WO2021167082A1

WO2021167082A1 - Information processing device, information processing method, and program

Info

Publication number: WO2021167082A1
Application number: PCT/JP2021/006429
Authority: WO
Inventors: 悠希中澤
Original assignee: 三菱パワー株式会社
Priority date: 2020-02-21
Filing date: 2021-02-19
Publication date: 2021-08-26
Also published as: JP2021131818A

Abstract

This information processing device is provided with: a storage unit which, having as reference detection values at least one of multiple detection values in one set, stores correction information for correcting detection times of multiple other detection values, and which stores unit space information indicating unit spaces created on the basis of multiple sets of time series of multiple detection values to which the detection times corrected on the basis of the correction information respectively correspond; a detection value acquisition unit which acquires multiple sets of time series of multiple detection values; a detection time correction unit which, on the basis of the correction information, corrects the detection times of the acquired multiple detection values; and a Mahalanobis distance calculation unit which calculates the Mahalanobis distance of the set of multiple detection values to which the corrected detection times respectively correspond.

Description

Information processing equipment, information processing methods and programs

This disclosure relates to information processing devices, information processing methods and programs. The present application claims priority based on Japanese Patent Application No. 2020-022866 filed in Japan on February 21, 2020, the contents of which are incorporated herein by reference.

Patent Document 1 discloses a plant condition monitoring device using the MT method (Maharanobis Taguchi System). Here, the MT method is a method of determining whether or not the unit space is normal based on the Mahalanobis distance, which represents the distance from the reference data group (normal data group) constituting the unit space. In the plant condition monitoring device described in Patent Document 1, the unit space used as a reference when calculating the Mahalanobis distance is, for example, a time when it goes back n days from the time when it goes back m days from the day when the state is monitored. It is created based on multiple sets of multidimensional state quantities in the time series acquired during the period up to.

In the control of a plant or the like, a certain time delay may occur between the change in the manipulated variable and the change in the controlled variable (for example, Patent Document 2).

Japanese Unexamined Patent Publication No. 2012-67757 Japanese Unexamined Patent Publication No. 2019-28824

Mahalanobis distance is a value that takes into account the correlation between multidimensional state quantities. As the absolute value of the correlation coefficient between each state quantity constituting the unit space increases, the unit space decreases and the Mahalanobis distance increases. If the unit space can be reduced and the Mahalanobis distance can be increased, it becomes easier to improve the accuracy and sensitivity of detecting an abnormal state and reduce the amount of data when creating a unit space. However, as described in Patent Document 2, for example, when a time delay occurs between the state quantities, the absolute value of the correlation coefficient between the state quantity without delay and the state quantity with delay decreases. In this case, it becomes difficult to reduce the unit space. Therefore, compared to the case where the unit space can be easily reduced, the Mahalanobis distance is used to determine whether the unit space is normal or not, and the unit space is created. There is a problem that it becomes difficult to properly perform the processing related to the distance.

The present disclosure has been made in view of the above circumstances, and an object of the present disclosure is to provide an information processing device, an information processing method, and a program capable of more appropriately performing processing related to Mahalanobis distance.

In order to solve the above problems, the information processing apparatus according to the present disclosure uses one of a plurality of detection values in a set as a reference detection value, and corrects information for correcting each detection time of the other plurality of detection values. And a storage unit that stores unit space information representing a unit space created based on a plurality of sets of time series of the plurality of detection values in which each detection time corrected based on the correction information corresponds to each other. , A detection value acquisition unit that acquires a plurality of sets of the plurality of detection values in a time series, a detection time correction unit that corrects each detection time of the acquired plurality of detection values based on the correction information, and the unit. A Mahalanobis distance calculation unit for calculating a Mahalanobis distance of a plurality of sets of the detected values whose corrected detection times correspond to each other based on spatial information is provided.

Further, the information processing apparatus according to the present disclosure is a storage unit that stores a plurality of sets of a plurality of detection values acquired in advance in chronological order, and a plurality of sets based on the plurality of the plurality of sets of the detection values. A correction information creation unit that creates correction information for correcting each detection time of each of the other plurality of detection values using one of the detection values as a reference detection value and stores the correction information in the storage unit, and the plurality of sets. Based on the plurality of sets of the time series of the plurality of the detected values in which the respective detection times of the plurality of the detected values are corrected based on the correction information, and the respective detection times correspond to each other based on the plurality of the detected values of the unit. It includes a unit space information creating unit that creates the unit space information representing a space and stores it in the storage unit.

The information processing method according to the present disclosure stores correction information for correcting each detection time of the other plurality of detection values using one of the plurality of detection values in one set as a reference detection value, and the correction information. Using a storage unit that stores unit space information representing a unit space created based on a plurality of sets of time series of the plurality of detection values whose detection times are corrected based on the above, the plurality of detection values A step of acquiring a plurality of sets of the time series of the above, a step of correcting each detection time of the acquired plurality of detection values based on the correction information, and each of the corrected detection times based on the unit space information. Has a step of calculating the Mahalanobis distance of a plurality of sets of the detected values corresponding to each other.

In addition, the information processing method according to the present disclosure uses a storage unit that stores a plurality of sets of a plurality of detected values acquired in advance in chronological order, and one set is based on the plurality of the detected values of the plurality of sets. A step of creating correction information for correcting each detection time of each of the other plurality of detection values using one of the plurality of detection values as a reference detection value and storing the correction information in the storage unit, and the plurality of sets. The unit space is based on a plurality of sets of time series of the plurality of detection values in which the detection times of the plurality of detection values are corrected based on the correction information based on the plurality of detection values. It has a step of creating the unit space information representing the above and storing the unit space information in the storage unit.

The program according to the present disclosure stores correction information for correcting each detection time of the other plurality of detection values using one of the plurality of detection values in one set as a reference detection value, and is based on the correction information. When the plurality of detected values are detected using a storage unit that stores unit space information representing the unit space created based on a plurality of sets of time series of the plurality of detected values whose corrected detection times correspond to each other. A step of acquiring a plurality of sets of a series, a step of correcting each detection time of the acquired plurality of detection values based on the correction information, and a step of correcting each of the corrected detection times based on the unit space information are mutually exclusive. Have the computer perform a step of calculating the Mahalanobis distance of the corresponding set of the plurality of detection values.

Further, the program according to the present disclosure uses a storage unit that stores a plurality of sets of a plurality of detection values acquired in advance in chronological order, and based on the plurality of the plurality of detection values of the plurality of sets, a plurality of sets. A step of creating correction information for correcting each detection time of a plurality of other detection values using one of the detection values of the above as a reference detection value and storing the correction information in the storage unit, and a plurality of the plurality of sets. Based on the detection value, each detection time of the plurality of detection values is corrected based on the correction information, and each detection time represents the unit space based on a plurality of sets of time series of the plurality of detection values corresponding to each other. The computer is made to execute the step of creating the unit space information and storing it in the storage unit.

According to the information processing apparatus, information processing method and program of the present disclosure, the processing related to the Mahalanobis distance can be performed more appropriately.

It is a block diagram which shows the structural example of the information processing apparatus which concerns on 1st Embodiment of this disclosure. It is a flowchart which shows the operation example of the information processing apparatus 1 shown in FIG. It is a schematic diagram which shows an example of the monitoring target of the monitoring device 5 shown in FIG. It is a schematic diagram for demonstrating the operation example of the information processing apparatus 1 shown in FIG. It is a schematic diagram for demonstrating the operation example of the information processing apparatus 1 shown in FIG. It is a schematic diagram for demonstrating the operation example of the information processing apparatus 1 shown in FIG. It is a schematic diagram for demonstrating the operation example of the information processing apparatus 1 shown in FIG. It is a schematic diagram for demonstrating the operation example of the information processing apparatus 1 shown in FIG. It is a block diagram which shows the structural example of the information processing apparatus which concerns on 2nd Embodiment of this disclosure. It is a block diagram which shows the structural example of the information processing apparatus which concerns on 3rd Embodiment of this disclosure. It is a schematic block diagram which shows the structure of the computer which concerns on at least one Embodiment.

<First Embodiment>
(Configuration of information processing device)
Hereinafter, the information processing apparatus according to the embodiment of the present disclosure will be described with reference to FIGS. 1 to 8. FIG. 1 is a block diagram showing a configuration example of an information processing device according to the first embodiment of the present disclosure. FIG. 2 is a flowchart showing an operation example of the information processing apparatus 1 shown in FIG. FIG. 3 is a schematic diagram showing an example of a monitoring target of the monitoring device 5 shown in FIG. 4 to 8 are schematic views for explaining an operation example of the information processing apparatus 1 shown in FIG. In each figure, the same reference numerals are used for the same or corresponding configurations, and the description thereof will be omitted as appropriate.

The information processing device 1 shown in FIG. 1 can be configured by using a computer such as a server or a personal computer or a computer and its peripheral devices, and includes hardware such as a computer and software such as a program executed by the computer. A processing unit 2 and a storage unit 3 are provided as a functional configuration composed of combinations. Further, the processing unit 2 includes a detection value acquisition unit 21, a detection time correction unit 22, a Mahalanobis distance calculation unit 23, a determination unit 24, a correction information creation unit 25, and a unit space information creation unit 26. Further, the correction information creation unit 25 has a correlation coefficient calculation unit 251 and a delay time definition unit 252. Further, the storage unit 3 stores the correction information 31, the unit space information 32, the detected value time series information (for creating the unit space) 33, and the detected value time series information (for evaluation) 34.

The processing unit 2 shown in FIG. 1 performs processing related to the Mahalanobis distance, such as a processing for determining whether or not it is normal using the Mahalanobis distance and a processing for creating a unit space.

The detection value acquisition unit 21 receives time-series data (time-series set) representing a plurality of types of detection values from the monitoring device 5 at a predetermined cycle, and stores the detection value time-series information (unit space creation) in the storage unit 3. (For) 33, or as detected value time series information (for evaluation) 34. Here, the monitoring device 5 will be described with reference to FIG. The time-series set means a plurality of sets of time-series data having different detection times, with a set of multiple types of detection values having the same detection time as one set.

As shown in FIG. 3, for example, in a plant such as a power plant 40, the monitoring device 5 uses sensors or the like provided in each part of the plant to determine a plurality of types of predetermined temperature, pressure, volume, output, rotation speed, and the like. The time series of physical quantities is acquired and transmitted as a detection value to the information processing apparatus 1 at a predetermined cycle. In the following, a plurality of detected values (k types of detected values) are represented by x1, x2, ..., Xk. Further, the n detection values x1 in the time series are x11, x21, ..., Xn1, the n detection values x2 in the time series are x12, x22, ..., Xn2, and the n detection values xk in the time series are x1k. It is expressed as x2k, ..., Xnk. Here, the detection values x11, x12, ..., X1k are a set of detection values having the same detection time (time t1), and the detection values x21, x22, ..., X2k have the same detection time (time t2). The detection values xn1, xn2, ..., Xnk are a set of detection values having the same detection time (time tun).

The power plant 40, the generator 41, the generator 42, the gas turbine 43, the steam turbine 44, the boiler 45, and the condenser 46 shown in FIG. 3 are provided. The generator 41 is driven by a gas turbine 43 to generate electric power (GTMW). The gas turbine 43 includes a compressor, a combustor, and a turbine, and the air sucked into the compressor becomes high-temperature and high-pressure air by the compressor and is guided to the combustor. In the combustor, fuel gas is supplied to high-temperature and high-pressure air and burned. The fuel gas burned in the combustor becomes high-temperature and high-pressure combustion gas and is supplied to the turbine to drive the turbine. Further, the combustion gas is supplied as exhaust gas to the boiler 45 via the turbine. The boiler 45 heats the hot water supplied from the condenser 46 and converts it into steam to rotate the steam turbine 44. The steam turbine 44 drives a generator 42, and the driven generator 42 generates electric power (STMW). Further, the steam obtained by rotating the steam turbine 44 is cooled by seawater in the condenser 46 and converted into hot water.

FIG. 4 schematically shows the time change of the flow rate of the fuel gas (kg / s), the output GTMW (MW) of the generator 41, and the output STMW (MW) of the generator 42. The flow rate of the fuel gas, the output GTMW of the generator 41, and the output STMW of the generator 42 correspond to, for example, any of the plurality of detected values x1, x2, ..., Xk described above. As shown in FIG. 4, since the gas turbine 43 responds quickly, the output GTMW of the generator 41 changes following the behavior of the combustion gas with almost no delay. On the other hand, since the steam turbine 44 is rotated by the steam generated by heating the boiler 45 with the exhaust gas of the gas turbine 43, the output STMW of the generator 42 is relative to the output GTMW (and fuel gas flow rate) of the generator 41. Therefore, the response is delayed and a time delay (ΔT) occurs.

On the other hand, the detection time correction unit 22 corrects each detection time of the acquired plurality of detected values based on the correction information 31. The correction information 31 is information for correcting each detection time of the other plurality of detection values by using one of the plurality of detection values in one set as the reference detection value. The correction information 31 includes, for example, information representing each delay time of each detection time of a plurality of other detection values with reference to the detection time of the reference detection value. For example, in the example shown in FIG. 4, when the fuel gas flow rate is used as the reference detection value, the delay time of the fuel gas flow rate and the delay time of the output GTMW of the electric machine 41 are set to zero in the correction information 31, and the generator 42 The delay time of the output STMW is set to ΔT. The reference detection value is a detection value that serves as a reference when defining the correction information 31, and is preferably a detection value that affects many behaviors of other detection values (detection values other than the reference detection value). For example, in the example shown in FIG. 4, in addition to the fuel gas flow rate, the output GTMW of the electric machine 41 and the like can be used. On the other hand, for example, it is desirable to avoid detected values such as atmospheric temperature and seawater temperature that have a small direct effect (small correlation) with the power generation output.

The time delay value is, for example, the absolute value of the correlation coefficient between the reference detection value and the other detection value when the time difference between the detection time of the reference detection value and the detection time of the other detection value is changed. The time difference that becomes the maximum value can be defined as the delay time. In FIG. 5, the horizontal axis represents the time difference between the detection time of the reference detection value and the detection time of other detection values, and the vertical axis represents the correlation coefficient between the reference detection value and the other detection values. FIG. 5 shows the correspondence between the time series of the reference detection values for a predetermined time from a certain detection time and the correlation coefficient with the time series of other detection values for a predetermined time from the detection time shifted by the time difference from the detection time. Show the relationship. When the time difference is zero, it corresponds to the correlation coefficient between the time series of the reference detection values for a predetermined time from the same detection time and the time series of other detection values. In this case, the absolute value of the correlation coefficient between the time series of the reference detection values for a predetermined time from a certain detection time and the time series of other detection values for a predetermined time from the time obtained by adding the delay time to the detection time is the maximum. It has become. Since the time when the correlation coefficient is maximized has a strong correlation, it can be defined as the delay time for correcting the detection time. The time difference can be shifted in units of 1 minute, for example.

The correlation coefficient is an index that measures the strength of the linear relationship between two random variables. Given the random variables X and Y with positive variance, the covariance is σXY and the standard deviation is the standard deviation. As σX and σY, the correlation coefficient ρ of the random variable X and the random variable Y is obtained by ρ = σXY / (σX · σY).

In the example shown in FIG. 4, the detection time correction unit 22 is delayed by, for example, the detection time of the detection value representing the flow rate of the fuel gas and the time delay (ΔT) of the detection value representing the output STMW of the generator 42. The detection time (time stamp) of the detection value representing the output STMW of the generator 42 is modified so that the detection time can be associated with each other.

As shown in FIGS. 6 and 7, when the curve representing the correlation coefficient between the reference detection value and the other detection value does not exceed the predetermined threshold value (1) as in the case of curve Re, for example. For the detected value, the time at which the correlation coefficient becomes maximum (Re_max) is not defined as the delay time, and for example, the delay time may be defined as zero. Note that FIG. 6 shows an example of curves Ra, Rb, Rc, Rd and Re of the correlation coefficient between the reference detection value and other detection values in the same manner as in FIG. 5, and FIG. 7 shows the maximum correlation coefficient. An example of the correspondence between the values Ra_max, Rb_max, Rc_max, Rd_max and Re_max and the delay time is shown.

Further, the Mahalanobis distance calculation unit 23 is based on the unit space information 32, and the corrected detection times of the detected values stored as the detected value time series information (for evaluation) 34 are a plurality of detected values corresponding to each other. Calculate the Mahalanobis distance of the set. Here, the unit space information 32 is information representing a unit space created based on a plurality of sets of time series of a plurality of detection values in which each detection time corrected based on the correction information 31 corresponds to each other.

Here, how to calculate the Mahalanobis distance will be explained with reference to mathematical formulas.

Assuming that the reference data group (normal data group) constituting the unit space is represented by a plurality of detected values xij (here, i represents n time series detection times from 1 to n, and j represents 1 to k). The average value Mj of each detected value xij is expressed by the following equation (representing k types of detected values up to). Note that i corresponds to a value corrected for each detected value by the detection time correction unit 22.

The standard deviation σj of each detected value xij is expressed by the following equation.

Further, the value Xij obtained by normalizing each detected value xij is expressed by the following equation.

Then, the correlation coefficient of all combinations for extracting 2 from k detected values is calculated by the following formula.

Further, the correlation coefficient matrix R is represented by the following equation, which is a matrix of k × k in which the diagonal element is 1 and the other elements in the p row and q column are rpq in the above equation.

Further, the inverse matrix R-1 of the correlation coefficient matrix R is expressed by the following equation.

Then, the squared value of the Mahalanobis distance D of a set of detected values (normalized values) Xm1, Xm2, ..., Xmk at the detection time tm is expressed by the following equation. In addition, m corresponds to the value corrected for each detection value by the detection time correction unit 22.

In the above formula, the unit space information 32 can be information representing the inverse matrix R-1 of the correlation coefficient matrix R.

Further, the determination unit 24 compares the Mahalanobis distance D with a predetermined threshold value, and a set of a plurality of detection values (a set of detection values (normalized values) Xm1, Xm2, ..., Xmk) is in a normal state. Determine if it exists.

Further, the correction information creation unit 25 creates correction information 31 based on a plurality of sets of time series of a plurality of detected values acquired in advance stored as detection value time series information (for unit space creation) 33. It is stored in the storage unit 3. At that time, the correlation coefficient calculation unit 251 changes each correlation coefficient with the reference detection value when each time difference between the detection time of the reference detection value and each detection time of the other plurality of detection values is changed. calculate. Further, the delay time definition unit 252 defines each time difference in which the absolute value of each correlation coefficient becomes the maximum value as each delay time.

Further, the unit space information creation unit 26 corrects each detection time of the plurality of detected values based on a plurality of sets of the plurality of detected values stored as the detected value time series information (for unit space creation) 33. Based on a plurality of sets of time series of a plurality of detection values whose detection times are corrected based on the above, unit space information 32 representing the unit space is created and stored in the storage unit 3.

(Example of operation of information processing device)
Next, an operation example of the information processing apparatus 1 shown in FIG. 1 will be described with reference to FIG. In the process shown in FIG. 2, first, the unit space information creation unit 26 acquires the past detection value for unit space creation from the detected value time series information (for unit space creation) 33 (step S11). Next, the correlation coefficient calculation unit 251 calculates the correlation coefficient of another detected value with respect to the reference detected value (step S12). Next, the delay time definition unit 252 defines the time of the maximum correlation coefficient as the delay time (step S13). Next, the unit space information creation unit 26 corrects the time of other detected values by the delay time and creates the unit space (step S14). Next, the detection value acquisition unit 21 (or the detection time correction unit 22) acquires the evaluation detection value from the monitoring device 5 (or the detection value time series information (for evaluation) 34) (step S15). Next, the detection time correction unit 22 corrects the time of another detected value for evaluation by the delay time (step S16). Next, the Mahalanobis distance calculation unit 23 calculates the Mahalanobis distance (MD) of the corrected evaluation detection value from the corrected unit space (step S17). Next, the determination unit 24 determines whether or not the Mahalanobis distance (MD) is equal to or less than a predetermined threshold value (step S18). When the Mahalanobis distance (MD) is equal to or less than a predetermined threshold value (when “YES” in step S18), the determination unit 24 determines that it is normal (step S19) and outputs the determination result (step S21). On the other hand, when the Mahalanobis distance (MD) is not equal to or less than a predetermined threshold value (when "NO" in step S18), the determination unit 24 determines that it is abnormal (step S20) and outputs the determination result (step S21).

(Examples of information processing equipment and effects)
As described above, according to the present embodiment, processing related to the Mahalanobis distance, such as creation of a unit space and calculation of the Mahalanobis distance, can be performed more appropriately.

In the embodiment of the present embodiment, for example, (1) the correlation coefficient of another detected value with respect to the reference detected value is calculated. At that time, a bundle of past detected values for creating a unit space is acquired, and the correlation coefficient of other detected values with respect to a preset reference detected value (for example, fuel flow rate) is set, for example, the time of the detected value. It is corrected and calculated every minute. Next, (2) the time of the maximum correlation coefficient is defined as the delay time. For example, the time when the correlation coefficient of the detected value obtained by correcting the time every minute is maximized is defined as the delay time of the detected value. Next, (3) the time of other detected values is corrected by the delay time to create a unit space. For example, the time of the time stamp of each past detection value (other than the reference detection value) for creating the unit space is corrected by the delay time of each detection value. Then, a unit space is created by using the reference detection value and the time-corrected detection value. Next, (4) the time of other detected values for evaluation is corrected by the delay time. For example, the time of the time stamp of each detection value (other than the reference detection value) for evaluation is corrected by the delay time of each detection value. Next, (5) the Mahalanobis distance of the corrected evaluation detection value is calculated from the corrected unit space. For example, the Mahalanobis distance of the time-corrected evaluation detection value is calculated with respect to the time-corrected unit space.

According to the present embodiment, a) due to the consideration of the dynamic delay, the correlation between the detected values becomes accurate (the variation becomes small) and the unit space becomes small, so that the evaluation detected value calculated from the unit space becomes The accuracy of Mahalanobis distance (MD) is improved. For example, as shown in FIG. 8, the region occupied by the iso-Mahalanobis distance curve 51 when the correlation is high is smaller than the region occupied by the iso-Mahalanobis distance curve 52 when the correlation is low. In FIG. 8, when the two detected values are xa and xb, the regions where the Mahalanobis distance D is equal are represented by

curves

51 and 52. For example, the Mahalanobis distance D (xa1, xb1) when the detection values for evaluation are xa1 and xb1 is outside the curve for the curve 51 and inside the curve for the curve 52, so that the judgment for the same distance can be made. It can be seen that the accuracy differs depending on the magnitude of the correlation.

Further, according to the present embodiment, b) the correlation between each detected value becomes accurate, so that the collection period of the bundle of detected values created in the unit space can be shortened. Further, c) Since the collection period of the bundle of detected values created in the unit space can be shortened, the surplus data storage capacity can be used to cover and monitor a wide range of operation patterns. For example, the output range is 100 MW to 200 MW ⇒ 0 MW to 200 MW and the like.

<Second Embodiment>
The information processing device 1a of the second embodiment is shown in FIG. Compared with the information processing device 1 shown in FIG. 1, the information processing device 1a shown in FIG. 9 has a processing unit 2a in which the correction information creation unit 25 and the unit space information creation unit 26 are omitted, and the detected value time series information (unit). It is composed of a storage unit 3a in which the space creation) 33 is omitted. According to the information processing apparatus 1a of the second embodiment, the Mahalanobis distance can be calculated based on the detected value corrected for the time delay by using the correction information 31 and the unit space information 32 prepared in advance.

<Third Embodiment>
The information processing device 1b of the third embodiment is shown in FIG. Compared with the information processing device 1 shown in FIG. 1, the information processing device 1b shown in FIG. 10 includes a detection time correction unit 22, a Mahalanobis distance calculation unit 23, a processing unit 2b in which the determination unit 24 is omitted, and a detection value. It is composed of a storage unit 3b in which the time-series information (for evaluation) 34 is omitted. According to the information processing apparatus 1b of the third embodiment, the correction information 31 can be created, and the unit space information 32 can be created based on the detected value corrected for the time delay.

(Other embodiments)
Although the embodiments of the present disclosure have been described in detail with reference to the drawings, the specific configuration is not limited to the embodiments, and includes design changes and the like within a range that does not deviate from the gist of the present disclosure. ..

<Computer configuration>
FIG. 11 is a schematic block diagram showing the configuration of a computer according to at least one embodiment.
The computer 90 includes a processor 91, a main memory 92, a storage 93, and an interface 94.
The

information processing devices

1, 1a and 1b described above are mounted on the computer 90. The operation of each processing unit described above is stored in the storage 93 in the form of a program. The processor 91 reads a program from the storage 93, expands it into the main memory 92, and executes the above processing according to the program. Further, the processor 91 secures a storage area corresponding to each of the above-mentioned storage units in the main memory 92 according to the program.

The program may be for realizing a part of the functions exerted on the computer 90. For example, the program may exert its function in combination with another program already stored in the storage or in combination with another program mounted on another device. In another embodiment, the computer may include a custom LSI (Large Scale Integrated Circuit) such as a PLD (Programmable Logic Device) in addition to or instead of the above configuration. Examples of PLDs include PAL (Programmable Array Logic), GAL (Generic Array Logic), CPLD (Complex Programmable Logic Device), and FPGA (Field Programmable Gate Array). In this case, some or all of the functions realized by the processor may be realized by the integrated circuit.

Examples of the storage 93 include HDD (Hard Disk Drive), SSD (Solid State Drive), magnetic disk, optical magnetic disk, CD-ROM (Compact Disc Read Only Memory), DVD-ROM (Digital Versatile Disc Read Only Memory). , Semiconductor memory and the like. The storage 93 may be internal media directly connected to the bus of the computer 90, or external media connected to the computer 90 via the interface 94 or a communication line. When this program is distributed to the computer 90 via a communication line, the distributed computer 90 may expand the program in the main memory 92 and execute the above processing. In at least one embodiment, the storage 93 is a non-temporary tangible storage medium.

<Additional notes>
The

information processing devices

1, 1a and 1b described in each embodiment are grasped as follows, for example.

(1) The

information processing devices

1 and 1a according to the first aspect have correction information for correcting each detection time of the other plurality of detection values by using one of the plurality of detection values in one set as a reference detection value. In addition to storing 31, unit space information 32 representing a unit space created based on a plurality of sets of time series of the plurality of detection values in which each detection time corrected based on the correction information 31 corresponds to each other is stored. The storage unit 3, the detection value acquisition unit 21 that acquires a plurality of sets of the plurality of detection values in a time series, and the detection time that corrects each detection time of the acquired plurality of detection values based on the correction information 31. A correction unit 22 and a Mahalanobis distance calculation unit 23 for calculating a Mahalanobis distance of a plurality of sets of the detection values whose corrected detection times correspond to each other based on the unit space information 32 are provided.

(2) The information processing device 1 according to the second aspect is the information processing device 1 of (1), and obtains the correction information 31 based on a plurality of sets of time series of the plurality of detected values acquired in advance. The correction information creation unit 25 that is created and stored in the storage unit 3 and the plurality of detection values in which each detection time obtained by correcting each detection time of the plurality of detection values acquired in advance based on the correction information 31 correspond to each other. A unit space information creating unit 26 that creates the unit space information 32 representing the unit space and stores the unit space information 32 in the storage unit 3 based on a plurality of sets of the time series of the above is further provided.

(3) The information processing device 1 according to the third aspect is the information processing device 1 of (2), and the correction information 31 is a plurality of other detected values based on the detection time of the reference detected value. When the correction information creation unit 25 changes each time difference between the detection time of the reference detection value and each detection time of a plurality of other detection values, including information representing each delay time of each detection time of the above. The correlation coefficient calculation unit 251 that calculates the change of each correlation coefficient with the reference detection value, and the delay time that defines each time difference at which the absolute value of each correlation coefficient becomes the maximum value as each delay time. The definition unit 252 and the like are included.

(4) The information processing device 1 according to the fourth aspect is the information processing device 1 of (1) to (3), in which the Mahalanobis distance is compared with a predetermined threshold value, and a plurality of the detected values are in a normal state. A determination unit 24 for determining whether or not the information is specified is further provided.

(5) The

information processing devices

1 and 1b according to the fifth aspect include a storage unit 3 that stores a plurality of sets of a plurality of detection values acquired in advance in time series, and a plurality of the detection values of the plurality of sets. Based on the above, correction information 31 for correcting each detection time of the other plurality of detection values using one of the plurality of detection values in one set as a reference detection value is created and stored in the storage unit 3. The information creation unit 25 and the plurality of detection values in which the detection times of the plurality of detection values are corrected based on the correction information 31 based on the plurality of detection values of the plurality of sets correspond to each other. A unit space information creating unit 26 that creates the unit space information 32 representing the unit space and stores the unit space information 32 in the storage unit 3 based on a plurality of sets of time series is provided.

According to the above-mentioned information processing device, information processing method and program, the processing related to the Mahalanobis distance can be performed more appropriately.

1, 1a, 1b Information processing device 2 Processing unit 21 Detection value acquisition unit 22 Detection time correction unit 23 Maharanobis distance calculation unit 24 Judgment unit 25 Correction information creation unit 26 Unit space information creation unit 251 Correlation coefficient calculation unit 252 Delay time definition Part 3 Storage part 31 Correction information 32 Unit space information 33 Detected value Time series information (for creating unit space)
34 Detected value time series information (for evaluation)

Claims

The correction information for correcting each detection time of the other plurality of detection values using one of the plurality of detection values in one set as the reference detection value is stored, and each detection time corrected based on the correction information is stored. A storage unit that stores unit space information representing a unit space created based on a plurality of sets of time series of a plurality of the detected values corresponding to each other.
A detection value acquisition unit that acquires a plurality of sets of the plurality of detection values in a time series,
A detection time correction unit that corrects each detection time of the acquired plurality of detection values based on the correction information, and a detection time correction unit.
A Mahalanobis distance calculation unit that calculates the Mahalanobis distance of a plurality of sets of the detection values whose corrected detection times correspond to each other based on the unit space information.
Information processing device equipped with.
A correction information creation unit that creates the correction information and stores it in the storage unit based on a plurality of sets of the detection values acquired in advance in a time series.
The unit space representing the unit space based on a plurality of sets of time series of the plurality of detection values in which each detection time of the plurality of detection values acquired in advance is corrected based on the correction information. A unit space information creation unit that creates information and stores it in the storage unit,
The information processing apparatus according to claim 1.
The correction information includes information representing each delay time of each detection time of a plurality of other detection values with reference to the detection time of the reference detection value.
The correction information creation unit
A correlation coefficient calculation unit that calculates a change in each correlation coefficient with the reference detection value when each time difference between the detection time of the reference detection value and each detection time of a plurality of other detection values is changed. ,
A delay time definition unit that defines each time difference in which the absolute value of each correlation coefficient becomes the maximum value as each delay time, and a delay time definition unit.
The information processing apparatus according to claim 2.
The information processing apparatus according to any one of claims 1 to 3, further comprising a determination unit for comparing the Mahalanobis distance with a predetermined threshold value and determining whether or not the plurality of detected values are in a normal state.
A storage unit that stores multiple sets of detected values acquired in advance in chronological order,
Based on the plurality of the plurality of detection values in the plurality of sets, correction information for correcting each detection time of the other plurality of detection values is created by using one of the plurality of detection values in one set as a reference detection value. The correction information creation unit stored in the storage unit and
Based on the plurality of detection values of the plurality of sets, each detection time of the plurality of detection values is corrected based on the correction information, and each detection time is corrected to a plurality of sets of time series of the plurality of detection values corresponding to each other. Based on this, a unit space information creation unit that creates the unit space information representing the unit space and stores it in the storage unit, and
Information processing device equipped with.
The correction information for correcting each detection time of the other plurality of detection values using one of the plurality of detection values in one set as the reference detection value is stored, and each detection time corrected based on the correction information is stored. Using a storage unit that stores unit space information representing a unit space created based on a plurality of sets of time series of a plurality of the detected values corresponding to each other,
A step of acquiring a plurality of sets of the plurality of detected values in a time series, and
A step of correcting each detection time of the acquired plurality of detection values based on the correction information, and
Based on the unit space information, a step of calculating the Mahalanobis distance of a plurality of sets of the detected values whose corrected detection times correspond to each other, and
Information processing method having.
Using a storage unit that stores multiple sets of detected values acquired in advance in chronological order,
Based on the plurality of the plurality of detection values in the plurality of sets, correction information for correcting each detection time of the other plurality of detection values is created by using one of the plurality of detection values in one set as a reference detection value. And the step of storing in the storage unit
Based on the plurality of detection values of the plurality of sets, each detection time of the plurality of detection values is corrected based on the correction information, and each detection time is corrected to a plurality of sets of time series of the plurality of detection values corresponding to each other. Based on the step of creating the unit space information representing the unit space and storing it in the storage unit,
Information processing method having.
The correction information for correcting each detection time of the other plurality of detection values using one of the plurality of detection values in one set as the reference detection value is stored, and each detection time corrected based on the correction information is stored. Using a storage unit that stores unit space information representing a unit space created based on a plurality of sets of time series of a plurality of the detected values corresponding to each other,
A step of acquiring a plurality of sets of the plurality of detected values in a time series, and
A step of correcting each detection time of the acquired plurality of detection values based on the correction information, and
Based on the unit space information, a step of calculating the Mahalanobis distance of a plurality of sets of the detected values whose corrected detection times correspond to each other, and
A program that causes a computer to run.
Using a storage unit that stores multiple sets of detected values acquired in advance in chronological order,
Based on the plurality of the plurality of detection values in the plurality of sets, correction information for correcting each detection time of the other plurality of detection values is created by using one of the plurality of detection values in one set as a reference detection value. And the step of storing in the storage unit
Based on the plurality of detection values of the plurality of sets, each detection time of the plurality of detection values is corrected based on the correction information, and each detection time is corrected to a plurality of sets of time series of the plurality of detection values corresponding to each other. Based on the step of creating the unit space information representing the unit space and storing it in the storage unit,
A program that causes a computer to run.