Data processing method, data diagnosis method, and data processing program

Abstract

The present method relates to a data processing method for processing data for calculating a Mahalanobis distance. The present method calculates an objective function for linearizing a combination of items included in unit space data, and calculates coordinate transformation parameters for performing coordinate transformation of the unit space data for each item, in order to minimize the objective function. Then, the calculated coordinate transformation parameters are used to perform coordinate transformation of the unit space data for each item. The unit space data that underwent the coordinate transformation is standardized for each item.

Description

[Technical Field]

[0001] The present disclosure relates to a data processing method, a data diagnostic method, and a data processing program. This application claims priority based on Japanese Patent Application No. 2022-166013, filed with the Japan Patent Office on October 17, 2022, the contents of which are incorporated herein by reference. [Background technology]

[0002] The Mahalanobis-Taguchi (MT) method is known as a method for determining normality / abnormality by taking into account the correlation between items (variables). The MT method calculates the Mahalanobis distance, which quantitatively indicates the degree to which the signal space, which is the group to be judged, is separated from the unit space, which is the normal group, as an index. A large Mahalanobis distance means that the signal space deviates from the unit space, and can be used as a measure for determining normality / abnormality in judgment targets that are difficult to quantify.

[0003] The MT method assumes that the population handled has a normal distribution, but generally, the population includes items with different units and items with non-normal distribution. In order to handle populations including such various items, the MT method performs standardization processing on the unit space and signal space (see, for example, Patent Document 1). For example, item X' obtained by standardizing item X can be obtained by the following equation using its average value Xavg and standard deviation σ: X´=(X-Xavg) / σ (1) The standardized item X' has a mean of "zero" and a variance of "1". [Prior art documents] [Patent documents]

[0004] [Patent Document 1] Japanese Patent Application Laid-Open No. 2012-252556 Summary of the Invention [Problem to be solved by the invention]

[0005] The MT method described above is a linear analysis method that assumes that the items being handled follow a normal distribution. Therefore, while it can reliably determine normality / abnormality when highly linear variables are used, its reliability may decrease when highly nonlinear items or items with non-normal distributions are included.

[0006] At least one embodiment of the present disclosure has been made in consideration of the above-mentioned circumstances, and aims to provide a data processing method, a data diagnosis method, and a data processing program that are capable of reliable abnormality determination even when dealing with a population that includes highly nonlinear items or items with a non-normal distribution. [Means for solving the problem]

[0007] In order to solve the above problem, a data processing method according to at least one embodiment of the present disclosure includes: 1. A data processing method for processing data for calculating Mahalanobis distance, comprising: calculating an objective function for linearizing a combination of each item included in the unit space data; calculating a coordinate transformation parameter for transforming the unit space data into coordinates for each of the items so as to minimize the objective function; a step of performing coordinate transformation on the unit space data for each of the items using the coordinate transformation parameters; standardizing the coordinate-transformed unit space data for each item; Equipped with.

[0008] In order to solve the above problem, a data diagnosis method according to at least one embodiment of the present disclosure includes: a data processing method according to at least one embodiment of the present disclosure; coordinate transforming signal space data using the coordinate transformation parameters; calculating the Mahalanobis distance as a degree of deviation of the signal space data after coordinate transformation from the unit space data after coordinate transformation; diagnosing the soundness of the signal space data based on the Mahalanobis distance; Equipped with.

[0009] In order to solve the above problem, a data processing program according to at least one embodiment of the present disclosure includes: A data processing program for processing data for calculating Mahalanobis distance, Using a computer device, calculating an objective function for linearizing a combination of each item included in the unit space data; calculating a coordinate transformation parameter for transforming the unit space data into coordinates for each of the items so as to minimize the objective function; a step of performing coordinate transformation on the unit space data for each of the items using the coordinate transformation parameters; standardizing the coordinate-transformed unit space data for each item; is possible. [Effects of the Invention]

[0010] According to at least one embodiment of the present disclosure, it is possible to provide a data processing method, a data diagnosis method, and a data processing program that are capable of reliable abnormality determination even when dealing with a population that includes highly nonlinear items or items with a non-normal distribution. [Brief explanation of the drawings]

[0011] [Figure 1] 1 is an overall configuration diagram of a plant monitoring device according to an embodiment; [Figure 2] 2 is an example of signal space data stored in the signal space file of FIG. 1. [Figure 3] 2 is an example of unit space data stored in the unit space file of FIG. 1. [Figure 4] FIG. 4 shows an example of a unit space corresponding to the unit space data. [Figure 5]2 is a flowchart showing a plant monitoring method carried out by the plant monitoring device of FIG. 1. [Figure 6] 1 is a flowchart illustrating a data processing method according to an embodiment. [Figure 7] 7 is a flowchart showing a method for calculating an objective function in step S201 of FIG. 6. [Figure 8] FIG. 10 is a diagram showing a combination of evaluation items used in calculating an objective function. [Figure 9] 7 is a flowchart showing another method for calculating the objective function in step S201 of FIG. 6. [Figure 10] FIG. 5 is a diagram showing a unit space corresponding to the result of processing the unit space data shown in FIG. 4. DETAILED DESCRIPTION OF THE INVENTION

[0012] Hereinafter, several embodiments of the present invention will be described with reference to the accompanying drawings. However, the configurations described as the embodiments or shown in the drawings are merely illustrative examples and are not intended to limit the scope of the present invention.

[0013] FIG. 1 is an overall configuration diagram of a plant monitoring system 100 according to one embodiment. The plant monitoring system 100 is a device for monitoring the operating state of a gas turbine power plant 1. The gas turbine power plant 1 includes a gas turbine 2 and a generator 6 that generates electricity by being driven by the gas turbine 2. The gas turbine 2 includes a compressor 3 that generates compressed air, a combustor 4 that mixes fuel with the compressed air and burns it to generate combustion gas, and a turbine 5 that is rotationally driven by the combustion gas. The rotor of the turbine 5 is connected to the generator 6 via the compressor 3, and the generator 6 generates electricity by rotation of the rotor.

[0014] The plant monitoring device 100 acquires state quantities for each of a plurality of evaluation items of the gas turbine power plant 1 and determines whether the operating state of the gas turbine power plant 1 is normal based on these state quantities. The plant monitoring device 100 monitors the operating state of the gas turbine power plant 1 using the Mahalanobis-Taguchi method (hereinafter referred to as the "MT method" as appropriate). Evaluation items of the gas turbine power plant 1 include, for example, the gas turbine output, the cavity temperature at a plurality of locations between the turbine rotor and the stationary part, the blade path temperature at a plurality of locations in the circumferential direction at the gas outlet of the turbine, the displacement amount at a plurality of locations in the circumferential direction of the turbine rotor, and the opening degree of various valves provided in the gas turbine. The gas turbine power plant 1 is provided with various state quantity detection means such as sensors to detect these state quantities.

[0015] The plant monitoring device 100 is composed of a computer device and includes a CPU 10 that executes various types of arithmetic processing, a main memory device 20 such as a RAM that serves as a work area for the CPU 10, an auxiliary memory device 30 such as a hard disk drive device in which various types of data, programs, etc. are stored, an input / output interface 40 for detecting various state quantities of the gas turbine power plant 1 and input / output devices such as a keyboard, mouse or touch panel, and display (not shown), and a recording / reproducing device 42 that records and reproduces data on a disk-type storage medium such as a CD or DVD.

[0016] The auxiliary storage device 30 stores in advance various programs including a plant monitoring program 34, a data processing program 33, an OS program, etc., for causing the computer to function as the plant monitoring device 100. The various programs including the data processing program 33 and the plant monitoring program 34 are loaded into the auxiliary storage device 30 from a disk-type storage medium via a storage / playback device 42. These programs may be loaded into the auxiliary storage device 30 from a portable memory such as a flash memory, or from an external device via a communication device (not shown).

[0017] Furthermore, the auxiliary storage device 30 is provided with the following files during the execution of the data processing program 33 and the plant monitoring program 34. That is, the auxiliary storage device 30 is provided with a signal space file 31 in which data (signal space data) of state quantities for each of a plurality of evaluation items of the gas turbine power plant 1 is stored, and a unit space file 32 in which data (unit space data) of a unit space that serves as a reference when determining the operating state of the plant is stored.

[0018] Fig. 2 shows an example of signal space data 31a stored in signal space file 31 of Fig. 1. As shown in Fig. 2, in the process of executing data processing program 33 and plant monitoring program 34, signal space data 31a, which is a collection of state quantities X, Y, Z for each of a plurality of evaluation items of gas turbine power plant 1, is stored in signal space file 31 in chronological order for each acquisition time T1, T2, ... of the state quantities.

[0019] Fig. 3 shows an example of unit space data 32a stored in the unit space file 32 of Fig. 1. As shown in Fig. 3, the unit space file 32 stores unit space data 32a, which is a collection of state quantities X, Y, and Z for each of a plurality of evaluation items, in correspondence with the signal space data 31a during the execution of the data processing program 33 and the plant monitoring program 34.

[0020] The CPU 10 functionally comprises a signal space data acquisition unit 11 for acquiring signal space data 31 a stored in the signal space file 31, a unit space data acquisition unit 12 for acquiring unit space data 32 a stored in the unit space file 32, a data processing unit 13 for performing data processing on the signal space data 31 a and the unit space data 32 a, a Mahalanobis distance calculation unit 14 for calculating a Mahalanobis distance using data processed by the data processing unit 13, and a plant state determination unit 15 for diagnosing the soundness of the operating state of the gas turbine power plant 1 depending on whether the Mahalanobis distance calculated by the Mahalanobis distance calculation unit 14 is within a predetermined threshold value.

[0021] Of the functional components of the CPU 10 described above, the data processing unit 13 functions when the CPU 10 executes a data processing program 33 stored in the auxiliary storage device 30. The signal space data acquisition unit 11, the unit space data acquisition unit 12, the Mahalanobis distance calculation unit 14, and the plant state determination unit 15 function when the CPU 10 executes a plant monitoring program 34 stored in the auxiliary storage device 30.

[0022] Next, the monitoring operation of the plant monitoring system 100 of this embodiment will be described. As described above, the plant monitoring by this plant monitoring system 100 utilizes the MT method. The basic content of the plant monitoring method using this MT method will be described with reference to Fig. 4. Fig. 4 shows an example of a unit space S corresponding to the unit space data 32a.

[0023] Here, the output of the generator 6 and the intake air temperature of the compressor 3 of the gas turbine power plant 1 are assumed to be state quantities X and Y, respectively, which are evaluation items (in this example, for ease of understanding, a case where there are two state quantities X and Y is described, but there may be three or more state quantities by including a state quantity Z, etc.). In the MT method, the Mahalanobis distance D is calculated as an evaluation index for whether or not the signal space data 31a indicating the operating state is abnormal, using the unit space S corresponding to the unit space data 32a, which is a collection of lattices that serve as the reference for these state quantities, as a reference. The Mahalanobis distance D takes a larger value as the degree of abnormality of the monitored object increases. Therefore, in the MT method, whether or not the operating state of the plant is abnormal is determined depending on whether or not the Mahalanobis distance D is within a predetermined threshold value Dc. In FIG. 4, the solid line surrounding the unit space S indicates the position where the Mahalanobis distance D becomes the threshold value Dc.

[0024] However, as indicated by the stars in Figure 4, the unit space data 32a exhibits nonlinear behavior due to its bends. The MT method described above is a linear analysis method that assumes that the items being evaluated follow a normal distribution. Therefore, when highly linear evaluation items are used, normal / abnormal determinations can be made with high reliability. However, when highly nonlinear items or items with non-normal distributions are included, the reliability may decrease. This issue can be resolved effectively by data processing, which will be described later.

[0025] FIG. 5 is a flowchart showing a plant monitoring method performed by the plant monitoring system 100 of FIG.

[0026] First, the signal space data acquisition unit 11 acquires the signal space data 31a stored in the signal space file 31 (step S100). Also, the unit space data acquisition unit 12 acquires the unit space data 32a stored in the unit space file 32 (step S101).

[0027] Next, the data processing unit 13 performs data processing on the signal space data 31a and unit space data 32a acquired in step S100 (step S102). Generally, the MT method is a linear analysis method that assumes that the evaluation items handled follow a normal distribution. Therefore, if an evaluation item with high nonlinearity or a non-normal distribution is included, its reliability may be reduced. Details of the data processing performed in step S4 will be described later. By performing data processing, the signal space data 31a and unit space data 32a used to calculate the Mahalanobis distance D can be coordinate-transformed to improve the linearity of the evaluation items or bring them closer to a normal distribution.

[0028] Next, the Mahalanobis distance calculation unit 14 calculates the Mahalanobis distance D using the processed signal space data 31a and unit space data 32a (step S103). Then, the plant state determination unit 15 compares the Mahalanobis distance D calculated in step S103 with a threshold value Dc to determine the soundness of the operating state of the plant (step S104).

[0029] Next, specific details of the data processing method performed in step S102 of Fig. 5 will be described. Fig. 6 is a flowchart showing a data processing method according to one embodiment. This data processing method is realized as a function of the data processing unit 13 by executing a data processing program 33. In this case, the data processing unit 13 functionally includes a first standardization unit 16, an objective function calculation unit 17, a coordinate transformation parameter calculation unit 18, a coordinate transformation operation unit 21, and a second standardization unit 22.

[0030] First, the first standardization unit 16 standardizes the unit space data 32a acquired in step S100 (step S200). The state quantities X, Y, and Z, which are multiple evaluation items included in the unit space data 32a, generally have different units and average values. In the standardization performed in step S200, these state quantities X, Y, and Z are converted so that the average value is "zero" and the standard deviation is "1" in order to compare them on an equal basis. For example, the state quantity X, which is one of the evaluation items included in the unit space data 32a, is standardized by the following equation using the average value Xavg and the standard deviation σ (the same applies to the other state quantities Y and Z). X´=(X-Xavg) / σ (2)

[0031] Next, the objective function calculation unit 17 calculates an objective function fx for linearizing the combination of each evaluation item included in the unit space data 32a (step S201), and the coordinate transformation parameter calculation unit 18 calculates coordinate transformation parameters for minimizing the objective function fx calculated in step S201 (step S202).

[0032] In step S202, coordinate transformation parameters included in any coordinate transformation formula can be used. In one embodiment, when the SHASH transformation is used, the state quantity X', which is the evaluation item, is transformed by the following formula using coordinate transformation parameters δ and ε ("X"" is the state quantity X' after the coordinate transformation): TIFF0007812006000001.tif20170

[0033] In another embodiment, the Yeo-Johnson transformation can be used as the coordinate transformation. In this case, the state quantity X', which is the evaluation item, is transformed by the following equation using a coordinate transformation parameter λ ("X" is the state quantity X' after the coordinate transformation): TIFF0007812006000002.tif30170

[0034] In another aspect, the coordinate transformation may be a Botzmann transformation, a double Botzmann transformation, a broken line transformation, etc. Any one of these exemplified coordinate transformations may be used alone, or two or more may be used in combination.

[0035] Next, the coordinate transformation calculation unit 21 performs coordinate transformation on the unit space data 32a for each item using the coordinate transformation parameters calculated in step S202 (step S203). Then, the second standardization unit 22 standardizes the unit space data 32a that has been coordinate transformed in step S203 for each item (step S204). The second standardization performed in step S204 is substantially similar to the first standardization performed in step S200.

[0036] Here, several specific examples of calculating the objective function fx in step S201 will be described. Fig. 7 is a flowchart showing a method for calculating the objective function fx in step S201 in Fig. 6, and Fig. 8 is a diagram showing combinations of evaluation items used in calculating the objective function fx.

[0037] First, a correlation coefficient R for each combination of evaluation items included in the unit space data 32a acquired in step S100 is calculated (step S300). In the example of FIG. 8, the correlation coefficient R for each combination of state quantities X, Y, and Z, which are three evaluation items, is shown. Specifically, the correlation coefficient R corresponding to the combination of state quantity X and state quantity Y is X-Y and the correlation coefficient R corresponding to the combination of state quantity X and state quantity Z. X-Z and the correlation coefficient R corresponding to the combination of state quantity Y and state quantity Z. Y-Z And, it is shown.

[0038] Next, the coefficient of determination R is calculated based on each correlation coefficient R calculated in step S300. 2 is calculated (step S301), and further, the coefficient of determination R 2 The self-information is calculated based on the above (step S302). The self-information is a concept of information theory, and is a measure of how unlikely an event (phenomenon) is to occur. The self-information can also be considered as a measure of how much information the event essentially contains. In this embodiment, the self-information I is defined by the following formula (coefficient of determination R 2 When approaches 1 (linearized), the self-information I is maximized). I=-log 10 (1-R 2 ) (5) Then, the total sum of the self-information calculated in step S303 is multiplied by "-1" to obtain the objective function fx (step S303).

[0039] FIG. 9 is a flowchart showing another method for calculating the objective function fx in step S201 of FIG.

[0040] First, the objective function calculation unit 17 calculates the skewness and kurtosis for each evaluation item included in the unit space data 32a acquired in step S100 (step S400). Specifically, the skewness and kurtosis are calculated by the following equations. TIFF0007812006000003.tif22170

[0041] Next, the objective function calculation unit 17 calculates the objective function fx by adding values ​​obtained by multiplying the squared values ​​of the skewness and kurtosis calculated in step S400 by different weighting coefficients (step S401). Specifically, the objective function fx is calculated by the following equation. fx=0.8×(kurtosis) 2 +0.2×(skewness) 2 (7)

[0042] By calculating the coordinate transformation parameters so as to minimize the objective function fx calculated based on the skewness and kurtosis in this way, it is possible to obtain the coordinate transformation parameters that will bring the skewness and kurtosis of each evaluation item closer to zero, i.e., closer to a normal distribution.

[0043] As described above, according to each of the above embodiments, by processing the signal space data 31a and the unit space data 32a for calculating the Mahalanobis distance D, the signal space data 31a and the unit space data 32a including evaluation items with high nonlinearity or non-normal distributions are coordinate-transformed. This makes it possible to improve the linearity of the signal space data 31a and the unit space data 32a or bring them closer to a normal distribution. For example, FIG. 10 is a diagram showing the unit space S' corresponding to the result of data processing of the unit space data 32a shown in FIG. 4. While the nonlinear behavior shown by the bends indicated by the stars in FIG. 4 is shown to be linear in FIG. 10. As a result, by calculating the Mahalanobis distance D based on the signal space data 31a and the unit space data 32a that have been coordinate-transformed in this manner, it is possible to accurately diagnose the soundness.

[0044] In addition, within the scope of the present disclosure, the components in the above-described embodiments may be replaced with well-known components as appropriate, and the above-described embodiments may be combined as appropriate.

[0045] The contents described in each of the above embodiments can be understood, for example, as follows.

[0046] (1) A data processing method according to one aspect includes: 1. A data processing method for processing data for calculating Mahalanobis distance, comprising: calculating an objective function for linearizing a combination of each item included in the unit space data; calculating a coordinate transformation parameter for transforming the unit space data into coordinates for each of the items so as to minimize the objective function; a step of performing coordinate transformation on the unit space data for each of the items using the coordinate transformation parameters; standardizing the coordinate-transformed unit space data for each item; Equipped with.

[0047] According to the above aspect (1), an objective function for linearizing the combination of each item included in the unit space data is calculated, and a coordinate transformation parameter for linearizing the combination of each item included in the unit space data is calculated so as to minimize the objective function. By performing coordinate transformation on the unit space data using the coordinate transformation parameter calculated in this manner, the unit space data can be linearized even if the unit space data includes items with high nonlinearity or items that do not follow a non-normal distribution. By using the unit space data linearized in this manner, the MT method enables highly reliable health diagnosis based on a group including items with high nonlinearity.

[0048] (2) In another embodiment, in the above embodiment (1), The step of calculating the objective function includes: calculating a correlation coefficient for each of the combinations; calculating a coefficient of determination based on the correlation coefficient; calculating a self-information content based on the coefficient of determination; calculating the objective function by multiplying the sum of the self-information by −1; Includes:

[0049] According to the above aspect (2), by multiplying the self-information based on the coefficient of determination calculated from the correlation coefficient for each combination of items included in the unit space data by −1, it is possible to suitably obtain an objective function for linearizing the combination of items included in the unit space data.

[0050] (3) In another embodiment, in the above embodiment (1), The step of calculating the objective function includes: Calculating skewness and kurtosis for each of the items; calculating the objective function by multiplying the squared values ​​of the skewness and the squared values ​​of the kurtosis by different weighting coefficients and adding the resulting values; Includes:

[0051] According to the above aspect (3), by adding values ​​obtained by multiplying the squared values ​​of the skewness and kurtosis for each item included in the unit space data by different weighting coefficients, an objective function for linearizing the combination of each item included in the unit space data can be suitably obtained.

[0052] (4) In another embodiment, in any one of the above (1) to (3), The coordinate transformation includes at least one of a SHASH transformation, a Yeo-Johnson transformation, a Botzmann transformation, a double Botzmann transformation, or a piecewise linear transformation.

[0053] According to the above aspect (4), by calculating the coordinate transformation parameters used in these transformation methods so as to minimize the objective function, it is possible to suitably linearize the combination of each item included in the unit space data. For coordinate transformation, any one of these transformation methods may be employed, or a combination of at least two of them may be employed, or the same method may be employed multiple times.

[0054] (5) In another embodiment, in any one of the above (1) to (4), further comprising a step of standardizing the unit space data for each of the items; In the step of calculating the objective function, an objective function for linearizing a combination of each of the items included in the standardized unit space data is calculated.

[0055] According to the above aspect (5), the objective function is calculated using unit space data that has been standardized in advance, which effectively reduces the computational load associated with calculating the objective function.

[0056] (6) A data diagnostic method according to one aspect includes: A data processing method according to any one of the above (1) to (3), coordinate transforming signal space data using the coordinate transformation parameters; calculating the Mahalanobis distance as a degree of deviation of the signal space data after coordinate transformation from the unit space data after coordinate transformation; diagnosing the soundness of the signal space data based on the Mahalanobis distance; Equipped with.

[0057] According to the above aspect (6), the signal space data to be diagnosed is also coordinate-transformed using the coordinate transformation parameters calculated to linearize the unit space data. That is, the unit space data and the signal space data are each coordinate-transformed using a common coordinate transformation parameter. By calculating the Mahalanobis distance based on the unit space and signal space data thus coordinate-transformed, it is possible to highly reliably diagnose the soundness of a population that includes highly nonlinear items or items with a non-normal distribution.

[0058] (7) A data processing program according to one aspect includes: A data processing program for processing data for calculating Mahalanobis distance, Using a computer device, calculating an objective function for linearizing a combination of each item included in the unit space data; calculating a coordinate transformation parameter for transforming the unit space data into coordinates for each of the items so as to minimize the objective function; a step of performing coordinate transformation on the unit space data for each of the items using the coordinate transformation parameters; standardizing the coordinate-transformed unit space data for each item; is possible.

[0059] According to the above aspect (7), an objective function for linearizing the combination of each item included in the unit space data is calculated, and a coordinate transformation parameter for linearizing the combination of each item included in the unit space data is calculated so as to minimize the objective function. By performing coordinate transformation on the unit space data using the coordinate transformation parameter calculated in this manner, the unit space data can be linearized even if the unit space data includes items with high nonlinearity or items that do not follow a non-normal distribution. By using the unit space data linearized in this manner, the MT method enables highly reliable health diagnosis based on a group including items with high nonlinearity. [Explanation of symbols]

[0060] 1. Gas turbine power plant 2. Gas turbine 3. Compressor 4 Combustor 5 Turbine 6. Generator 10 CPU 11 Signal space data acquisition unit 12 Unit space data acquisition section 13 Data Processing Unit 14 Mahalanobis distance calculation section 15 Plant status determination section 16 1st Standardization Department 17 Objective function calculation section 18 Coordinate transformation parameter calculation unit 20 Main storage 21 Coordinate transformation calculation unit 22 2nd Standardization Department 30 Auxiliary storage 31 Signal Space Files 31a Signal Space Data 32 Unit Space File 32a Unit spatial data 33 Data Processing Program 34 Plant Monitoring Program 40 Input / Output Interface 42 Recording and playback equipment 100 Plant monitoring equipment D Mahalanobis distance Dc threshold

Claims

1. A data processing method for processing data for calculating a Mahalanobis distance as an evaluation index for determining whether or not an operating state of a plant is abnormal, using a plant monitoring device comprising a computer device for monitoring the operating state of the plant, comprising: a step of calculating, by the plant monitoring device, an objective function for linearizing a combination of each item included in unit space data, which is data of a unit space that serves as a reference when determining the operating state; calculating, by the plant monitoring device, coordinate transformation parameters for performing coordinate transformation on the unit space data for each of the items so as to minimize the objective function; a step of performing coordinate transformation of the unit space data for each of the items using the coordinate transformation parameters by the plant monitoring device; standardizing the coordinate-transformed unit space data for each item by the plant monitoring device; Equipped with The step of calculating the objective function includes: calculating a correlation coefficient for each of the combinations; calculating a coefficient of determination based on the correlation coefficient; calculating a self-information content based on the coefficient of determination; calculating the objective function by multiplying the sum of the self-information by −1; data processing methods, including

2. A data processing method for processing data for calculating a Mahalanobis distance as an evaluation index for determining whether or not an operating state of a plant is abnormal, using a plant monitoring device comprising a computer device for monitoring the operating state of the plant, comprising: a step of calculating, by the plant monitoring device, an objective function for linearizing a combination of each item included in unit space data, which is data of a unit space that serves as a reference when determining the operating state; calculating, by the plant monitoring device, coordinate transformation parameters for performing coordinate transformation on the unit space data for each of the items so as to minimize the objective function; a step of performing coordinate transformation of the unit space data for each of the items using the coordinate transformation parameters by the plant monitoring device; standardizing the coordinate-transformed unit space data for each item by the plant monitoring device; Equipped with The step of calculating the objective function includes: Calculating skewness and kurtosis for each of the items; calculating the objective function by multiplying the squared values ​​of the skewness and the squared values ​​of the kurtosis by different weighting coefficients and adding the resulting values; data processing methods, including

3. 3. The data processing method according to claim 1, wherein the coordinate transformation includes at least one of a SHASH transformation, a Yeo-Johnson transformation, a Bötzmann transformation, a double Bötzmann transformation, and a polygonal transformation.

4. further comprising a step of standardizing the unit space data for each of the items; 3. The data processing method according to claim 1, wherein the step of calculating the objective function includes calculating an objective function for linearizing a combination of each of the items included in the standardized unit space data.

5. A data processing method according to claim 1 or 2; coordinate transforming signal space data using the coordinate transformation parameters; calculating the Mahalanobis distance as a degree of deviation of the signal space data after coordinate transformation from the unit space data after coordinate transformation; diagnosing the soundness of the signal space data based on the Mahalanobis distance; A data diagnostic method comprising:

6. A data processing program for processing data to calculate a Mahalanobis distance as an evaluation index for determining whether signal space data indicating the operating state of a plant is abnormal, comprising: Using a computer device, calculating an objective function for linearizing a combination of each item included in unit space data, which is data of a unit space that serves as a reference when determining the operating state; calculating a coordinate transformation parameter for transforming the unit space data into coordinates for each of the items so as to minimize the objective function; a step of performing coordinate transformation on the unit space data for each of the items using the coordinate transformation parameters; standardizing the coordinate-transformed unit space data for each item; is executable, The step of calculating the objective function includes: calculating a correlation coefficient for each of the combinations; calculating a coefficient of determination based on the correlation coefficient; calculating a self-information content based on the coefficient of determination; calculating the objective function by multiplying the sum of the self-information by −1; a data processing program,

7. A data processing program for processing data for calculating a Mahalanobis distance as an evaluation index for determining whether signal space data indicating an operating state of a plant is abnormal, comprising: Using a computer device, calculating an objective function for linearizing a combination of each item included in unit space data, which is data of a unit space that serves as a reference when determining the operating state; calculating a coordinate transformation parameter for transforming the unit space data into coordinates for each of the items so as to minimize the objective function; a step of performing coordinate transformation on the unit space data for each of the items using the coordinate transformation parameters; standardizing the coordinate-transformed unit space data for each item; is executable, The step of calculating the objective function includes: Calculating skewness and kurtosis for each of the items; calculating the objective function by multiplying the squared values ​​of the skewness and the squared values ​​of the kurtosis by different weighting coefficients and adding the resulting values; a data processing program,

Images