JP3005593B2 - Diagnostic algorithm construction method and device - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method and apparatus for constructing a diagnostic algorithm, and more particularly, to a diagnostic algorithm construction method suitable for constructing a diagnostic algorithm for diagnosing abnormalities of various devices such as automobiles and home appliances. Regarding the device.

[0002]

2. Description of the Related Art An acoustic mission (AE) method is known as a method for diagnosing an abnormality of a device. A diagnostic device using the AE method is disclosed in, for example,
2-282258 and JP-A-62-19755.
What is described in the gazette is known.

In the former device, diagnosis is performed using a diagnosis algorithm specific to a diagnosis target, and in the latter device, a diagnosis algorithm is constructed using specific AE characteristic parameters. It is supposed to.

[0004]

In the above-mentioned prior art, no consideration is given to the versatility with which highly accurate diagnosis can be made for various types of diagnostic objects. In addition, it is necessary to develop a diagnostic algorithm, and the development efficiency is low. That is, in the case of the former device, a signal that has passed through a band of a specific wavelength is extracted from the AE signal, the signal is compared with a reference value, and the failure of the bearing is predicted from the comparison result. However, since the reference value differs depending on the diagnosis target, it cannot be used for devices other than the diagnosis target. Also, in the case of the latter device, the nature of the abnormality is mainly classified, and it is difficult to determine the abnormality determination level of the diagnosis target with high accuracy. Therefore, it is necessary to perform a diagnosis using an AE characteristic value that is optimal for the diagnosis target. Is impossible. Therefore, in the method applied to both devices, when performing highly accurate diagnosis, it is necessary to use a diagnostic device developed for each diagnostic object.

An object of the present invention is to provide a diagnostic algorithm construction method and an apparatus thereof capable of performing highly accurate diagnosis on various types of diagnostic objects.

[0006]

In order to achieve the above object, the present invention provides, as a first method, a method for diagnosing a plurality of diagnostic objects in a diagnostic object group divided into those having normal properties and those having abnormal properties. Measure the values related to the parameters and check each diagnostic parameter.
Statistical processing is performed on the measured values of the parameters, and based on the processing results, the effective parameters are selected from the diagnostic parameter group.
When extracting a plurality of diagnostic parameters predicted as parameters and determining a diagnostic level for judging the quality of the diagnosis target based on the measured values of the extracted effective diagnostic parameters , Any of the valid diagnostic parameters
From the measured values for this parameter,
Generates a division scale to be used and supports the generated scale
The normal value of the division scale is used as the distribution characteristic of the diagnosed object.
Abnormal distribution characteristics and abnormal values that indicate abnormal values on the division scale
Generate cloth properties and the intersection or boundary of each generated distribution property
The value is determined to a normal and abnormal diagnostic level.
This adopts a diagnostic algorithm construction method.

[0007] As a second method, values relating to a plurality of diagnostic parameters are measured for a group of diagnostic objects classified into those having normal properties and those having abnormal properties, and values for each diagnostic parameter are measured. Statistical processing is performed on the measured values, and based on the processing results, diagnostic parameters predicted as effective parameters are extracted from the diagnostic parameter group, and the extracted effective diagnostic parameters are extracted. The diagnostic level for judging the quality of the properties of the object to be diagnosed based on the measured values of the data
At this time, any of the effective diagnostic parameters
-Interval from the measured value of this parameter to this parameter
Generate division scales and diagnose corresponding to the generated scales
Normal indicating the normal value of the division scale as the distribution characteristic of the target
The distribution characteristic and the abnormal distribution characteristic that indicates the abnormal value of the
Generate and normalize the intersection or boundary value of each generated distribution characteristic
And a diagnosis algorithm construction method characterized in that the diagnosis level is determined to a diagnosis level of abnormality .

[0008] As a third method, those properties are normal and different
Multiple diagnoses for the usual diagnostic group
Measure the values related to the diagnostic parameters, and
Statistical processing is performed on the measured values of
Effective parameters from the group of diagnostic parameters based on the
Extract and extract multiple diagnostic parameters
Based on the measurements of the effective diagnostic parameters
Determine the diagnostic level for judging the quality of the diagnosis target.
When determining the parameters of the effective diagnostic parameters
Select two arbitrary parameters and select each of the selected diagnostic parameters.
From the measured values of the diagnostic group obtained for the
The correlation between the diagnostic parameters and each diagnostic parameter is shown.
The distribution characteristics of the diagnosis target group as
Boundary between normal diagnosis target and abnormal diagnosis target based on distribution characteristics
The value is determined to a normal and abnormal diagnostic level.
This adopts a diagnostic algorithm construction method.

As a fourth method , a plurality of diagnostic subjects are divided into a normal group and an abnormal group, and values relating to a plurality of diagnostic parameters are measured for each diagnostic subject in each group. The average value of the normal group is calculated according to the measurement value obtained from the group to be diagnosed of the normal group among the measurement values of each diagnostic parameter, and the normal group is calculated from the average value and the measured value. The variance of the loop is calculated for each diagnostic parameter,
The variance of the abnormal group is calculated for each diagnostic parameter from the average value and the measured value of the abnormal group, and the variance calculated for the normal group and the variance calculated for the abnormal group. Calculate the variance rate of each diagnostic parameter from the ratio of, and extract a plurality of effective diagnostic parameters from the diagnostic parameter group based on the calculated variance rate. The deviation amount between the measured value of the normal group and the average value of the normal group is calculated for each of the diagnostic parameters for the extracted diagnostic parameters, and the normal group is calculated from the calculated deviation amount.
Calculating an average value and a standard deviation of the deviation amount of the loop, calculating an individual certainty factor for each of the effective diagnostic parameters from these calculated values, and calculating an overall certainty factor based on each individual certainty factor;
A diagnostic algorithm construction method for determining a diagnostic level for determining the quality of the property of the diagnostic object according to the calculated value is employed.

As a fifth method , a value of a characteristic parameter group for a plurality of diagnostic objects is measured for each parameter, and the measured value is compared with an average value of a normal diagnostic object group. The deviation amount of each diagnostic parameter is calculated for each diagnostic object for each diagnostic parameter, the average deviation amount and the standard deviation of the normal diagnostic object are calculated from the calculated deviation amounts, and each diagnostic parameter is calculated from these calculated values. The method uses a diagnostic algorithm construction method that calculates the individual confidence for each data, calculates the overall confidence based on each individual confidence, and determines the diagnostic level for judging the quality of the diagnosis target according to the calculated value. It was done. As a sixth method, including any one of the first to fifth methods, parameters for cross - data are parameters relating to the waveform and frequency response of the diagnostic object - Oh with data
The diagnostic parameters related to waveform characteristics include an average value, a maximum value, a minimum value, a peak value, the number of occurrences, an energy value, an average duration, and an average rise time. 1 peak frequency and its component ratio, 2nd
The diagnostic algorithm construction method is characterized by including a peak frequency and its component ratio, a primary rotation frequency and this component ratio, and a secondary rotation frequency and its component ratio.

As a first apparatus, a measuring means for measuring values relating to a plurality of diagnostic parameters for a group of diagnostic objects including those having normal properties and abnormal properties, and a normal value among the measured values of the measuring means. The average value for the normal diagnosis target group was calculated according to the measurement values obtained from the various diagnosis target groups for each diagnostic parameter.
Average value calculating means for calculating each parameter, and first variance calculating means for calculating a variance for a normal diagnosis target for each diagnostic parameter from the measured value of the measuring means and the calculated value of the average value calculating means. The variance of an abnormal diagnosis target is calculated from the measured value of the measuring means and the calculated value of the average value calculating means for each diagnostic parameter.
A second variance calculating means for each parameter, a variance calculating means for calculating a variance of each diagnostic parameter from a ratio of the calculated values of the variance calculating means, and a variance calculating means based on the calculated values of the variance calculating means. Parameter extracting means for extracting a plurality of effective diagnostic parameters from the diagnostic parameter group, and measuring each diagnostic object based on the measured values of the measuring means and the calculated values of the average value calculating means The deviation amount between the value and the average value is calculated for each effective diagnostic parameter extracted by the parameter extraction unit for each diagnostic object, and the deviation amount is calculated from the calculated value of the deviation amount calculation unit. Calculating means for calculating a deviation average value and a standard deviation of the group of diagnosis targets, individual confidence calculating means for calculating individual confidence for each effective diagnostic parameter from the calculated value of the arithmetic means, and individual confidence. Total confidence based on the value calculated by the degree calculation means The diagnostic algorithm constructing device comprises: a total confidence calculating means for issuing a diagnostic algorithm; and a diagnostic level determining means for determining a diagnostic level for determining whether the property of the diagnosis target is good or bad according to the value calculated by the total certainty calculating means. is there.

As a second apparatus, a measuring means for measuring a plurality of diagnostic parameters with respect to a plurality of diagnostic objects, and a measured value obtained from a normal diagnostic object group among the measured values of the measuring means. Average value calculating means for calculating an average value for a normal diagnostic object group for each diagnostic parameter in accordance with the above, and a measured value for each diagnostic object based on the measured value of the measuring means and the calculated value of the average value calculating means. Amount calculating means for calculating a deviation amount from the average value for each diagnostic parameter for each diagnostic parameter, and calculating a deviation amount average value and a standard deviation of a normal diagnostic object group from the calculated value of the deviation amount calculating means. Calculating means for calculating;
Individual confidence calculation means for calculating individual confidence for each diagnostic parameter from the operation value of the operation means, and total confidence calculation means for calculating total confidence based on the value calculated by the individual confidence calculation means. And a diagnosis level determining means for determining a diagnosis level for determining whether the property of the diagnosis target is good or bad according to the value calculated by the total certainty degree calculating means.

[0013]

The values of a plurality of diagnostic parameters are measured for a group of diagnostic objects having normal and abnormal properties. Then, statistical processing is performed on the measured values for each diagnostic parameter, and the diagnostic parameters predicted as effective parameters based on the processing results.
A plurality of parameters are extracted, and a diagnostic level for judging the normality of the diagnosis target is determined based on the measured values of the extracted effective diagnostic parameters. Therefore, the diagnostic level is determined based on the measured values of the effective diagnostic parameters, so that highly accurate diagnosis can be performed for various diagnostic objects .

[0014]

An embodiment of the present invention will be described below with reference to the drawings.

FIG. 1 is a flowchart showing a processing method of a diagnostic algorithm construction method according to the present invention, and FIG. 2 shows an entire configuration of the apparatus according to the present invention. In FIG. 2, a plurality of rotating bodies such as a steam turbine, a generator, a water wheel, and a rolling mill are used as a diagnosis target 10, and each diagnosis target 10 has a normal property and an abnormal property. Classified. Each diagnostic object 10 is detected by a detector 12 with respect to various diagnostic parameters. The output signal of the detector 12 is sent to a computer 18 via an amplifier 14 and an A / D converter 16. The computer 18 includes a waveform analysis unit 20, a frequency analysis unit 22, and an arithmetic processing unit 24. The arithmetic processing unit 24 is connected to a display unit 26 and an input / output unit 28. The computer 18 comprises a detector 12, an amplifier 14, A
Along with the A / D converter 16, the measuring means is constituted, and the average value calculating means, the first variance calculating means, the second variance calculating means, the dispersion ratio calculating means, the parameter extracting means, the shift amount calculating means,
The calculation means, the individual certainty calculating means, the total certainty calculating means and the diagnosis level determining means are constituted.
Then, the computer 18 statistically processes values relating to a plurality of diagnostic parameters based on the detection signal from the detector 12, and
The diagnosis level is determined based on the processing result.

As diagnostic parameters, diagnostic parameters relating to waveform characteristics and diagnostic parameters relating to frequency characteristics are used. As shown in Tables 1 and 2, the AE parameters relating to the frequency characteristics include an average value, an average peak value, and an energy value. The AE parameters relating to the frequency characteristics include the first frequency. And its component ratio, the first peak frequency and its component ratio, the rotational primary frequency and its component ratio, the rotational secondary frequency and its component ratio, and the like.

[0017]

[Table 1]

[0018]

[Table 2]

The AE parameters relating to the waveform characteristics are measured by the methods shown in Table 3, and the measurement results are displayed on the display unit 26 as shown in FIG. The AE parameters relating to the frequency characteristics are measured by the method shown in Table 4, and the measurement results are displayed on the display unit 26 as characteristics as shown in FIG.

[0020]

[Table 3]

[0021]

[Table 4]

Next, the operation of the apparatus shown in FIG. 2 will be described with reference to the flowchart of FIG.

First, the plurality of diagnostic objects 10 are divided into normal products and abnormal products, and n normal products and m abnormal products are set (step 100). Is performed (step 102). That is, a detector 12 is installed on each diagnosis object 10 and each diagnosis object 1
The detection signal from 0 is transferred to the computer 18, the computer 18 performs a waveform analysis and a frequency analysis of the detection signal, and the analysis results are processed in the arithmetic processing unit 24. That is, the arithmetic processing unit 24 processes the measured value of the AE parameter relating to the waveform characteristic and the measured value of the AE parameter relating to the frequency characteristic.

In performing this arithmetic processing, the following equations (1) to (11) are used.
The average value of the E parameter is calculated according to equation (1) (step 104).

[0025]

(Equation 1)

Next, the variance of the normal product is calculated from the measured value of the normal product and the average value of the normal product according to the equation (2) (Swip 106). Further, the variance of the abnormal product is calculated from the measured value of the abnormal product and the average value of the normal product according to the equation (3) (step 1).
08). Then, the dispersion ratio is calculated according to the equation (4) as a ratio or a difference between the dispersion of the normal product and the dispersion of the abnormal product (step 110). These processes are performed for each AE parameter, and a plurality of effective parameters are extracted from the dispersion ratio of each AE parameter (step 112). For example, four AE parameters are selected in descending order of the dispersion ratio. In this case, when AE parameters extracted from the AE parameter group include items of the same family (see Table 1), the family items are sequentially moved up as the top one item. If the four items are of the same type, the top item of the other type is replaced with the fourth item.

4 items of AE as effective AE parameters
After the parameters have been extracted, the amount of deviation of each AE parameter from the average value of normal products is calculated for each diagnosis target 10 according to equation (5) (step 114). Then, based on the deviation amount from the average value of the normal product, the average deviation amount and the standard deviation of the normal product are calculated according to the equations (6) and (7). Then, the individual certainty factor of each effective AE parameter is determined based on these calculated values (step 116).

In determining the individual certainty factor, in this embodiment, the criterion of the individual certainty factor R is set as follows.

[0029]

(Equation 2)

When obtaining the individual certainty factor of each effective AE parameter, as shown in FIG.
A correlation diagram relating to the parameter deviation is created, and the individual certainty factor R is determined based on the maximum value of the deviation of the normal product.
As shown in FIG. 6, the distribution characteristic of each diagnosis target 10 is created with respect to the effective AE parameter, and the individual certainty factor is determined from the distribution characteristic.

As shown in FIG. 5 or FIG. 6, individual confidences R1, R2, and R3 are selected as those having high confidence as individual confidence R, and total confidence M is calculated using each individual confidence R1, R2, and R3. (Step 118).

In calculating the total certainty factor M, the fuzzy inference method includes a minimum value operation method, a multiplication operation method, a maximum value operation method, a combine function method, and the like. The certainty factor M is to be calculated.

When calculating the overall confidence M,
Equations (8) to (11) are used. First, the first overall certainty factor M1 is obtained according to the equation (8) using the certainty factor R1 of the first effective parameter.

Next, the second overall certainty factor M2 is obtained in accordance with the equation (9) using the certainty factor R2 of the second effective parameter.

Hereinafter, similarly, the third overall certainty M3 and the final overall certainty Ms are obtained according to the equations (10) and (11).

When these processes are performed, as shown in FIG. 7, the individual certainty factors R1, R2, and R3, for which the distinction between the normal product and the abnormal product cannot be clearly identified, determine the total certainty factor M. This makes it possible to clearly distinguish normal products from abnormal products.

After calculating the final overall confidence Ms, FIG.
As shown in the above, the final overall confidence Ms is plotted on the horizontal axis,
A distribution chart is created with the frequency on the vertical axis, and the diagnostic object 10
The intersection of a normal product and an abnormal product is determined as a diagnostic level as a diagnostic level for determining the quality of the properties of the product, and a normal product and an abnormal product are determined according to the determination level (step 120). Thereby, a diagnostic algorithm can be constructed. In this case, a trial calculation is performed so that the value of the standard deviation of a normal product becomes a value with a high correct answer rate.

When determining the determination level, as shown in FIG. 9, the maximum value of the final total certainty factor Ms is set to 1
It is also possible to divide the value of the final total certainty factor Ms into four and use the judgment level of the property of the diagnosis target 10 as normal, small abnormal, medium abnormal, or large abnormal by the value of the final total certainty factor Ms. .

In the above-described embodiment, a case where normal products and abnormal products are classified as diagnosis targets has been described. However, effective AE parameters can be obtained from measured values of the diagnosis target 10 including normal products and abnormal products. When extracted, a diagnosis algorithm for various diagnosis targets 10 can be constructed by the device shown in FIG. 10 as a diagnosis on an actual line.

In this case, an effective parameter analyzing section 30 is used for the computer 18 instead of the waveform analyzing section 20 and the frequency analyzing section 22, and the effective parameter analyzing section 30 is used for a specific effective parameter. By performing waveform analysis and frequency analysis and outputting the analysis results to the arithmetic processing unit 24, as shown in FIG. 11, measurement of each of the diagnostic objects 10 with respect to effective AE parameters P1, P2, and P3. , The determination level for the diagnosis target 10 can be determined.

In deciding the judgment level, as shown in FIG. 12, a division scale for the effective AE parameter P1 is obtained, and the number of samples corresponding to this scale is divided into a normal product and an abnormal product and distributed. In this way, the normal product distribution characteristic and the abnormal product distribution characteristic are created, and the intersection or boundary value of each distribution characteristic can be set to the determination level P1 (S). Further, as shown in FIG. 13, the distribution of valid AE parameters P1 and P2 between the normal product and the abnormal product is created, and the boundary value of the zone between the normal product and the abnormal product is determined at the determination level P1 (S1). ) P2 (S).

If a known knowledge using a database is introduced in addition to the statistical processing and the knowledge engineering processing, a more reliable diagnosis algorithm can be constructed. For example, the characteristic form of the AE signal is checked in advance, and this characteristic form is used as a reference when determining effective parameters.

[0043]

As described above, according to the present invention,
The values of a plurality of diagnostic parameters are measured for a group of diagnostic subjects classified into those having normal properties and those having abnormal properties, statistical processing is performed on the measured values, and effective parameters are obtained from the processing results. Since the diagnostic parameters predicted as parameters are extracted and the diagnostic level is determined based on the measured values of the extracted effective diagnostic parameters, high accuracy is achieved for various diagnostic objects. Can make a diagnosis,
A highly reliable diagnostic algorithm can be constructed for various diagnostic objects.

[Brief description of the drawings]

FIG. 1 is a flowchart for explaining the operation of the present invention.

FIG. 2 is an overall configuration diagram of an apparatus to which the present invention is applied.

FIG. 3 is an AE detection waveform diagram.

FIG. 4 is a frequency characteristic diagram of an AE signal.

FIG. 5 is a correlation diagram of a shift amount.

FIG. 6 is a diagram showing a relationship between a shift amount and a frequency.

FIG. 7 is a diagram for explaining a relationship between an individual certainty factor and a total certainty factor;

FIG. 8 is a diagram for explaining the relationship between the total certainty factor and the frequency.

FIG. 9 is a diagram for explaining the relationship between the total certainty factor and the determination level.

FIG. 10 is a configuration diagram showing another embodiment of the present invention.

FIG. 11 is a flowchart for explaining the operation of the apparatus shown in FIG. 10;

FIG. 12 is a diagram for explaining another method of determining a determination level.

FIG. 13 is a diagram for explaining another method for determining a determination level.

[Description of Signs] 10 ... Diagnosis object 12 ... Detector 14 ... Amplifier 16 ... A / D converter 18 ... Computer 20 ... Waveform analysis unit 22 ... Frequency analysis unit 24 ... Operation processing unit 26 ... Display unit 28 ... On Output unit 30 ... Effective parameter analysis unit

────────────────────────────────────────────────── ─── Continuation of the front page (72) Inventor Shozo Michimoto 2520, Oaza Takaba, Katsuta City, Ibaraki Pref. Sawa Plant, Hitachi, Ltd. JP-A-62-282258 (JP, A) JP-A-62-19755 (JP, A) (58) Fields investigated (Int. Cl. ⁷ , DB name) G01N 29/14 G01M 19/00

Claims (8)

(57) [Claims] 1. A method for measuring values of a plurality of diagnostic parameters for a group of diagnostic subjects classified into normal and abnormal ones, and statistically calculating the measured values for each of the diagnostic parameters. Based on the result of the processing, extract a plurality of diagnostic parameters predicted as effective parameters from a group of diagnostic parameters based on the processing result, and extract the extracted effective diagnostic parameters. When determining the diagnostic level for determining the quality of the properties of the diagnosis target based on the measured value of,
For any of the valid diagnostic parameters
From the measured values, the division scale for this parameter is calculated.
Generated and distribution of diagnostic objects corresponding to the generated scale
Normal distribution characteristics that indicate the normal value of the section scale as characteristics
Generates an abnormal distribution characteristic that indicates the abnormal value of the section
The intersection or boundary value of each generated distribution characteristic is diagnosed as normal or abnormal.
A diagnostic algorithm constructing method characterized in that the diagnostic algorithm is determined at a disconnection level . 2. A method for measuring a plurality of diagnostic parameters for a group of diagnostic subjects classified into normal and abnormal ones, and statistically calculating the measured values for each of the diagnostic parameters. Based on the result of the processing, extracting a diagnostic parameter that is predicted to be an effective parameter from the diagnostic parameter group based on the processing result, and extracting the extracted effective diagnostic parameter.
Effective in determining the diagnostic level for determining the quality of the properties of the diagnosis target based on the measured values of
Measurement of any of the diagnostic parameters
Generate a division scale for this parameter from a constant value
And the distribution characteristics of the diagnostic object corresponding to the generated scale
Distribution characteristics indicating the normal value of the section division scale
Generates an abnormal distribution characteristic indicating the outlier of the scale
The intersection or boundary value of each distribution characteristic is evaluated for normal and abnormal diagnostic levels.
A diagnostic algorithm constructing method characterized in that the algorithm is determined in a simple manner. 3. Classification is divided into normal and abnormal properties.
Multiple diagnostic parameters for the diagnostic group
Measured values, and measured values for each diagnostic parameter
Statistical processing is performed on the
Diagnosis predicted to be effective parameters from the parameter group
Extract multiple effective diagnostic parameters and extract the extracted effective diagnostic parameters.
Based on the measured values of the parameters,
When determining a diagnostic level for determining pass / fail,
Any of the extracted effective diagnostic parameters
Select two parameters and for each selected diagnostic parameter
From the obtained measured values of the group to be diagnosed,
Diagnosis as information indicating the correlation with each diagnostic parameter
Generate distribution characteristics of the target group and normalize from the generated distribution characteristics
The boundary value between a normal diagnosis target and an abnormal diagnosis target is
A diagnostic algorithm construction method characterized in that the diagnostic algorithm is determined . 4. A plurality of diagnostic subjects are divided into a normal group and an abnormal group, and values relating to a plurality of diagnostic parameters are measured for each diagnostic subject in each group, and each diagnostic subject is measured. The average value of the normal group is calculated according to the measured values obtained from the diagnostic group of the normal group among the measured values of the parameters, and the variance of the normal group is calculated from the average value and the measured value. It was calculated for each diagnostic parameter, and the variance of the abnormal group was calculated for each diagnostic parameter from the average value and the measured value for the abnormal group, and was calculated for the normal group. The variance of each diagnostic parameter is calculated from the ratio or difference between the variance and the variance calculated for the abnormal group, and an effective diagnostic parameter is selected from the diagnostic parameter group based on the calculated variance. A plurality of parameters are extracted, and the measured values for The deviation from the average value for the normal group is calculated for each of the diagnostic parameters for the extracted diagnostic parameters, and the average deviation and the standard deviation for the normal group are calculated from the calculated deviation. From these calculated values, an individual certainty factor for each effective diagnostic parameter is calculated, a total certainty factor is calculated based on each individual certainty factor, and the quality of the diagnosis target is determined based on the calculated value. Algorithm construction method that determines the diagnostic level for the diagnosis. 5. The method according to claim 5, wherein the value of the characteristic parameter group for each of the plurality of diagnostic objects is measured for each parameter, and the difference between the measured value and the average value of the normal diagnostic object group is determined. Calculate each diagnostic parameter for each diagnostic object, calculate the average deviation and standard deviation of the normal diagnostic target from the calculated deviation, and calculate individual diagnostic parameters for each diagnostic parameter from these calculated values. A diagnostic algorithm construction method for calculating a certainty factor, calculating a total certainty factor based on each individual certainty factor, and determining a diagnostic level for determining whether the property of the diagnosis target is good or bad according to the calculated value. 6. The diagnostic parameter is a parameter relating to a waveform characteristic and a frequency characteristic of a diagnosis target, and is related to a waveform characteristic.
The diagnostic parameters used are average, maximum, minimum, and wave height.
Value, number of occurrences, energy value, average duration, average rise
Including time, diagnostic parameters for frequency characteristics are
The first peak frequency and its component ratio, the second peak frequency and its component
Component ratio, rotation primary frequency and its component ratio, rotation secondary frequency
The diagnostic algorithm construction method according to any one of claims 1 to 5 , comprising a number and a component ratio thereof . 7. Measuring means for measuring values relating to a plurality of diagnostic parameters for a group of diagnostic objects including those having normal and abnormal properties, and a group of normal diagnostic objects among the measured values of the measuring means. Average value calculating means for calculating an average value for a normal diagnosis target group for each diagnostic parameter in accordance with the measured values obtained from the above, and a normal value based on the measured value of the measuring means and the calculated value of the average value calculating means. First variance calculating means for calculating the variance for the diagnostic object for each diagnostic parameter, and variance for the abnormal diagnostic object based on the measured value of the measuring means and the calculated value of the average value calculating means. A second variance calculating means for each parameter, a variance calculating means for calculating a variance of each diagnostic parameter from a ratio of the calculated values of the variance calculating means, and a variance calculating means for calculating the variance of each diagnostic parameter. For effective diagnosis from the group of diagnostic parameters A parameter extracting means for extracting a plurality of parameters; and a deviation amount between the measured value and the average value for each diagnostic object based on the measured value of the measuring means and the calculated value of the average value calculating means. A shift amount calculating means for calculating each effective diagnostic parameter extracted by the extracting means for each diagnostic object, and a shift amount average value and a standard deviation of a normal diagnostic object group are calculated from the calculated values of the shift amount calculating means. Calculating means for calculating, and each effective diagnostic parameter from the calculated value of the calculating means.
An individual certainty factor calculating means for calculating an individual certainty factor of the data, a total certainty factor calculating means for calculating the total certainty factor based on the calculated value of the individual certainty factor calculating means, and a diagnosis according to the calculated value of the total certainty factor calculating means. A diagnostic algorithm construction device comprising: a diagnostic level determining unit that determines a diagnostic level for determining whether the properties of a target are good or bad. 8. Measuring means for measuring values relating to a plurality of diagnostic parameters for a plurality of diagnostic objects, and a normal diagnosis in accordance with a measured value obtained from a normal group of diagnostic objects among the measured values of the measuring means. Mean value calculating means for calculating an average value for the target group for each diagnostic parameter, and a measured value and an average value for each diagnostic object based on the measured value of the measuring means and the calculated value of the average value calculating means. Deviation calculating means for calculating the deviation amount of each diagnostic parameter for each diagnostic parameter, and calculating means for calculating the average deviation amount and the standard deviation of the normal diagnosis target group from the calculated value of the deviation amount calculating means. And individual confidence calculation means for calculating individual confidence for each diagnostic parameter from a calculation value of the calculation means; and total confidence calculation for calculating total confidence based on the calculation value of the individual confidence calculation means. Means,
A diagnostic algorithm construction device comprising: a diagnostic level determining unit that determines a diagnostic level for determining whether the properties of a diagnosis target are good or bad according to a value calculated by a total certainty factor calculating unit.