JPH08129691A - Fault diagnostic device utilizing expert system - Google Patents

Abstract

PURPOSE: To provide an operator with fault prediction information regarding object facilities before a fault occurs by deciding a fault factor in case of fault occurrence on the basis of defined decision conditions and a decision rule. CONSTITUTION: The expert system defines the quality judgment as diagnostic and prediction definition item as to fault on an elevator and the tread judgment and difference judgment as to measurement data as monitor information on the elevator as diagnostic rule 6 and prediction rule 7. Then, an estimated fault position is defined. If the elevator gets out of order, the fault data of monitor information 3 gathered by an on-line system 8 changes and an inference engine 4 is actuated to perform inference according to the diagnostic rule 6 of a knowledge base 2, and an estimated fault position is extracted and displayed of a man-machine device 5.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention, when an equipment such as an elevator or an escalator installed in a station premises fails, or before a failure occurs, a failure occurrence portion, a failure occurrence portion, or a failure occurrence time can be determined. The present invention relates to a failure diagnosis device using an expert system that estimates and informs an operator or a maintenance worker of a site where a failure has occurred and assists in promptly taking countermeasures.

[0002]

2. Description of the Related Art FIG. 6 is a block diagram showing the configuration of a conventional fault diagnosis apparatus using an expert system.
In the figure, 1 is a target machine facility (hereinafter referred to as target facility), 2 is a knowledge database that stores diagnostic rules for the target facility 1 using a knowledge expression such as a rule base, and 3 is online from the target facility 1 The input monitoring information, 4 is an inference engine for performing a failure diagnosis by using the monitoring information 3 and the knowledge database 2, and 5 is a display of the failure diagnosis result inferred by the inference engine 4 and the input of data for constructing the knowledge database 2. It is a man-machine device for.

Next, the operation will be described. Now, when a failure occurs in the target equipment 1, the monitoring information 3 changes. The inference engine 4 matches the predetermined monitoring information 3 with the knowledge database 2 to identify the faulty part of the target equipment 1 and output the diagnosis result to the man-machine device 5. Also,
The inference engine 4 operates by registering in advance the know-how related to failure diagnosis of an expert (skilled person) from the man-machine device 5 as a rule base in the knowledge database 2.

[0004]

The conventional failure diagnosis device using the expert system is constructed as described above, and the expert system is used to carry out the failure diagnosis based on the result of the failure, and the diagnosis result is displayed. Since it is provided to operators and maintenance workers, it is effective in reducing the time from the occurrence of a failure to the recovery of the failure, so-called downtime, while the function as preventive maintenance to prevent the failure is provided. There is a problem in that it cannot be used for preventive maintenance in order to prevent failures before it has it.

The present invention has been made in order to solve the above-mentioned problems, and a failure diagnosis apparatus using an expert system capable of providing failure prediction information about a target facility to an operator before a failure occurs. Aim to get.

Another object of the present invention is to obtain a failure diagnosis device using an expert system that can simplify the system configuration.

[0007]

According to a first aspect of the present invention, there is provided a failure diagnosis apparatus using an expert system, wherein a judgment condition and a judgment condition for a phenomenon that may occur in a target equipment are defined by a matrix structure. A failure is made by defining means and failure diagnosis at the time of failure of each part of the target equipment and failure prediction before failure occurrence based on the judgment type and the judgment condition and the judgment rule defined by the judgment condition defining means. A failure factor determining means for determining a factor is provided.

A failure diagnosing apparatus using an expert system according to a second aspect of the present invention comprises a judgment type and a judgment condition to be applied to a phenomenon that is currently occurring, based on monitoring information obtained from the target equipment. A diagnosis rule and a prediction rule including a judgment type and a judgment condition to be applied to a state of phenomenon change when the above phenomenon changes with time, a degree to an abnormal state, a probability of occurrence of an abnormality in the future, etc. And a determination condition defining means defined by a matrix structure.

A failure diagnosis apparatus using the expert system according to the third aspect of the present invention is a pass / fail judgment applied for making a diagnosis based on binary information for the monitoring information of the target equipment as the judgment type and its judgment condition. And conditions for making a pass / fail judgment, and conditions for making an upper / lower limit judgment and its upper / lower limit judgment applied to make a diagnosis based on threshold information defining the upper and lower limits, and based on a fuzzy function The fuzzy judgment applied for making a diagnosis and the conditions for making the fuzzy judgment, the stage judgment applied for making a diagnosis based on the human sensory criteria, and the condition for making the stage judgment, and the time The tendency judgment to be applied for making a diagnosis based on the tendency that changes with the progress of and the conditions for making the judgment, and the difference amount between the phenomenon that is currently occurring and the phenomenon that has occurred in the past. It is obtained by a condition of making a difference determination and the difference decision to apply in order to perform the diagnosis.

A fault diagnosis apparatus using an expert system according to a fourth aspect of the present invention has a configuration for collecting monitoring information such as operation data / fault data on target equipment and test data obtained by performing inspection online. It is a thing.

A failure diagnosis apparatus using an expert system according to a fifth aspect of the present invention is provided with a configuration for collecting monitoring information such as operation data / fault data regarding target equipment and test data obtained by performing an inspection offline. It is a thing.

A failure diagnosis apparatus using an expert system according to a sixth aspect of the present invention is a knowledge database that combines monitoring information such as operation data and failure data related to target equipment and test data collected by inspection. The failure diagnosis and the failure prediction are performed based on the diagnosis rule such as the judgment type and the judgment condition applied to the phenomenon that is currently occurring, and the part where the failure occurs is predicted.

A failure diagnosis apparatus using an expert system according to a seventh aspect of the present invention is a knowledge database that combines monitoring information such as operation data and failure data of target equipment and test data collected by inspection. It is equipped with a configuration for performing failure diagnosis prediction based on the prediction rule of (3) and predicting a site where a failure occurs.

[0014]

According to another aspect of the present invention, there is provided a failure diagnosis apparatus using an expert system, wherein a judgment type and a judgment condition for a phenomenon that may occur in a target equipment are defined by a matrix structure, and the defined judgment type and the judgment are defined. Based on conditions and judgment rules, failure diagnosis is performed for each part of the target equipment at the time of failure occurrence and failure prediction before failure occurrence is performed to determine failure factors, and failure is caused by failure diagnosis and failure prediction for the target equipment. It is possible to provide the operator with failure prediction information of a portion that is expected to occur.

The determination condition defining means of the fault diagnosis apparatus utilizing the expert system according to the second aspect of the present invention comprises a determination rule to be applied to a phenomenon which is currently occurring and a diagnostic rule consisting of the determination condition and the passage of time. When the phenomenon changes with the phenomenon, the predictive rule including the state of the phenomenon change and the degree to the abnormal state where the failure occurs, the determination type and the determination condition for determining the accuracy of the abnormality occurrence are the target equipment. It is possible to provide the operator with the failure prediction information of the part that is expected to fail by combining the failure diagnosis and the failure prediction of the target equipment by defining the monitoring information obtained from the operator.

According to a third aspect of the present invention, there is provided a failure diagnosis device using an expert system, which is a threshold for defining a pass / fail judgment for making a diagnosis of monitoring information based on binary information and its judgment condition, and an upper limit and a lower limit. Upper and lower limit judgment for making a diagnosis such as the degree of approach to a state regarded as abnormal based on information and its judgment conditions, and fuzzy judgment for making a diagnosis about the probability of abnormal occurrence based on a fuzzy function And its determination conditions, and the stage determination and its determination conditions for making a diagnosis based on the sensory criteria by the human five senses,
Difference between the current judgment and the past judgment, and the tendency judgment and the judgment conditions for making a diagnosis based on the tendency of the monitoring data that changes with the passage of time until the state of occurrence of an abnormality occurs. The difference judgment for making a diagnosis based on the quantity and the judgment condition are combined to perform the failure diagnosis and the failure prediction for the target equipment, and the failure prediction information of the part that is expected to fail is generated before the failure occurs. It is possible to provide to.

A failure diagnosis apparatus using an expert system according to a fourth aspect of the present invention monitors by collecting on-line monitoring information such as operation data / fault data regarding target equipment and test data detected by inspection. It is possible to efficiently collect information, combine failure diagnosis and failure prediction of target equipment, and provide the operator with failure prediction information of a site where a failure is expected before the occurrence of failure.

A failure diagnosis apparatus using the expert system according to the fifth aspect of the present invention collects monitoring information such as operation data / failure data relating to the target equipment and test data obtained by performing an inspection off-line to perform failure diagnosis. It is also possible to perform failure prediction, realize the simplification of the system configuration off-line, and provide the operator with failure prediction information of a part that is expected to fail before a failure occurs.

A fault diagnosis apparatus using an expert system according to the invention of claim 6 combines monitoring information such as operation data / fault data regarding target equipment and test data collected by performing inspection, and only a diagnosis rule of a knowledge database is provided. Performs failure diagnosis and failure prediction based on
To realize a failure diagnosis device using an expert system with a simplified system configuration.

According to a seventh aspect of the present invention, there is provided a fault diagnosis apparatus utilizing an expert system, which combines monitoring information such as operation data / fault data relating to target equipment and test data collected by performing inspections, thereby eliminating the need for a diagnosis rule. A fault diagnosis device using an expert system with a simplified system configuration is realized by performing fault diagnosis and fault prediction based only on the prediction rules of the knowledge database.

[0021]

【Example】

Example 1. An embodiment of the present invention will be described below with reference to the drawings. First Embodiment FIG. 1 is a block diagram showing the configuration of a failure diagnosis device using an expert system according to a first embodiment of the present invention. In FIG. 1, parts that are the same as or correspond to those in FIG. 6 are given the same reference numerals and description thereof is omitted. In the figure, 6 is a diagnostic rule for determining whether or not an abnormality has occurred in a part corresponding to the monitoring information based on the monitoring information,
Reference numeral 7 is a prediction rule for predicting a failure occurrence based on the monitoring information. Reference numeral 8 is an online system for collecting monitoring information online from the target facility 1. The diagnosis rule 6 and the prediction rule 7 correspond to the judgment rule and the failure factor judgment means, the inference engine 4 corresponds to the failure factor judgment means, and the man-machine device 5 corresponds to the judgment condition definition means.

FIG. 2 is a diagnosis / prediction definition table for defining diagnosis / prediction conditions as to which judgment type and judgment conditions are applied to each piece of monitoring information to perform failure diagnosis prediction. Monitoring information S1 obtained from the equipment 1
The definition contents of the determination type and the determination condition applied to each of S2, S3, ... Are set. This definition content is a pass / fail judgment that determines whether the monitor information is good or bad, a degree of change of the monitor information is determined, and a tendency determination that predicts a period until reaching a state considered abnormal, A difference judgment for observing the degree of deterioration based on the difference amount from the previous time in the monitoring information and making a judgment to detect abnormalities early, a fuzzy judgment that calculates the probability of occurrence of an abnormality based on a fuzzy function using fuzzy theory, There are upper and lower limit judgments that judge the degree to which the monitoring information is approaching abnormal based on the upper and lower limit values that have been set, and stage judgments that judge the monitoring information based on the human senses. It is set based on the know-how obtained from

FIG. 3 is a causal relationship table in which the relationship between the diagnosis / prediction conditions and the estimated failure part is defined in a matrix structure. Based on the know-how that an expert has gained from many years of experience, the judgment type And the determination condition and the like in the determination type are set in association with the estimated failure site.
These diagnostic prediction definition table and causal relationship table are registered in the knowledge database 2 as diagnostic rules 6 and prediction rules 7.

In the diagnosis / prediction definition table shown in FIG. 2, diagnosis / prediction definition items (various judgment types such as quality judgment and tendency judgment) to be applied to each monitoring information such as S1, S2, S3, ... ) Is defined, the quality determination is applied to the monitoring information S1, and the tendency determination and the difference determination are applied to the monitoring information S2.

Next, the operation will be described. In this case, elevator failure monitoring will be described as an example. First, as a diagnostic rule 6 and a predictive rule 7 by a so-called expert system, which utilizes the know-how that an expert has gained from many years of experience, as shown in FIG. The quality item is defined as a definition item, and the tendency determination and the difference determination are defined for the measurement data S2 that is the monitoring information of the elevator. Next, the quality judgment of the elevator failure data S1 is defined by one diagnosis / prediction condition A in the causal relationship table shown in FIG. 3, and "a" is estimated as a failure part estimated when a phenomenon satisfying this diagnosis / prediction condition occurs. Is defined.

When a failure occurs in the elevator of the target equipment 1, the failure data S1 of the monitoring information 3 collected by the online system 8 changes, the inference engine 4 is activated, and inference is made by the diagnosis rule 6 of the knowledge database 2. Then, “a” is extracted as the estimated failure part and displayed on the man-machine device 5.

Similarly, the elevator measurement data S2 is constantly observed and taken into the inference engine 4 as the monitoring information 3. If there is a change in this data, the inference engine 4 is started and the inference is made by the prediction rule 7 of the knowledge database 2. The estimated failure part “b” and the failure occurrence prediction time are extracted and displayed on the man-machine device 5. As the predicted occurrence time of the failure, the tendency and difference amount of the measurement data S2, which changes with time of the monitoring information 3 that is constantly observed, are determined by the trend determination and the difference determination, and the timing of the failure occurrence is predicted. It is determined.

FIG. 4 is an explanatory view showing a display screen showing the diagnosis and prediction result displayed on the man-machine device 5, and in this case, the monitoring information name 11 (S2 in this case) which is the factor of the diagnosis and prediction judgment. The failure part 12 (in this case, part "b") inferred as a result of the diagnosis prediction is displayed. Furthermore, according to the prediction rule 7, the diagnosis / prediction condition B for the tendency determination
When it becomes clear that the tendency of the monitoring information S2 satisfies
As the prediction information, the recommended replacement time 13 of the parts and the state of the monitoring information used for the prediction that changes with time are displayed as a graph 14.

Therefore, according to the first embodiment, not only the location of the failure that has occurred by the expert system is specified, but also the appearance of the failure that may occur in the future in the specific area of the target equipment from the monitoring information such as measurement data. It is possible to present the operator with the failure prediction information such as the time and the monitoring information used for the failure occurrence prediction.

Further, in the embodiment described above, the judgment type and the judgment condition for the phenomenon that may occur in the target equipment are defined on the matrix by the matrix structure based on the know-how of the skilled person. Even if the operator is not accustomed to the operation of the engine 4, it is possible to easily change the determination type and the determination condition shown in FIGS. 2 and 3 by using the man-machine device 5.

Embodiment 2 FIG. In the first embodiment, the monitoring information 3 is fetched from the target equipment 1 online, but an expert system capable of predicting failure diagnosis is used even if, for example, inspection data is collected offline and fetched as the monitoring information 3. It is possible to construct a fault diagnosis device that has

FIG. 5 is a block diagram showing the structure of a failure diagnosis apparatus using the expert system according to the second embodiment. 5, parts that are the same as or correspond to those in FIG. 1 are assigned the same reference numerals and explanations thereof are omitted. In the figure, reference numeral 21 is a data collection device for collecting monitoring information off-line, for example, a pen computer. In this data collection device 21, for example, the operator reads the state of each part of the target equipment from a state display device that detects and displays the state of each part of the target equipment, and inputs it by pen input or the like.

Reference numeral 22 is an input device for inputting monitoring information such as inspection data collected by the data collecting device 21 via a storage medium such as an IC card or a floppy disk.

Therefore, in the second embodiment, not only the failure part is specified by the expert system, but also the appearance time of the failure that may occur in the future in the specific part of the target equipment from the monitoring information such as the measurement data collected offline. Alternatively, it is possible to present the operator with failure prediction information such as monitoring information used for the failure occurrence prediction.

Example 3. Although the diagnosis rule 6 and the prediction rule 7 are provided as the knowledge database in the first embodiment, the failure diagnosis prediction is substantially performed by incorporating only the diagnosis rule 6 as in the case where the prediction rule 7 is incorporated. It is also possible to configure a failure diagnosis device using an expert system that enables the above.

A fault diagnosis apparatus using the expert system of this embodiment is defined by the upper and lower limits in the upper and lower limit determination, for example, which determines the monitoring information based on the set upper and lower limits. The range is set as a range in which a failure is considered to occur after a certain period of time, and it is determined from the monitoring information such as inspection data whether the inspection data is within the range, and the time of future failure occurrence. Can be predicted, and since the prediction rule is not provided, the system configuration can be simplified.

Embodiment 4 FIG. In the first embodiment, the knowledge database 2 is configured to include the diagnosis rule 6 and the prediction rule 7. However, by incorporating only the prediction rule 7 in the knowledge database 2, a failure prediction or a failure is considered. It is also possible to configure a failure diagnosis device that uses an expert system that enables failure diagnosis of an intermediate state that is moving, and the system configuration can be simplified.

[0038]

As described above, according to the first aspect of the present invention, the determination condition defining means for defining the determination type and the determination condition for the phenomenon that may occur in the target equipment by the matrix structure, and the determination condition definition Based on the determination type defined by the means and its determination conditions and determination rules,
An abnormality such as a failure that occurs in the target equipment is configured so as to include a failure factor determination means that determines failure factors by performing failure diagnosis and failure prediction before failure occurrence for each part of the target equipment. There is an effect that a failure diagnosis device using an expert system that can provide the failure prediction information about the information to the operator before the failure occurs.

According to the second aspect of the present invention, there is a change in the phenomenon that occurs when the phenomenon changes with the passage of time and a diagnostic rule consisting of the judgment type and the judgment conditions applied to the phenomenon that is currently occurring. Since the prediction rule consisting of the judgment type and its judgment conditions to be applied is configured to have a judgment condition defining means for defining the monitoring information obtained from the target equipment by a matrix structure, an abnormality such as a failure occurring in the target equipment There is an effect that a failure diagnosis device using an expert system that can provide the failure prediction information about the information to the operator before the failure occurs.

According to the third aspect of the present invention, as the judgment type and its judgment condition, the quality judgment and the condition for making the quality judgment, the upper and lower limit judgment and the condition for making the upper and lower limit judgment, and the fuzzy judgment are used. And conditions for performing the fuzzy determination, stage determination and conditions for performing the stage determination, trend determination and conditions for performing the trend determination, and difference determination and conditions for performing the difference determination. Since it is configured to be provided, failure prediction information about an abnormality such as a failure that occurs in the target equipment, the failure judgment, upper and lower limit judgment, fuzzy judgment, step judgment, tendency judgment,
There is an effect that a failure diagnosis device using an expert system that can provide the operator before a failure occurs by using the difference determination or the like can be obtained.

According to the invention of claim 4, the monitoring information such as the operation data / fault data regarding the target equipment and the test data obtained by the inspection is collected online, so that the operator can monitor the monitoring information. There is an effect that a failure diagnosis device that uses an expert system can be provided that can quickly provide operator with failure prediction information about the target equipment by monitoring information that has been efficiently collected without imposing a collection burden. .

According to the invention of claim 5, the monitoring information such as the operation data / fault data regarding the target equipment and the test data obtained by the inspection is collected off-line, so that the system configuration is complicated. There is an effect that a failure diagnosis device using an expert system can be provided that can provide the operator with failure prediction information about the target equipment before the occurrence of a failure without performing the above.

According to the sixth aspect of the present invention, the monitoring information of the operation data and the failure data of the target equipment and the monitoring information such as the test data collected by the inspection are combined, and the failure diagnosis is performed based on the diagnosis rule of the knowledge database. And, since it is configured to perform failure prediction, it is possible to provide a failure diagnosis device that uses an expert system with a simplified system configuration that can provide failure prediction information about target equipment to an operator before a failure occurs. There is.

According to the invention of claim 7, the monitoring information of the operation data and the failure data of the target equipment and the monitoring information such as the test data collected by the inspection are combined, and the failure diagnosis and the failure are performed by the prediction rule of the knowledge database. Since the system is configured to perform prediction and output the failure prediction information about the failure site as a result of the diagnosis, it is possible to provide the operator with failure prediction information about the target equipment before the failure occurs. There is an effect that a fault diagnosis device using an expert system that is simplified can be obtained.

[Brief description of drawings]

FIG. 1 is a block diagram showing a configuration of a failure diagnosis device using an expert system according to a first embodiment of the present invention.

FIG. 2 is an explanatory diagram showing a diagnostic prediction definition table for defining a diagnostic prediction condition for which diagnostic prediction is to be performed for each piece of monitoring information in the failure diagnostic device using the expert system according to the first embodiment of the present invention. is there.

FIG. 3 is an explanatory diagram showing a causal relationship table in which a relationship between a diagnosis / prediction condition and an estimated failure site in the failure diagnosis device using the expert system according to the first embodiment of the present invention is defined by a matrix structure.

FIG. 4 is an explanatory diagram showing a display screen showing a diagnosis prediction result displayed on the man-machine device in the failure diagnosis device using the expert system according to the first embodiment of the present invention.

FIG. 5 is a block diagram showing a configuration of a failure diagnosis device using an expert system according to a second embodiment of the present invention.

FIG. 6 is a block diagram showing a configuration of a conventional failure diagnosis device using an expert system.

[Explanation of symbols]

1 target equipment, 2 knowledge database, 3 monitoring information,
4 inference engine (failure factor determining means), 5 man-machine device (determination condition defining means), 6 diagnostic rule (determination rule, failure factor determining means), 7 prediction rule (determination rule, failure factor determining means), 8 online system .

─────────────────────────────────────────────────── ─── Continuation of the front page (51) Int.Cl. ⁶ Identification number Reference number within the agency FI Technical display location G06F 11/22 360 D (72) Inventor Mitsuo Abe 1-4, Mei Station, Nakamura-ku, Nagoya Tokai Passenger Railway Co., Ltd. (72) Inventor Genji Torii 1-4-1, Mei Station, Nakamura-ku, Nagoya City Tokai Passenger Railway Co., Ltd. (72) Inventor Shinichi Arimura 1-1, Wadasaki-cho, Hyogo-ku, Kobe No. 2 Inside Sanryo Electric Co., Ltd. (72) Inventor Hidekazu Tanaka 1-2 1-2 Wadasaki-cho, Hyogo-ku, Kobe Sanryo Electric Co., Ltd.

Claims (7)

[Claims] 1. A combination of monitoring information such as operation data and failure data regarding target equipment and monitoring information such as test data collected by performing inspection, failure diagnosis is performed based on a judgment rule of a knowledge database, and a failure is detected as a diagnosis result. In a failure diagnosis device using an expert system for specifying a part, a judgment condition defining means for defining a judgment type and a judgment condition for a phenomenon that may occur in the target equipment by a matrix structure, and the judgment condition defining means. A failure factor determination means for determining a failure factor by performing a failure diagnosis at the time of occurrence of a failure for each part of the target equipment and a failure prediction before the failure based on the defined determination type and its determination condition and the determination rule. A failure diagnosis device using an expert system, which is characterized by being equipped. 2. The determination condition defining means, based on the monitoring information obtained from the target facility, determines a diagnosis rule including a determination type and a determination condition to be applied to a phenomenon that is currently occurring, and a lapse of time. It is necessary to define a predictive rule consisting of a condition change state, a degree of an abnormal condition, a probability of occurrence of anomaly in the future, etc., and a prediction rule including the condition for judgment in the case where the above phenomenon changes accordingly. A failure diagnosis device using the expert system according to claim 1. 3. As the judgment type and its judgment condition, a pass / fail judgment applied for making a diagnosis based on binary information with respect to the monitoring information, a condition for performing the pass / fail judgment, and an upper limit and a lower limit are defined. The conditions for performing upper and lower limit judgments and upper and lower limit judgments that are applied to make a diagnosis based on threshold information, and the fuzzy judgments that are applied to make a diagnosis based on a fuzzy function and the fuzzy judgments when performing the fuzzy judgments. Tendency applied to make a diagnosis based on the conditions and human sensory criteria, conditions to make the step judgment, and a tendency to make a diagnosis based on the tendency to change over time Conditions for making a judgment and its tendency judgment, a difference judgment applied for making a diagnosis based on a difference amount between a phenomenon that is currently occurring and a phenomenon that has occurred in the past, and conditions for making the difference determination A little A fault diagnosis apparatus using the expert system according to claim 2, characterized in that it is provided at least. 4. The monitoring information such as operation data and failure data regarding the target equipment, and monitoring information such as test data obtained by performing inspection are collected online. A failure diagnosis device using the expert system according to item 1. 5. The method according to claim 1, wherein monitoring information such as operation data and failure data regarding the target equipment and monitoring information such as test data obtained by the inspection are collected off-line. A failure diagnosis device using the expert system according to item 1. 6. A combination of monitoring information such as operation data and failure data regarding target equipment and monitoring information such as test data collected by performing inspection, and performing failure diagnosis and failure prediction based on the diagnosis rule of the knowledge database, 3. A failure diagnosis device using an expert system according to claim 2, wherein the failure part is predicted before the failure occurs. 7. A combination of monitoring information such as operation data and failure data related to the target equipment and monitoring information such as test data collected by performing inspection, failure diagnosis prediction based on the prediction rule of the knowledge database, and diagnosis prediction 3. A failure diagnosis device using an expert system according to claim 2, which predicts a part that will fail as a result.