JPH06301878A - Maintenance guidance system - Google Patents

Abstract

(57) [Abstract] [Purpose] Infer an anomaly cause that may be related to the anomaly from anomaly input information from the target device a1 using an anomaly cause inference knowledge base a2-4 and an inference unit a2-3. A maintenance guide system that displays all of these causes on the CRTa3 is displayed from the most probable cause of the abnormality, and the cause is grasped and the recovery is speeded up. [Arrangement] An abnormality cause occurrence time storage means a2-5 is provided to store the occurrence time of the abnormality cause so far, and the abnormality cause display order updating means a2-6 may be used for displaying the abnormality cause. From the most probable cause of anomaly, for example, the anomaly cause with the largest sum of the reciprocal of the time difference between the past occurrence time and the current time is displayed in an array, or the average occurrence up to the previous time for each anomaly cause is displayed. The time difference is displayed in order from the largest value obtained by dividing the elapsed time from the previous failure occurrence time to the present.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention is directed to inferring a cause of an abnormality and instructing a method of recovering the abnormality cause based on a knowledge base for inferring an abnormal state and an abnormal cause of a target device. The present invention relates to a target device maintenance guide system. In the drawings below, the same reference numerals indicate the same or corresponding parts.

[0002]

2. Description of the Related Art FIG. 1 is a block diagram of a maintenance guide system as an embodiment of the present invention. The conventional technique will be described below with reference to this drawing. It should be noted that conventionally, the means a2-5 and a2-6 in FIG. This Figure 1
In the above, the maintenance guide system includes a target device a1 which is a target of the maintenance guide, a personal computer main body a2, and a CRTa3. The personal computer main body a2 includes an input unit a2-1 for inputting the state of the target device, a man-machine interface unit a2-2, an inference unit a2-3, and a knowledge base a2-4 for inferring an abnormality cause.
Further, the CRTa3 can display an abnormality cause display screen a3-1, an inspection instruction display screen a3-2, an inspection result input screen a3-3, and a recovery method display screen a3-4.

In this maintenance guide system, information on the abnormality of the target device a1 is received from the state input section a2-1 of the target device. Then, the information is transferred to the man-machine interface a2-2 and the inference unit a2.
Pass to -3. The inference unit a2-3, which has received the abnormality information of the target device, infers the cause of the abnormality using the received abnormality information and the knowledge base a2-4 for inferring the abnormality cause, and uses the inference result as the man-machine interface unit a2-. Pass to 2. The man-machine interface part a2-2 displays the received inference result on the display screen a3-1 of the cause of abnormality of the CRTa3.
To display. Then, when there is a request from the operator to display the inspection content of the abnormality cause, the inspection instruction screen a3-2 for each abnormality cause is displayed on the CRTa3, and the inspection result input screen for the operator to input the inspection result. a3-
3 is displayed. Then, it is determined from the input inspection result whether or not the cause of the abnormality is really abnormal. If it is not abnormal, it is deleted from the abnormality cause list, and if it is abnormal, a recovery method screen a3-4 for the abnormality is displayed.

FIG. 6 is a flow chart for explaining the operation of a conventional maintenance guide system. Note that b1 to b1
The reference numeral 0 indicates a step in the figure. The software of the maintenance guide system is usually
The abnormality information of the target device is read (b1) and it is determined whether an abnormality has occurred (b2). When no abnormality has occurred (b2, branch NO), reading of abnormality information (b1) and determination of whether an abnormality has occurred (b2) are repeated until the abnormality occurs. When an abnormality occurs (b
2, branch YES), the reason why the abnormality has occurred, that is, the cause of the abnormality is inferred based on the knowledge base (b3). At this time, the cause of the abnormality, which is the inference result, is not limited to only one, but all possible causes are obtained with high probability (confidence). Then, the cause of the abnormality, which is the result of the inference, is displayed on the screen a3-1 (b
4). And the instruction from the operator is accepted (b5),
If the instruction is an instruction to display the inspection method for the cause of the abnormality (b6, branch NO), the inspection instruction screen a3-2 is displayed (b7), and the inspection result is input to the operator using the screen a3-3. (B8), it is judged whether the cause of the abnormality inferred from the input result is really the cause, and if it is normal, it is deleted from the list of causes of abnormality (b9). Is displayed (b4). If the operator's instruction is a display request for a recovery method instruction screen for the cause of the abnormality (b6, branch YE
S), and a recovery method screen a3-4 for the corresponding cause of abnormality is displayed (b10).

[0005]

Since the conventional maintenance guide system displays a plurality of causes of anomalies inferred from the knowledge base in the order of high certainty assumed at the time of construction of the knowledge base, There is a possibility that it does not sufficiently match the actual cause of the abnormality. For this reason, it is necessary to reflect the actual occurrence result of the abnormality cause after the knowledge base is constructed in the knowledge base, and it is necessary to rewrite the priority order of displaying the abnormality cause in the knowledge base. And there was a problem that manpower was required to rewrite the knowledge base. Therefore, it is an object of the present invention to provide a maintenance guide system that can solve this problem.

[0006]

In order to solve the above-mentioned problems, a maintenance guide system according to a first aspect of the present invention is based on input information at the time of an abnormal state of a device to be maintained (such as a target device a1). , Inferring the cause of the abnormality (via the inference unit a2-3) from the knowledge base (a2-4 etc.) for inferring the cause of the abnormality,
In a maintenance guide system that displays (on CRTa3 or the like) the input information and all the causes of abnormality that may be related to this abnormality, a means (a2-5, etc.) for storing the occurrence time of the abnormality cause, and an abnormality in this storage means. Anomaly cause display order that calculates the magnitude of this possibility of anomaly causes that may be related to the anomaly from the data group of the occurrence time for each cause, and displays the above in the order of the probability. Update means (a2-6 etc.) shall be provided.

Further, in the maintenance guide system according to a second aspect, in the maintenance guide system according to the first aspect, the abnormality cause display order updating means indicates the reciprocal of the time difference between the past occurrence time and the current time for each abnormality cause. The sum is calculated, and the display order of the causes of abnormality is rearranged in the order of magnitude of the sum of the reciprocals.

Further, in the maintenance guide system according to claim 3, in the maintenance guide system according to claim 1, the abnormality cause display order updating means calculates an average occurrence time difference of the abnormality causes up to the last time for each abnormality cause, and The display order of the cause of abnormality is rearranged in the order of magnitude of the value obtained by dividing the elapsed time from the failure of (1) to the present by the average difference in occurrence time.

[0009]

According to the present invention, in the main body a2 of the maintenance guide system shown in FIG. 1, the storage means a2-5 for storing and storing the time of occurrence of each abnormality cause, and each abnormality cause in this storage means a2-5. From the past occurrence time data group of each, the magnitude of each possibility is calculated for all the abnormality causes that are inferred to be related to the relevant abnormality of the target device, and the abnormality cause is displayed in order of this magnitude. An abnormality cause display order updating means a2-6 is provided.

In the invention according to claim 2, the abnormality cause display order updating means a2-6 has the past occurrence time and the present time of the abnormality cause already recorded in the storage means a2-5 according to the abnormality cause obtained as a result of the inference. The sum of the reciprocals of the time difference from the time is calculated, and the causes of the anomalies are arranged and displayed in descending order of the sum of the reciprocals. As a result, the cause of abnormality that has recently occurred or the cause of abnormality that has occurred many times can be displayed with priority.

In the invention according to claim 3, the abnormality cause display order updating means a2-6 calculates an average occurrence time difference of the abnormality causes up to the previous time for each abnormality cause obtained as a result of the inference when the abnormality occurs. , The causes of the abnormalities are arranged and displayed in descending order of the value obtained by dividing the elapsed time from the previous failure to the present by this average occurrence time difference. As a result, it is possible to give priority to and display the cause of an error in which the elapsed time from the previous failure to the present exceeds the average time difference of occurrence of the error cause or the error cause close to the average time difference of occurrence of the error cause.

[0012]

Embodiments of the present invention will be described below with reference to FIGS. (1) About Embodiment 1: FIG. 2 is a flow chart of the maintenance guide system in the first embodiment of the present invention. The reference numerals b1 to b10 and c1 and c2 indicate steps in the figure. The major difference between the conventional flow chart described in FIG. 6 and FIG. 2 is that the abnormality cause occurrence time storage means of FIG. a
After saving (accumulating) in 2-5 (c2) and inferring the cause of the abnormality (b3), the cause of the abnormality is saved in the above-mentioned recording medium via the abnormality cause display order updating means a2-6 in FIG. The reason is to sort (classify) the causes of abnormality and change the display order (c1) using the past information regarding.

Next, FIG. 3 and FIG. 4 are respectively related to the saving of the information on the cause of abnormality to the recording medium (c2) and the change of the display order by sorting the cause of abnormality using the past information on the cause of abnormality (c1). I will explain it in detail. FIG. 3 is a flowchart of software that records the date and time when the cause of the abnormality occurred. Here, the reference numerals c2-1 to c2-4 indicate steps in the figure. First, it is checked whether or not the inspection result of the inspection item for each cause of abnormality is input (c2-1). If the inspection result has been input (branch YES), it is determined whether or not the corresponding abnormality cause inferred from the inspection result is the cause of the actual abnormality (c2-2). If it is determined that an abnormality has occurred due to that reason (c2-2, YES branch), the current date and time is recorded as the date and time of occurrence of the relevant abnormality cause (c2-
3). Then, the above procedure is repeatedly carried out for all the causes of abnormality inferred this time (c2-4).

When the inference of the cause of abnormality in FIG. 2 is completed (b
In 3), the causes of abnormalities are arranged in the order inferred based on the knowledge base data. That is, they are arranged in the order of registration of the causes of abnormality in the knowledge base. The sum of the reciprocals of the time difference between the past occurrence time and the current time is calculated for each anomaly cause obtained as a result of inference, and the anomaly causes are arranged and displayed in descending order of the sum of the reciprocals.

Calculation of the sum total of the reciprocal of the time difference between the past abnormality occurrence time and the current time for the same abnormality cause is shown in FIG.
Will be explained below. Here, d0 to d2 and e
The symbols 1 to e3 indicate the occurrence time of the abnormality in the figure and the time difference between each abnormality occurrence time and the current time. First, when the time difference between the past abnormality occurrence time and the current time is, for example, the time difference e1 between the previous abnormality occurrence time d1 and the current time d0, the previous abnormality occurrence time d1 may be subtracted from the current time d0. That is, the time difference e between the previous abnormality occurrence time d1 and the current time d0
1 is obtained by Expression 1.

[0016]

## EQU00001 ## e1 = (current time d0)-(previous abnormality occurrence time d1) [Equation 1] Similarly, the time difference e2 between the abnormality occurrence time d2 two times before and the current time d0 is obtained by Equation 2.

[0017]

## EQU00002 ## e2 = (current time d0)-(previous abnormality occurrence time d2) [Equation 2] Further, the sum S1 of the reciprocal of the time difference between the past abnormality occurrence time and the current time is obtained by Equation 3.

[0018]

## EQU00003 ## S1 = (1 / e1) + (1 / e2) + (1 / e3) + ... [Equation 3] Then, for each abnormality cause in the inference result, the past abnormality occurrence time and the current time are calculated. The sum S1 of the reciprocal of the time difference is obtained, and the causes of abnormalities are arrayed and displayed in the order of increasing S1. In the present invention,
The larger the reciprocal of the time difference between the past abnormality occurrence time and the current time, that is, the closer the previous occurrence time d1 is to the current time d0, the smaller the time difference e1 becomes. To do.

In addition, since the time difference from the current time is calculated for all past abnormality occurrence times and the sum of the reciprocals thereof is calculated, if there is no other recently caused abnormality cause, the abnormality cause with a large number of occurrences is given priority and displayed. To do. (2) Regarding Example 2: The basic flow chart is the same as in FIG. The difference lies in a method of sorting the causes of abnormalities and changing the display order (c1) by using past information about the causes of abnormalities saved in the recording medium after inferring the causes of abnormalities (b3). The saving (c2) of the information on the cause of the abnormality in the recording medium is the same as in the first embodiment, and the time of occurrence of the cause of the abnormality is recorded.

The change of the display order (c1) by sorting the cause of abnormality using the past information about the cause of abnormality in the second embodiment will be described with reference to FIG. FIG. 5 is an explanatory diagram of the calculation of the average occurrence time difference up to the last time for the same cause of abnormality. The symbols d0 to d3 and f0, g1 to g3 are the time when the abnormality occurs in the figure,
And the time difference between each abnormality occurrence time. First, the average occurrence time difference S2 of the abnormality causes up to the previous time is the sum of the time differences of the occurrence times of the abnormality causes up to the last time (g1 + g2 + g3 ...).
Can be divided by a value obtained by subtracting 1 from the number n of occurrences of abnormalities in the past. That is, it is obtained by the equation 4.

[0021]

[Equation 4] S2 = (g1 + g2 + g3 ...) / (Number of abnormal occurrences n-1) [Equation 4] Also, a value obtained by dividing the elapsed time f0 from the previous failure to the present by the average occurrence time difference S2 of the above-mentioned abnormal causes. S3 is equation 5
Required by.

[0022]

[Equation 5] S3 = f0 / S2 [Equation 5] Then, a value S3 obtained by dividing the elapsed time f0 from the previous failure to the present by the average occurrence time difference S2 of the above-mentioned abnormality causes for all the abnormality causes of the inference result. The abnormal causes are arranged and displayed in descending order of S3.

[0023]

【The invention's effect】

(1) Regarding the invention relating to claim 2, by storing the time of occurrence of the cause of the abnormality, it is possible to accumulate information relating to the abnormality unique to the target apparatus. In addition, the sum of the reciprocals of the time differences between the past occurrence time and the current time of the anomaly cause already recorded for each anomaly cause obtained as a result of inference is calculated, and the anomaly causes are arranged and displayed in descending order of the sum of the reciprocals. As a result, the cause of abnormality that occurred first recently can be displayed with priority on the cause of abnormality that occurred next, so that the cause of abnormality that often occurs in succession can be found early.

(2) Regarding the invention according to claim 3, when an abnormality occurs, an average occurrence time difference of the abnormality causes up to the previous time is calculated for each abnormality cause obtained as a result of inference, and the difference from the previous failure to the present time is calculated. By arranging and displaying the causes of abnormalities in descending order of the value obtained by dividing the elapsed time by this average occurrence time difference, the elapsed time from the previous failure to the present is given priority over the abnormal cause that exceeds the average occurrence time difference, and Since the abnormal cause close to the average time difference between the abnormal causes and the elapsed time from the previous failure to the present can be displayed preferentially, the abnormal cause, which often occurs periodically, can be found early.

[Brief description of drawings]

FIG. 1 is a configuration diagram of a maintenance guide system as an embodiment of the present invention.

FIG. 2 is a flowchart for explaining the operation of FIG.

3 is a flowchart of detailed software for recording the date and time when the cause of abnormality in FIG. 2 is generated.

FIG. 4 is an explanatory diagram of the calculation of the sum of the reciprocals of the time difference between the past abnormality occurrence time and the current time as the first embodiment of the present invention.

FIG. 5 is an explanatory diagram of calculation of an average occurrence time difference of the cause of abnormality up to the previous time as the second embodiment of the present invention.

FIG. 6 is a flowchart for explaining the operation of a conventional maintenance guide system.

[Explanation of symbols]

 a1 Target device a2 Personal computer main body a2-1 Target device status input part a2-2 Man-machine interface part a2-3 Inference part a2-4 Abnormal cause inference knowledge base a2-5 Abnormal cause occurrence time storage means a2-6 Abnormal cause Display order updating means a3 CRT a3-1 Abnormality cause display screen a3-2 Inspection instruction display screen a3-3 Inspection result input screen a3-4 Recovery method display screen

Claims (3)

[Claims] 1. A cause of an abnormality is inferred from a knowledge base for inferring the cause of the abnormality based on input information when an abnormality occurs in a device state to be maintained, and the input information and this abnormality are related. In the maintenance guide system that displays all the causes of abnormalities that can be obtained, the means for storing the time of occurrence of the cause of abnormality and the data group of the time of occurrence for each cause of abnormality in this storage means are used to identify the cause of abnormality that may be related to the abnormality. A maintenance guide system comprising: abnormality cause display order updating means for calculating the degree of possibility of this relation and performing the display in the order of the degree of possibility. 2. The maintenance guide system according to claim 1, wherein the abnormality cause display order updating means calculates the sum of the reciprocals of the time difference between the past occurrence time and the current time for each abnormality cause, and A maintenance guide system characterized in that the display order of the causes of abnormalities is rearranged in order of magnitude of total sum. 3. The maintenance guide system according to claim 1, wherein the abnormality cause display order updating means calculates an average occurrence time difference of the abnormality causes up to the last time for each abnormality cause, and an elapsed time from the last failure to the present time. The maintenance guide system is characterized in that the display order of the cause of abnormality is rearranged in order of magnitude of the value obtained by dividing by the average time difference.