Feature extraction method, program, and device

Abstract

To provide a feature extraction method, a program, and a device that can extract feature quantities more appropriately.SOLUTION: The feature amount extraction method for extracting feature amounts from multiple unit data, where the multiple unit data are obtained by associating measurement data measured by a sensor with a predetermined physical quantity, and serve as a unit of processing. The feature amount extraction method includes: a change point detection step (step S3) of detecting change points of the measurement data in the multiple unit data; a position matching process (step S4) of matching positions of a physical quantity of the multiple unit data using the detected change point as an origin; and a feature extraction step (step S5) of extracting the feature amounts from the multiple unit data while matching the positions of the physical quantity.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The present invention relates to a feature extraction method, a program, and an apparatus for extracting features from a plurality of unit data. [Background technology]

[0002] For example, industrial production facilities are equipped with sensors such as temperature sensors, pressure sensors, and flow rate sensors. The measurement data obtained by each sensor is stored in association with a predetermined physical quantity such as time. When analyzing the stored measurement data, the difference between unit data, which is the unit of processing, becomes an important feature.

[0003] Conventionally, multiple unit data are extracted from continuously accumulated measurement data, the unit data are superimposed at positions derived from the analyst's knowledge and / or insight, graphed, and the analyst visually checks to extract areas where differences are observed between the unit data, generating feature quantities.The generated feature quantities are analyzed using machine learning and statistical analysis methods to identify feature quantities that affect product quality and equipment operation, and are used to identify the causes of quality defects and develop productivity improvement measures.

[0004] For example, Patent Documents 1 to 4 listed below disclose methods for extracting feature amounts from unit data. [Prior art documents] [Patent documents]

[0005] [Patent Document 1] Patent No. 4756707 [Patent Document 2] Patent No. 4837520 [Patent Document 3] Patent No. 5348977 [Patent Document 4] Patent No. 5912880 Summary of the Invention [Problem to be solved by the invention]

[0006] The conventional feature extraction methods described above have had the following problems. For example, unit data in industrial production equipment may start when the production equipment starts operating. However, even if the start points of each unit data are used as the origin and the positions of the physical quantities of each unit data are aligned and compared, differences between the unit data may not be clearly detected. For example, by overlapping each unit data set with the physical quantity position midway between the unit data as the origin, differences between the unit data that would be hidden when aligned with the start points as the origin may become clear. That is, unit data may contain extra data, and the extra data associated with the unit data may hide differences between the unit data that should be extracted.

[0007] There was also another problem, as described below. That is, analysts would gradually slide the unit data along the axis of the physical quantity until the curve drawn by the unit data matched most closely, and the remaining differences between the unit data were used as features. However, the determination that the curve drawn by the unit data best matches depends on the analyst's experience, is not quantified, and can have low reproducibility.

[0008] Furthermore, there was another problem as follows. Namely, the generation of features also depended on the analyst's awareness, and there was a risk of oversight. In other words, it was not possible to generate features that covered all of the unit data.

[0009] The present invention has been made to solve at least one of the above-described problems, and one of its objects is to provide a feature extraction method, program, and device that can extract features more appropriately. [Means for solving the problem]

[0010] A feature extraction method according to one aspect of the present invention is a feature extraction method for extracting feature values ​​from a plurality of unit data, wherein the plurality of unit data are measurement data measured by a sensor associated with a predetermined physical quantity and serve as a unit of processing, and includes a change point detection step for detecting a change point in the measurement data of the plurality of unit data, an alignment step for aligning the positions of the physical quantities of the plurality of unit data with the detected change point as an origin, and a feature extraction step for extracting feature values ​​from the plurality of unit data with the positions of the physical quantities aligned.

[0011] A feature extraction method according to another aspect of the present invention is a feature extraction method for extracting feature amounts from a plurality of unit data, wherein the plurality of unit data are measurement data measured by a sensor associated with a predetermined physical quantity and serve as units of processing. The feature extraction method includes an alignment step of aligning the physical quantities of the plurality of unit data, and a feature extraction step of extracting feature amounts from the plurality of unit data with the physical quantities aligned. The alignment step includes a first alignment step of aligning the physical quantities of the plurality of unit data with the position of an arbitrary physical quantity as an origin, a second alignment step of calculating a difference area between the plurality of unit data whose physical quantities have been aligned, a third alignment step of calculating a difference area sum by adding up the difference areas, a fourth alignment step of repeating the first to third alignment steps while changing the origin, and a fifth alignment step of determining the origin at which the difference area sum is minimum after repeating the first to third alignment steps in the fourth alignment step. In the feature extraction step, feature amounts are extracted from the plurality of unit data with the physical quantities aligned at the origin at which the difference area sum is minimum.

[0012] A feature extraction method according to yet another aspect of the present invention is a feature extraction method for extracting features from a plurality of unit data, wherein the plurality of unit data are measurement data measured by a sensor associated with a predetermined physical quantity and serve as units of processing, and includes an alignment step of aligning the positions of the physical quantities of the plurality of unit data, and a feature extraction step of extracting features from the plurality of unit data in a state in which the positions of the physical quantities are aligned, wherein the feature extraction step divides the plurality of unit data into a plurality of intervals with an arbitrary physical quantity width, and calculates a predetermined statistical quantity as a feature for each of the plurality of intervals.

[0013] A feature extraction program according to one embodiment of the present invention is a feature extraction program for extracting features from a plurality of unit data, wherein the plurality of unit data are measurement data measured by a sensor associated with a predetermined physical quantity and serve as units of processing, and the program causes a computer to realize a change point detection function for detecting change points in the measurement data of the plurality of unit data, an alignment function for aligning the positions of the physical quantities of the plurality of unit data using the detected change point as the origin, and a feature extraction function for extracting features from the plurality of unit data with the positions of the physical quantities aligned.

[0014] A feature extraction program according to another aspect of the present invention is a feature extraction program for extracting feature amounts from a plurality of unit data, wherein the plurality of unit data are measurement data measured by a sensor associated with a predetermined physical quantity and serve as units of processing. The feature extraction program is for causing a computer to realize an alignment function for aligning the positions of the physical quantities of the plurality of unit data and a feature extraction function for extracting feature amounts from the plurality of unit data with the physical quantities aligned. The alignment function includes: a first alignment function for aligning the positions of the physical quantities of the plurality of unit data with the position of an arbitrary physical quantity as an origin; a second alignment function for calculating a difference area between the plurality of unit data after the physical quantities have been aligned; a third alignment function for calculating a difference area sum by adding up the difference areas; a fourth alignment function for repeating the first to third alignment functions while changing the origin; and a fifth alignment function for determining the origin at which the difference area sum is minimum after the fourth alignment function has repeated the first to third alignment functions. The feature extraction function extracts feature amounts from the plurality of unit data with the physical quantities aligned at the origin at which the difference area sum is minimum.

[0015] A feature extraction program according to yet another aspect of the present invention is a feature extraction program for extracting features from a plurality of unit data, wherein the plurality of unit data are measurement data measured by a sensor associated with a predetermined physical quantity and serve as units of processing, and the feature extraction program is for causing a computer to realize an alignment function for aligning the positions of the physical quantities of the plurality of unit data and a feature extraction function for extracting features from the plurality of unit data with the positions of the physical quantities aligned, and the feature extraction function divides the plurality of unit data into a plurality of intervals with an arbitrary physical quantity width and calculates a predetermined statistical quantity as a feature for each of the plurality of intervals.

[0016] A feature extraction device according to one aspect of the present invention is a feature extraction device for extracting feature values ​​from a plurality of unit data, wherein the plurality of unit data are measurement data measured by a sensor associated with a predetermined physical quantity and serve as a unit of processing. The feature extraction device includes a change point detection unit that detects change points in the measurement data of the plurality of unit data, an alignment unit that aligns the positions of the physical quantities of the plurality of unit data using the detected change point as an origin, and a feature extraction unit that extracts feature values ​​from the plurality of unit data with the positions of the physical quantities aligned.

[0017] A feature extraction device according to another aspect of the present invention is a feature extraction device for extracting feature amounts from a plurality of unit data, wherein the plurality of unit data are measurement data measured by a sensor associated with a predetermined physical quantity and serve as units of processing. The feature extraction device includes an alignment unit that aligns the positions of the physical quantities of the plurality of unit data, and a feature extraction unit that extracts feature amounts from the plurality of unit data with the physical quantities aligned. The alignment unit includes a first alignment unit that aligns the positions of the physical quantities of the plurality of unit data with the position of an arbitrary physical quantity as an origin, a second alignment unit that calculates a difference area between the plurality of unit data after the physical quantities have been aligned, a third alignment unit that calculates a difference area sum by adding up the difference areas, a fourth alignment unit that repeats the processing of the first to third alignment units while changing the origin, and a fifth alignment unit that determines the origin at which the difference area sum is minimum after the fourth alignment unit has repeated the processing of the first to third alignment units. The feature extraction unit extracts feature amounts from the plurality of unit data with the physical quantities aligned at the origin at which the difference area sum is minimum.

[0018] A feature extraction device according to yet another aspect of the present invention is a feature extraction device for extracting features from a plurality of unit data, wherein the plurality of unit data are measurement data measured by a sensor associated with a predetermined physical quantity and serve as units of processing, and includes an alignment unit that aligns the positions of the physical quantities of the plurality of unit data, and a feature extraction unit that extracts features from the plurality of unit data with the physical quantities aligned, wherein the feature extraction unit divides the plurality of unit data into a plurality of intervals with an arbitrary physical quantity width, and calculates a predetermined statistical quantity as a feature for each of the plurality of intervals. [Effects of the Invention]

[0019] According to a feature extraction method, program, and device of one aspect of the present invention, features are extracted from multiple unit data while aligning the positions of the physical quantities of the multiple unit data, with the change point of the detected measurement data as the origin.This makes it possible to avoid differences between the unit data being hidden by extraneous data accompanying the unit data, and allows for more appropriate feature extraction.

[0020] Furthermore, according to a feature extraction method, program, and device according to another aspect of the present invention, feature values ​​are extracted from a plurality of unit data while the positions of the physical quantities are aligned at the origin where the sum of the differential areas is smallest. This allows feature values ​​to be extracted in a state where the curves drawn by the unit data are considered to be most consistent, without relying on the experience of the analyst, and allows for more appropriate feature extraction.

[0021] Furthermore, according to a feature extraction method, program, and device according to yet another aspect of the present invention, a plurality of unit data are divided into a plurality of intervals using an arbitrary physical quantity width, and a predetermined statistical quantity is calculated as a feature for each of the plurality of intervals, thereby making it possible to generate features that cover the entirety of the unit data, and to extract features more appropriately. [Brief explanation of the drawings]

[0022] [Figure 1] 1 is a flowchart showing a feature extraction method according to the first embodiment of the present invention. [Figure 2]FIG. 2 is an explanatory diagram showing the unit data extraction step in FIG. [Figure 3] FIG. 2 is an explanatory diagram showing the change-point detection process of FIG. [Figure 4] 2 is a flowchart showing the alignment process of FIG. 1 in more detail. [Figure 5] FIG. 5 is an explanatory diagram showing the first alignment step of FIG. 4. [Figure 6] FIG. 5 is an explanatory view showing the second alignment step of FIG. 4. [Figure 7] FIG. 5 is an explanatory view showing the fifth alignment step of FIG. 4. [Figure 8] FIG. 2 is an explanatory diagram showing the feature extraction process of FIG. 1. [Figure 9] FIG. 2 is a block diagram showing a feature extraction program according to the first embodiment of the present invention. [Figure 10] 1 is a block diagram showing a feature extraction device according to a first embodiment of the present invention. [Figure 11] 10 is a flowchart showing a feature extraction method according to second and third embodiments of the present invention. [Figure 12] FIG. 10 is a block diagram showing a feature extraction program according to a second embodiment of the present invention. [Figure 13] FIG. 10 is a block diagram showing a feature extraction device according to a second embodiment of the present invention. [Figure 14] FIG. 11 is a block diagram showing a feature extraction program according to a third embodiment of the present invention. [Figure 15] FIG. 10 is a block diagram showing a feature extraction device according to a third embodiment of the present invention. DETAILED DESCRIPTION OF THE INVENTION

[0023] Hereinafter, embodiments of the present invention will be described with reference to the drawings. The present invention is not limited to each embodiment, and the components can be modified and embodied without departing from the spirit of the present invention. Furthermore, various inventions can be formed by appropriately combining multiple components disclosed in each embodiment. For example, some components may be omitted from all the components shown in the embodiments. Furthermore, components of different embodiments may be appropriately combined.

[0024] Embodiment 1 First, a feature extraction method according to a first embodiment of the present invention will be described with reference to Figs. 1 to 8. Fig. 1 is a flowchart showing the feature extraction method according to the first embodiment of the present invention, Fig. 2 is an explanatory diagram showing the unit data extraction step of Fig. 1, and Fig. 3 is an explanatory diagram showing the change-point detection step of Fig. 1. Fig. 4 is a flowchart showing the alignment step of Fig. 1 in more detail, Fig. 5 is an explanatory diagram showing the first alignment step of Fig. 4, Fig. 6 is an explanatory diagram showing the second alignment step of Fig. 4, and Fig. 7 is an explanatory diagram showing the fifth alignment step of Fig. 4. Fig. 8 is an explanatory diagram showing the feature extraction step of Fig. 1.

[0025] The feature extraction method shown in Fig. 1 is a method for extracting feature values ​​from a plurality of unit data. The unit data is measurement data measured by a sensor associated with a predetermined physical quantity, and serves as a unit of processing. In this embodiment, the plurality of unit data are assumed to be continuously stored in a database.

[0026] The sensor may be any sensor, such as a temperature sensor, a pressure sensor, or a flow rate sensor. Therefore, the measurement data measured by the sensor may correspond to any physical quantity (first physical quantity), such as temperature, pressure, or flow rate. The physical quantity (second physical quantity) to which the measurement data is associated is typically time, but may also be other measurement data measured by another sensor. For example, the temperature at a predetermined location may be associated with the pressure or flow rate at that location, or the temperature at a predetermined location may be associated with the temperature, pressure, or flow rate at another location. In other words, the first physical quantity measured by the sensor may be associated with a second physical quantity of a different type from the first physical quantity, or may be associated with the same type of second physical quantity.

[0027] Another example of a physical quantity (second physical quantity) with which measurement data is associated is a physical position or distance. For example, in a tunnel kiln for continuously heat-treating industrial products along a predetermined trajectory, the temperature (first physical quantity) measured at each location may be associated with the distance (second physical quantity) from the entrance of the tunnel kiln. Yet another example of a physical quantity (second physical quantity) with which measurement data is associated is a wavelength. For example, in a gas analyzer or frequency analyzer for investigating the properties of a test sample, the wavelength (second physical quantity) may be associated with the strength of the response at that wavelength (first physical quantity).

[0028] A database can be understood as a set of data in which measurement data such as those described above is continuously stored. Continuous does not necessarily mean uninterrupted in an analog sense, but may also include discrete data such as measurement data at predetermined time intervals. The database can be stored in any hardware, such as a hard disk, memory, computer, or server. The feature extraction method can be performed by any device, such as a computer or dedicated circuit. The device that performs the feature extraction method may include hardware that stores the database, or may be provided separately from the hardware and connected to it.

[0029] Although the sensor can measure any object, in this embodiment, the sensor is described as measuring a batch-type production facility for industrial products. More specifically, an example will be described in which a sensor measures the temperature at a predetermined position in a heat treatment furnace that heat treats industrial products for a predetermined time, and the sensor measurement data is associated with time and continuously stored in a database. The temperature of the heat treatment furnace is raised after an industrial product is inserted into the heat treatment furnace, maintained at a high temperature for a predetermined period of time, and then lowered. The temperature of the heat treatment furnace is maintained at a predetermined low temperature until the next industrial product to be treated is inserted into the heat treatment furnace.

[0030] A database showing such temperature changes in a heat treatment furnace is shown in Figure 2. More specifically, the database shown in Figure 2 includes first unit data D1 relating to the production of a first industrial product, second unit data D2 relating to the production of a second industrial product, and third unit data D3 relating to the production of a third industrial product. The first to third industrial products are the same type of products that are subjected to the same heat treatment.

[0031] As shown in Figure 2, the temperature may be increased and decreased stepwise rather than at a uniform rate. In the example shown in Figure 2, after an industrial product is inserted into the heat treatment furnace, the temperature of the heat treatment furnace is increased from a first temperature T1 to a second temperature T2 at a predetermined rate and maintained at the second temperature T2 for a predetermined period of time. Next, the temperature is increased from the second temperature T2 to a third temperature T3 at a predetermined rate and maintained at the third temperature T3 for a predetermined period of time. Then, the temperature is decreased from the third temperature T3 to the first temperature T1. When decreasing from the third temperature T3 to the first temperature T1, the temperature is decreased at a predetermined first rate and then at a second rate that is slower than the first rate.

[0032] A processing unit may be a set of measurement data associated with a predetermined physical quantity, intended to find differences between the sets of measurement data. A processing unit may be a set of measurement data during each process when the same type of process is repeatedly performed. More specifically, a processing unit may be a set of measurement data during each batch process in a batch production facility for industrial products. A processing unit may also be a set of measurement data from the inlet to the outlet in a continuous production facility for industrial products, such as a tunnel kiln. A processing unit may also be a set of measurement data during each measurement process in a test sample measurement device, such as a gas analyzer or a frequency analyzer.

[0033] 1, the feature extraction method of the first embodiment includes a unit data extraction step (step S1), a noise removal step (step S2), a change point detection step (step S3), a registration step (step S4), and a feature extraction step (step S5). Each step can be performed in this order.

[0034] The unit data extraction step (step S1) is a step of extracting a plurality of unit data serving as processing units from a database. In this step, first to fourth unit data are extracted from the database shown in FIG. 2. For example, the timing at which a state of a first temperature T1 is maintained for a predetermined time (the timing at which a first physical quantity is maintained in a first state for a predetermined time) is recognized as the start point of the unit data, and the timing at which the state of the first temperature T1 is maintained again for a predetermined time from the start point through second and third temperatures T2 and T3 is recognized as the end point of the unit data, thereby extracting the unit data. For example, the first temperature T1 or a temperature between the first temperature T1 and the second temperature T2 can be set as a threshold, and the timing at which the state at the first temperature T1 is maintained for a predetermined time can be set as the threshold. Data not included in the first to fourth unit data may exist between each of the first to fourth unit data. Alternatively, the end point of one unit data may be set as the start point of the next unit data.

[0035] The noise removal step (step S2) is a step of removing noise from the extracted multiple unit data. Noise is high-frequency fluctuation contained in the measurement data, and can be removed by, for example, obtaining a moving average of the measurement data or applying a low-pass filter to the measurement data.

[0036] The change-point detection step (step S3) is a step of detecting change points in the measurement data in the plurality of unit data from which noise has been removed. A change point in the measurement data can be detected as a timing at which the second-order differential values ​​of the plurality of unit data exceed a predetermined value or become discontinuous. FIG. 3A shows the unit data from which noise has been removed, FIG. 3B shows the first-order differential values ​​of the unit data, and FIG. 3C shows the relationship between the second-order differential values ​​of the unit data and the pulse P contained in the second-order differential values. As shown in FIG. 3C, a pulse P appears in the second-order differential value of the unit data at the position of a change point in the measurement data contained in the unit data. By detecting the timing at which the second-order differential value exceeds a predetermined value or becomes discontinuous, the position of the pulse P can be detected, and the position of the pulse P can be detected as a change point. Alternatively, a change point may be detected by other methods, such as detecting a point at which the measurement data of the noise-removed unit data changes by more than a predetermined value, or detecting a timing at which the first-order differential value of the unit data becomes discontinuous.

[0037] The alignment step (step S4) is a step of aligning the positions of the physical quantities of the plurality of unit data from which noise has been removed, with the detected change point as the origin. In this embodiment, the position of the time axis of the unit data is adjusted so that the time positions (positions of the physical quantities) detected as change points coincide with each other. FIG. 5 shows an example in which a plurality of unit data whose positions on the time axis have been adjusted are superimposed. When a plurality of change points are detected in the change point detection step, a specific change point may be selected at least initially according to a predetermined rule, and the selected change point may be used as the origin for alignment. In the example of FIG. 5, the change point at the timing when the temperature reaches the second temperature T2 is set as the origin.

[0038] As shown in FIG. 4, the alignment process (step S4) in the feature extraction method of this embodiment 1 includes a first alignment process (step S4-1), a second alignment process (step S4-2), a third alignment process (step S4-3), a fourth alignment process (step S4-4), and a fifth alignment process (step S4-5).

[0039] The first alignment step (step S4-1) is a step of aligning the positions of the physical quantities of the plurality of unit data from which noise has been removed, with the change point as the origin. The processing in this step is the same as the processing in the alignment step described above.

[0040] The second alignment step (step S4-2) is a step of calculating the difference area between a plurality of unit data whose physical quantities have been aligned. As shown in FIG. 6, the difference area can be understood as the sum of the areas of the differences between the unit data when the measurement data (temperature) is on the vertical axis and the physical quantity associated with the measurement data is on the horizontal axis. In other words, it can be understood as the sum of the differences between the measurement data of one unit data and the measurement data of another unit data at each physical quantity position. The difference between the measurement data of one unit data and the measurement data of another unit data may be treated as an absolute value before the sum is calculated.

[0041] In the second alignment process of this embodiment, after determining average data (average unit data) of the aligned multiple unit data, the difference area between the average data and the multiple unit data is determined. The average data can be determined by averaging the measurement data (temperature) of each unit data at each physical quantity (time) position. Figure 6 shows the difference area between the average data and the second unit data. Alternatively, one of the multiple unit data may be selected according to a predetermined rule, and the difference area between the selected unit data and the other unit data may be determined.

[0042] The third alignment step (step S4-3) is a step of adding up the difference areas to find the sum of the difference areas. In the third alignment step of this embodiment, the sum of the difference areas is found by adding up the first difference area between the average data and the first unit data, the second difference area between the average data and the second unit data, and the third difference area between the average data and the third unit data.

[0043] The fourth alignment step (step S4-4) is a step in which the first to third alignment steps are repeated while changing the origin. The origin can be changed by shifting it from the current origin by a predetermined physical quantity width (time width). In the initial first alignment step, the unit data is roughly aligned by using the change point as the origin, and in the fourth alignment step, the origin is gradually changed (i.e., the position of the unit data is finely adjusted) to find the minimum difference area sum, and the difference area sum can be collected. Alternatively, the origin may be sequentially changed to a change point different from the change point used in the initial first alignment step.

[0044] The fifth alignment step (step S4-5) is a step of determining the origin at which the differential area sum is minimum after repeating the first to third alignment steps in the fourth alignment step. Fig. 7 shows an example of a graph showing the relationship between the differential area sum collected in the fourth alignment step and the value of the physical quantity (time) used as the origin when calculating the differential area sum. The origin at which the differential area sum is minimum is the origin at which the deviation between each unit data is smallest, and can be used as the optimal origin for correcting the deviation between each unit data.

[0045] 1, the feature extraction step (step S5) is a step of extracting feature amounts from a plurality of unit data in a state where the positions of the physical quantities are aligned. In the feature extraction step of this embodiment, feature amounts are extracted from a plurality of unit data in a state where the positions of the physical quantities are aligned at the origin determined in the fifth alignment step (the origin where the difference area sum is smallest).

[0046] The extracted feature quantities and the extraction method are arbitrarily determined depending on the processing purpose of the unit data. In the feature quantity extraction process of this embodiment, as shown in FIG. 8 , a plurality of unit data are divided into a plurality of intervals S by an arbitrary physical quantity width, and a predetermined statistical quantity is calculated as a feature quantity for each of the plurality of intervals S. The physical quantity width used for dividing into intervals S can be set appropriately depending on the processing purpose of the unit data and / or the data acquisition period, for example, every 5 minutes when the data interval of the unit data is 1 minute. Furthermore, considering that statistical quantities will be calculated in subsequent processing, the physical quantity width used for dividing into intervals S is preferably set so that three or more measurement data points are included in one interval. The intervals S can be set for each predetermined physical quantity width based on the origin. The intervals S can be set adjacent to each other. The entire physical quantity range of the unit data may be divided into intervals S. Alternatively, a portion of the physical quantity range of the unit data, such as a portion of the range where the measurement data exceeds a predetermined value (i.e., a portion of the range where significant measurement data is included), may be divided into intervals S. Examples of the calculated statistics include the average value and / or the slope (rate of change) of the measurement data (temperature) for each section S. By comparing and examining the feature values ​​extracted from each of the aligned unit data, the differences in events in each unit data can be more reliably understood.

[0047] Next, FIG. 9 is a block diagram showing a feature extraction program 1 according to the first embodiment of the present invention. The feature extraction program 1 shown in FIG. 9 is for extracting feature values ​​of unit data, which serve as processing units, from a database 2 in which measurement data measured by a sensor is associated with a predetermined physical quantity and continuously stored. The feature extraction program 1 is loaded into a computer 3 and executed to cause the computer 3 to perform functions corresponding to each step of the feature extraction method described above. The database 2 shown in FIG. 9 is the measurement data measured by a sensor that is associated with a predetermined physical quantity and continuously stored. Although the database 2 is shown external to the computer 3 in FIG. 9, the database 2 may also be stored in hardware included in the computer 3. Furthermore, although a single computer 3 is shown in FIG. 9, the computer 3 shown in FIG. 9 may be implemented by multiple computers. In this case, some functions may be implemented by one computer and other functions may be implemented by another computer.

[0048] When the feature extraction program 1 is loaded into the computer 3 and executed, the functions realized by the computer 3 include a unit data extraction function 31 that extracts multiple unit data from the database 2, a noise removal function 32 that removes noise from the extracted multiple unit data, a change point detection function 33 that detects change points in the measurement data in the multiple unit data from which noise has been removed, an alignment function 34 that aligns the positions of the physical quantities of the multiple unit data from which noise has been removed, using the detected change point as the origin, and a feature extraction function 35 that extracts features from the multiple unit data with the positions of the physical quantities aligned.

[0049] The change point detection function 33 can detect, as a change point, the timing at which the second-order differential values ​​of a plurality of unit data exceed a predetermined value or become discontinuous.

[0050] The alignment function 34 includes a first alignment function 341 that aligns the positions of the physical quantities of multiple unit data from which noise has been removed, using the change point as the origin; a second alignment function 342 that calculates the difference area between multiple unit data whose physical quantities have been aligned; a third alignment function 343 that adds up the difference areas to calculate the difference area sum; a fourth alignment function 344 that repeats the first to third alignment functions 341 to 343 while changing the origin; and a fifth alignment function 345 that determines the origin at which the difference area sum is minimum after the fourth alignment function 344 repeats the first to third alignment functions 341 to 343.The feature extraction function 35 can extract features from the multiple unit data with the positions of the physical quantities aligned at the origin at which the difference area sum is minimum.

[0051] The second alignment function 342 can calculate the average data of the aligned plurality of unit data, and then calculate the difference area between the average data and the plurality of unit data.

[0052] The feature extraction function 35 can divide a plurality of unit data into a plurality of intervals S with an arbitrary physical quantity width, and calculate a predetermined statistical quantity as a feature for each of the plurality of intervals S.

[0053] As described above, each function corresponds to a step in the feature extraction method, and the explanation of each step in the feature extraction method also applies to each function.

[0054] Next, FIG. 10 is a block diagram showing a feature extraction device 4 according to the first embodiment of the present invention. The feature extraction device 4 shown in FIG. 10 extracts features of unit data, which serve as processing units, from a database 2 in which measurement data measured by a sensor is associated with a predetermined physical quantity and continuously stored. The feature extraction device 4 includes units (functional blocks) for executing the steps of the feature extraction method described above. The database 2 shown in FIG. 10 is the measurement data measured by a sensor that is associated with a predetermined physical quantity and continuously stored. Although the database 2 is shown outside the feature extraction device 4 in FIG. 10, the database 2 may be stored in hardware included in the feature extraction device 4. Furthermore, although a single feature extraction device 4 is shown in FIG. 10, the feature extraction device 4 shown in FIG. 10 may be realized by multiple devices. In this case, some functional blocks may be realized by one device and other functional blocks may be realized by another device.

[0055] The feature extraction device 4 includes a unit data extraction unit 41 that extracts a plurality of unit data from the database 2, a noise removal unit 42 that removes noise from the extracted plurality of unit data, a change point detection unit 43 that detects change points in the measurement data in the plurality of unit data from which noise has been removed, a position alignment unit 44 that aligns the positions of the physical quantities of the plurality of unit data from which noise has been removed, using the detected change point as the origin, and a feature extraction unit 45 that extracts feature quantities from the plurality of unit data with the positions of the physical quantities aligned.

[0056] The change point detection unit 43 can detect, as a change point, the timing at which the second-order differential values ​​of a plurality of unit data exceed a predetermined value or become discontinuous.

[0057] The alignment unit 44 includes a first alignment unit 441 that aligns the positions of the physical quantities of multiple unit data from which noise has been removed, using the change point as the origin; a second alignment unit 442 that calculates the difference area between multiple unit data whose physical quantities have been aligned; a third alignment unit 443 that adds up the difference areas to calculate the difference area sum; a fourth alignment unit 444 that repeats the first to third alignment units 441 to 443 while changing the origin; and a fifth alignment unit 445 that determines the origin at which the difference area sum is minimum after the fourth alignment unit 444 has repeated the first to third alignment units 441 to 443.The feature extraction unit 45 can extract features from the multiple unit data with the positions of the physical quantities aligned at the origin at which the difference area sum is minimum.

[0058] After determining the average data of the plurality of unit data that have been subjected to the alignment, the second alignment section 442 can determine the difference area between the average data and the plurality of unit data.

[0059] The feature extraction unit 45 can divide a plurality of unit data into a plurality of intervals S with an arbitrary physical quantity width, and calculate a predetermined statistical quantity as a feature for each of the plurality of intervals S.

[0060] As described above, each unit corresponds to each step of the feature extraction method, and the explanation of each step of the feature extraction method also applies to each unit.

[0061] In the feature extraction method, program, and device of this embodiment, feature values ​​are extracted from multiple unit data while aligning the positions of the physical quantities of the multiple unit data, with the change point of the detected measurement data as the origin. This makes it possible to avoid differences between the unit data being hidden by extraneous data accompanying the unit data, and allows for more appropriate feature extraction.

[0062] Furthermore, since the timing at which the second-order differential values ​​of a plurality of unit data items exceed a predetermined value or become discontinuous is detected as a change point, it is possible to more reliably detect a change point.

[0063] Furthermore, feature quantities are extracted from multiple unit data sets while aligning the physical quantities at the origin where the sum of the differential areas is minimized. This allows feature quantities to be extracted under conditions where the curves of the unit data sets are considered to be most consistent, without relying on the analyst's experience, resulting in more accurate feature quantity extraction. This configuration is particularly useful when the range of physical quantities between change points varies between unit data sets. For example, in a batch-type device involving a vacuum process, there is a program that transitions to the next step, such as heating, when a predetermined degree of vacuum is reached. The time required to reach a predetermined degree of vacuum may vary depending on the ambient environment, such as the season, which means that the range of physical quantities between change points may differ between unit data sets. In such cases, it is difficult to determine which change point is best to align the origin. The above configuration can solve this problem.

[0064] Furthermore, since the difference area between the average data and the plurality of unit data is calculated after calculating the average data of the plurality of unit data that have been aligned, the calculation load can be reduced compared to when calculating the difference area between all the unit data, and the risk of bias in the results can be reduced compared to when calculating the difference area based on a specific unit data.

[0065] Furthermore, multiple unit data are divided into multiple intervals S with an arbitrary physical quantity width, and a predetermined statistical quantity is calculated as a feature for each of the multiple intervals S, so that the feature can be generated by covering the entire unit data, and the feature can be extracted more appropriately.

[0066] Embodiment 2 FIG. 11 is a flowchart showing a feature extraction method according to a second embodiment of the present invention. The feature extraction method according to the second embodiment is a method in which the change-point detection step (step S3) according to the first embodiment is omitted. The alignment step (step S4) according to the second embodiment includes the first alignment step (step S4-1), the second alignment step (step S4-2), the third alignment step (step S4-3), the fourth alignment step (step S4-4), and the fifth alignment step (step S4-5) described with reference to FIG. 4. The first alignment step (step S4-1) according to the second embodiment aligns the positions of the physical quantities of a plurality of unit data from which noise has been removed, using the position of any physical quantity as the origin. That is, although the origin is determined based on the change point in the first embodiment, the origin may be determined according to another method / criteria, such as the physical quantity position specified by the operator. The rest is the same as the feature extraction method according to the first embodiment.

[0067] Next, FIG. 12 is a block diagram showing a feature extraction program 1 according to a second embodiment of the present invention. The feature extraction program 1 according to the second embodiment is a program in which the change-point detection function 33 is omitted from the functions to be implemented by a computer described in the first embodiment. The alignment function 34 according to the second embodiment includes a first alignment function 341, a second alignment function 342, a third alignment function 343, a fourth alignment function 344, and a fifth alignment function 345. The first alignment function 341 according to the second embodiment aligns the positions of physical quantities of multiple unit data from which noise has been removed, using the position of an arbitrary physical quantity as the origin. That is, while the origin is determined based on the change point in the first embodiment, the origin may be determined according to other methods / criteria, such as the position of a physical quantity specified by an operator. The remaining aspects are the same as those of the feature extraction program according to the first embodiment.

[0068] Next, FIG. 13 is a block diagram showing a feature extraction device 4 according to a second embodiment of the present invention. The feature extraction device according to the second embodiment is a device in which the change point detection unit 43 described in the first embodiment is omitted. The alignment unit 44 according to the second embodiment includes a first alignment unit 441, a second alignment unit 442, a third alignment unit 443, a fourth alignment unit 444, and a fifth alignment unit 445. The first alignment unit 441 according to the second embodiment aligns the positions of the physical quantities of a plurality of unit data from which noise has been removed, using the position of an arbitrary physical quantity as the origin. That is, while the origin is determined based on the change point in the first embodiment, the origin may be determined according to another method / criteria, such as the position of a physical quantity specified by an operator. The rest is the same as the feature extraction device according to the first embodiment.

[0069] In this way, the origin may be determined according to a method / criteria different from that of the first embodiment. In the feature extraction method, program, and device of the second embodiment, feature quantities are extracted from multiple unit data sets while aligning the physical quantities at the origin where the sum of the differential areas is minimized. Therefore, feature quantities can be extracted under conditions where the curves of the unit data sets are considered to be most consistent, without relying on the analyst's experience, resulting in more appropriate feature extraction. This configuration is particularly useful when the range of physical quantities between change points varies between unit data sets. For example, in a batch-type device involving a vacuum process, there is a program that transitions to the next step, such as heating, when a predetermined degree of vacuum is reached. The time required to reach a predetermined degree of vacuum may vary depending on the ambient environment, such as the season, and therefore the range of physical quantities between change points may differ between unit data sets. In such cases, it is difficult to determine which change point is best to align the origin. The above-described configuration can solve this problem.

[0070] Embodiment 3 The flowchart showing the feature extraction method of the third embodiment is similar to the flowchart showing the feature extraction method of the second embodiment, and therefore FIG. 11 is also used. That is, the feature extraction method of the third embodiment is a method in which the change-point detection step (step S3) of the first embodiment is omitted, as in the second embodiment. Furthermore, while the alignment step (step S4) of the second embodiment has been described as including the first to fifth alignment steps (steps S4-1 to S4-5), the alignment step (step S4) of the third embodiment does not include the first to fifth alignment steps (steps S4-1 to S4-5). In the alignment step (step S4), the positions of the physical quantities of the plurality of unit data from which noise has been removed are aligned by an arbitrary method / criteria, such as an operator's operation. In the feature extraction step (step S5), the plurality of unit data are divided into a plurality of sections S with an arbitrary physical quantity width, and a predetermined statistical quantity is calculated as a feature for each of the plurality of sections S. The rest is similar to the feature extraction methods of the first and second embodiments.

[0071] Next, FIG. 14 is a block diagram showing a feature extraction program 1 according to a third embodiment of the present invention. The feature extraction program of the third embodiment is a program in which the change-point detection function 33 and the first to fifth alignment functions 341 to 345 are omitted from the functions to be implemented by a computer described in the first embodiment. The alignment function 34 of the second embodiment aligns the positions of the physical quantities of a plurality of unit data from which noise has been removed, using any method / criteria such as an operator's operation. The feature extraction function 35 divides the plurality of unit data into a plurality of intervals S with any physical quantity width, and calculates a predetermined statistical quantity as a feature for each of the plurality of intervals S. The rest is the same as the feature extraction program of the first and second embodiments.

[0072] Next, Fig. 15 is a block diagram showing a feature extraction device 4 according to a third embodiment of the present invention. The feature extraction device according to the third embodiment is a device in which the change-point detection unit 43 and the first to fifth alignment units 441 to 445 described in the first embodiment are omitted. The alignment unit 44 according to the second embodiment aligns the positions of the physical quantities of a plurality of unit data from which noise has been removed, using any method / criterion such as an operator's operation. The feature extraction unit 45 divides the plurality of unit data into a plurality of intervals S with any physical quantity width, and calculates a predetermined statistical quantity as a feature for each of the plurality of intervals S. The rest is the same as the feature extraction devices according to the first and second embodiments.

[0073] In this way, the positions of the physical quantities of a plurality of unit data may be aligned according to a method / standard different from those in Embodiments 1 and 2. In the feature extraction method, program, and device as in Embodiment 3, a plurality of unit data are divided into a plurality of intervals S with an arbitrary physical quantity width, and a predetermined statistical quantity is calculated as a feature for each of the plurality of intervals S, so that features can be generated covering all of the unit data, and more appropriate feature extraction can be achieved.

[0074] In the first to third embodiments, it has been described that a plurality of unit data are continuously stored as a database, but it is also possible to extract features from a plurality of unit data that have been individually prepared in advance. That is, the unit data extraction step, the unit data extraction function, and the unit data extraction unit for extracting a plurality of unit data from a database may be omitted.

[0075] Furthermore, in the first to third embodiments, it has been described that noise is removed from a plurality of unit data, but in cases where there is extremely little noise in the plurality of unit data, for example, the noise removal process, noise removal function, and noise removal unit for removing noise may be omitted. [Explanation of symbols]

[0076] 1: Feature extraction program 2: Database 3: Computer 31: Unit data extraction function 32: Noise reduction function 33: Change point detection function 34: Alignment function 341: First alignment function 342: Second alignment function 343: Third alignment function 344: 4th alignment function 345: 5th alignment function 35: Feature extraction function 4: Feature extraction device 41: Unit data extraction section 42: Noise removal section 43: Change point detection unit 44: Alignment section 441: First alignment section 442: Second alignment section 443: Third alignment section 444: 4th alignment part 445: 5th alignment part 45: Feature extraction unit

Claims

1. A feature extraction method for extracting feature values ​​from a plurality of unit data, wherein the plurality of unit data are measurement data measured by a sensor associated with a predetermined physical quantity and serve as a unit of processing; a change point detection step of detecting a change point of the measurement data in the plurality of unit data; an alignment step of aligning positions of the physical quantities of the plurality of unit data with the detected change point as an origin; a feature extraction step of extracting the feature from the plurality of unit data while aligning the positions of the physical quantities, The alignment step includes: a first alignment step of aligning positions of the physical quantities of the plurality of unit data using the change point as an origin; a second alignment step of calculating a difference area between the plurality of unit data after aligning the physical quantities; a third alignment step of adding up the difference areas to obtain a sum of the difference areas; a fourth alignment step of repeating the first to third alignment steps while changing the origin; a fifth alignment step of determining an origin at which the difference area sum is minimum after repeating the first to third alignment steps in the fourth alignment step; Including, In the feature extraction step, the feature is extracted from the plurality of unit data in a state where the physical quantity is aligned with an origin at which the difference area sum is minimum. Feature extraction methods.

2. In the change point detection step, timings at which second-order differential values ​​of the plurality of unit data items exceed a predetermined value or become discontinuous are detected as the change points. The feature extraction method according to claim 1 .

3. In the second alignment step, average data of the aligned plurality of unit data is calculated, and then the difference area between the average data and the plurality of unit data is calculated. The feature extraction method according to claim 1 or 2.

4. In the feature extraction step, the plurality of unit data are divided into a plurality of sections by an arbitrary physical quantity width, and a predetermined statistical quantity is calculated as the feature for each of the plurality of sections. The feature extraction method according to any one of claims 1 to 3.

5. the plurality of unit data are continuously stored as a database, a unit data extraction step of extracting the plurality of unit data from the database; further comprising the change point detection step detects the change points in the plurality of unit data extracted in the unit data extraction step; The feature extraction method according to any one of claims 1 to 4.

6. a noise removal step of removing noise from the extracted plurality of unit data. further comprising In the change point detection step, a change point of the measurement data in the plurality of unit data from which the noise has been removed is detected. The feature extraction method according to any one of claims 1 to 5.

7. A feature extraction method for extracting feature values ​​from a plurality of unit data, wherein the plurality of unit data are measurement data measured by a sensor associated with a predetermined physical quantity and serve as a unit of processing; an alignment step of aligning positions of the physical quantities of the plurality of unit data; a feature extraction step of extracting the feature from the plurality of unit data while aligning the positions of the physical quantities; Including, The alignment step includes: a first alignment step of aligning positions of the physical quantities of the plurality of unit data with a position of any one of the physical quantities as an origin; a second alignment step of calculating a difference area between the plurality of unit data after aligning the physical quantities; a third alignment step of adding up the difference areas to obtain a sum of the difference areas; a fourth alignment step of repeating the first to third alignment steps while changing the origin; a fifth alignment step of determining an origin at which the difference area sum is minimum after repeating the first to third alignment steps in the fourth alignment step; Including, In the feature extraction step, the feature is extracted from the plurality of unit data in a state where the physical quantity is aligned with an origin at which the difference area sum is minimum. Feature extraction methods.

8. A feature extraction program for extracting feature values ​​from a plurality of unit data, the plurality of unit data being measurement data measured by a sensor associated with a predetermined physical quantity and serving as a unit of processing; a change point detection function for detecting a change point of the measurement data in the plurality of unit data; a positioning function for aligning positions of the physical quantities of the plurality of unit data with the detected change point as an origin; a feature extraction function for extracting the feature from the plurality of unit data while aligning the positions of the physical quantities; This is a feature extraction program that enables a computer to achieve the above. The alignment function includes: a first alignment function that aligns positions of the physical quantities of the plurality of unit data using the change point as an origin; a second alignment function for calculating a difference area between the plurality of unit data items after aligning the physical quantities; a third alignment function for calculating a sum of the difference areas by adding up the difference areas; a fourth alignment function that repeats the first to third alignment functions while changing the origin; a fifth alignment function that determines an origin at which the difference area sum is minimum after the fourth alignment function repeats the first to third alignment functions; and Including, the feature extraction function extracts the feature from the plurality of unit data in a state where the physical quantity is aligned at an origin where the difference area sum is minimum; Feature extraction program.

9. the change point detection function detects, as the change point, a timing at which second-order differential values ​​of the plurality of unit data exceed a predetermined value or become discontinuous; The feature extraction program according to claim 8.

10. the second alignment function calculates average data of the aligned plurality of unit data, and then calculates the difference area between the average data and the plurality of unit data.

10. The feature extraction program according to claim 8 or 9.

11. the feature extraction function divides the plurality of unit data into a plurality of sections by an arbitrary physical quantity width, and calculates a predetermined statistical quantity as the feature for each of the plurality of sections; 11. The feature extraction program according to claim 8.

12. the plurality of unit data are continuously stored as a database, A unit data extraction function for extracting the plurality of unit data from the database. We will further realize this on computers, the change point detection function detects the change points in the plurality of unit data extracted by the unit data extraction function; 12. The feature extraction program according to claim 8.

13. a noise removal function for removing noise from the extracted plurality of unit data; We will further realize this on computers, the change point detection function detects change points of the measurement data in the plurality of unit data from which the noise has been removed; 13. The feature extraction program according to claim 8.

14. A feature extraction program for extracting feature values ​​from a plurality of unit data, the plurality of unit data being measurement data measured by a sensor associated with a predetermined physical quantity and serving as a unit of processing; a position alignment function for aligning positions of the physical quantities of the plurality of unit data; a feature extraction function for extracting the feature from the plurality of unit data while aligning the positions of the physical quantities; This is a feature extraction program that enables a computer to achieve the above. The alignment function includes: a first alignment function that aligns positions of the physical quantities of the plurality of unit data with a position of any one of the physical quantities as an origin; a second alignment function for calculating a difference area between the plurality of unit data items after aligning the physical quantities; a third alignment function for calculating a sum of the difference areas by adding up the difference areas; a fourth alignment function that repeats the first to third alignment functions while changing the origin; a fifth alignment function that determines an origin at which the difference area sum is minimum after the fourth alignment function repeats the first to third alignment functions; and Including, the feature extraction function extracts the feature from the plurality of unit data in a state where the physical quantity is aligned at an origin where the difference area sum is minimum; Feature extraction program.

15. A feature extraction device for extracting feature values ​​from a plurality of unit data, wherein the plurality of unit data are measurement data measured by a sensor associated with a predetermined physical quantity and serve as a unit of processing; a change point detection unit that detects change points of the measurement data in the plurality of unit data; a positioning unit that aligns positions of the physical quantities of the plurality of unit data with the detected change point as an origin; a feature extraction unit that extracts the feature from the plurality of unit data while aligning the positions of the physical quantities; Equipped with The alignment unit is a first alignment unit that aligns positions of the physical quantities of the plurality of unit data using the change point as an origin; a second alignment unit that calculates a difference area between the plurality of unit data items after aligning the physical quantities; a third alignment unit that calculates a sum of the difference areas by adding up the difference areas; a fourth alignment unit that repeats the processes of the first to third alignment units while changing the origin; a fifth alignment unit that determines an origin at which the difference area sum is minimum after the fourth alignment unit has repeated the processes of the first to third alignment units; and Including, the feature extraction unit extracts the feature from the plurality of unit data in a state where the physical quantities are aligned at an origin where the difference area sum is minimum; Feature extraction device.

16. the change point detection unit detects, as the change point, a timing at which second-order differential values ​​of the plurality of unit data exceed a predetermined value or become discontinuous. The feature extraction device according to claim 15.

17. the second alignment unit calculates average data of the aligned plurality of unit data, and then calculates the difference area between the average data and the plurality of unit data.

17. The feature extraction device according to claim 15 or 16.

18. the feature extraction unit divides the plurality of unit data into a plurality of sections by an arbitrary physical quantity width, and calculates a predetermined statistical quantity as the feature for each of the plurality of sections; 18. The feature extraction device according to claim 15.

19. the plurality of unit data are continuously stored as a database, a unit data extraction unit that extracts the plurality of unit data from the database; Furthermore, the change point detection unit detects the change points in the plurality of unit data extracted by the unit data extraction unit.

19. The feature extraction device according to claim 15.

20. a noise removal unit that removes noise from the extracted plurality of unit data; Furthermore, the change point detection unit detects change points of the measurement data in the plurality of unit data from which the noise has been removed.

20. The feature extraction device according to claim 15.

21. A feature extraction device for extracting feature values ​​from a plurality of unit data, wherein the plurality of unit data are measurement data measured by a sensor associated with a predetermined physical quantity and serve as a unit of processing; a positioning unit that aligns positions of the physical quantities of the plurality of unit data; a feature extraction unit that extracts the feature from the plurality of unit data while aligning the positions of the physical quantities; Equipped with The alignment unit is a first alignment unit that aligns positions of the physical quantities of the plurality of unit data with a position of any one of the physical quantities as an origin; a second alignment unit that calculates a difference area between the plurality of unit data items after aligning the physical quantities; a third alignment unit that calculates a sum of the difference areas by adding up the difference areas; a fourth alignment unit that repeats the processes of the first to third alignment units while changing the origin; a fifth alignment unit that determines an origin at which the difference area sum is minimum after the fourth alignment unit has repeated the processes of the first to third alignment units; and Equipped with the feature extraction unit extracts the feature from the plurality of unit data in a state where the physical quantities are aligned at an origin where the difference area sum is minimum; Feature extraction device.

Images