Next, referring to Figure 25 the data processing apparatus 100 and the substrate processing apparatus 200 of the present embodiment will be described. Figure 25 is a schematic view of the data processing apparatus 100 and the substrate processing apparatus 200 of the present embodiment. As Figure 25 indicated, the substrate processing apparatus 200 includes a substrate processing apparatus 200A and a substrate processing apparatus 200B. The substrate processing apparatus 200A and the substrate processing apparatus 200B are communicably connected to the data processing apparatus 100.

[0351] Typically, the substrate processing apparatus 200A and the substrate processing apparatus 200B are arranged at sites separated from each other. For example, the substrate processing apparatus 200A and the substrate processing apparatus 200B can be arranged at different sites in the same country. Alternatively, the substrate processing apparatus 200A and the substrate processing apparatus 200B can be arranged in different countries.

[0352] The data processing apparatus 100 is, for example, a server. The substrate processing apparatus 200A is capable of performing information communication with the substrate processing apparatus 200B via the data processing apparatus 100, for example.

[0353] In addition, in the description with reference to Figure 25 the data processing apparatus 100 can communicate with the substrate processing apparatus 200A and the substrate processing apparatus 200B, but the substrate processing apparatus 200A and the substrate processing apparatus 200B can not exist at the same time.

[0354] The above describes the embodiments of the present application with reference to the drawings. However, the present application is not limited to the described embodiments, and can be implemented in various forms without departing from the gist thereof. Also, various applications can be formed by appropriately combining a plurality of constituent elements disclosed in the described embodiments. For example, several constituent elements can be deleted from all the constituent elements shown in the embodiments. Further, the constituent elements between different embodiments can be appropriately combined. For the convenience of understanding, the drawings schematically show each constituent element on the main body, and the thickness, length, number, interval, and the like of each constituent element shown in the drawings are sometimes different from the actual ones in order to facilitate the drawings. Also, the material, shape, size, and the like of each constituent element shown in the described embodiments are examples, and are not particularly limited, and various changes can be made without substantially departing from the effects of the present application.

Claims

1. A data processing method comprising the steps of: acquiring a plurality of time series data obtained by a substrate processing apparatus; acquiring an evaluation value of each of the plurality of time series data; classifying each of the plurality of time series data into any one of a plurality of categories based on the evaluation value; extracting time series data corresponding to a category in which time series data exhibiting a tendency of abnormality is included among the plurality of categories, as extracted time series data; classifying the extracted time series data into any one of a plurality of clusters by cluster processing; changing control of the substrate processing apparatus; acquiring time series data from the substrate processing apparatus whose control has been changed, as changed time series data; acquiring an evaluation value of the changed time series data; classifying the changed time series data into any one of the plurality of categories based on the evaluation value of the changed time series data; in a case where the changed time series data is classified into the category, extracting the changed time series data as the extracted time series data; in a case where the changed time series data is not classified into the category, storing information related to a change content of the substrate processing apparatus whose control has been changed; and generating a learning database in which a cluster corresponding to the extracted time series data is given information on a change content of the substrate processing apparatus corresponding to the time series data, that is, cause and countermeasure information, the plurality of clusters including a rise delay cluster, a wave peak variation cluster, and a low wave peak cluster, the rise delay cluster including a processing liquid concentration drop / valve retightening as the cause and countermeasure information, the wave peak variation cluster including a substrate holding force drop / overcurrent suppression as the cause and countermeasure information, and the low wave peak cluster including an environment unevenness / sensor restart as the cause and countermeasure information.

2. The data processing method according to claim 1, wherein in the step of acquiring the evaluation value, the evaluation value of each of the plurality of time series data is acquired by comparison of each of the plurality of time series data with reference data.

3. The data processing method according to claim 2, wherein in the step of acquiring the evaluation value, the evaluation value in a case where a difference between the time series data and a value of the reference data is large is larger than the evaluation value in a case where the difference is small.

4. The data processing method according to claim 1 or 2, further comprising the step of: switching display of an evaluation value chart and the extracted time series data, the evaluation value chart indicating a change in the evaluation value of the extracted time series data.

5. The data processing method according to claim 1 or 2, further comprising the step of: switching display of an evaluation value chart and the extracted time series data, the evaluation value chart indicating an evaluation value of the extracted time series data classified into any one of the plurality of clusters.

6. A storage medium storing a program that causes a computer to execute the steps of: ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ acquiring a plurality of time series data obtained by a substrate processing apparatus; acquiring an evaluation value of each of the plurality of time series data; classifying the plurality of time series data into any one of a plurality of categories based on the evaluation value; extracting time series data corresponding to a category in which time series data showing a tendency to exhibit abnormality, i.e., an object category, from among the plurality of categories, as extracted time series data; performing clustering processing on the extracted time series data to classify the extracted time series data into any one of a plurality of clusters; changing control of the substrate processing apparatus; acquiring time series data from the substrate processing apparatus after the control change, as changed time series data; acquiring an evaluation value of the changed time series data; classifying the changed time series data into any one of the plurality of categories based on the evaluation value of the changed time series data; in a case where the changed time series data is classified into the object category, extracting the changed time series data as the extracted time series data; in a case where the changed time series data is not classified into the object category, storing information related to a change content of the substrate processing apparatus after the control change; and generating a learning database in which a cluster corresponding to the extracted time series data is given information on a change content of the substrate processing apparatus corresponding to the time series data, i.e., cause and countermeasure information, the plurality of clusters include a rise delay cluster, a wave peak variation cluster, and a low wave peak cluster, the rise delay cluster contains a processing liquid concentration drop / valve retightening as the cause and countermeasure information, the wave peak variation cluster contains a substrate holding force drop / overcurrent suppression as the cause and countermeasure information, and the low wave peak cluster contains an environment unevenness / sensor restart as the cause and countermeasure information.

Images